结论: 你可以花费大大少于Blue Air, Philips或者IQAir这些大品牌净化器的钱，轻松清除掉你家里的颗粒污染。
方法: Anna在晚上睡觉之前将空气净化器开到最大功率，然后在起床后关闭净化器。为了测量粒子，她使用一个 Dylos 粒子计数器，它能测出每0.01平方米中，0.5微米及以上颗粒和2.5微米及以上颗粒的数量。粒子计数器每一小时测试一次空气。
净化器寿命: 所有测试都结束后，IQ Air显示其前过滤还能使用1,931个小时，活性炭滤网还能使用3,077个小时，HEPA滤网还能使用1,910个小时。因此，净化器的状况还是很好的。
异常值: 在这11个测试中，有一天的数据是很显著的异常值。5月22日，IQ Air仅过滤了68%的0.5微米粒子。一般测试出现很差的结果是因为室外空气在夜间变脏很多。但是5月22日这一天，室外空气中的颗粒浓度在74和110微克之间浮动变化，这并不算反常的情况。
DIY 1.0 在正常大小的房间（15平方米以下）使用，大炮可以用于更大的空间。
1. 这个测试是在白天进行。一些人在直觉上认为夜晚由于人们活动和路面汽车的减少空气污染也会有所下降（但是数据显示这种直觉是错误的——北京午夜的PM 2.5污染水平是最高的）。
就算在一个门窗紧闭的房间， 我们的测试发现，在关掉净化器后，颗粒物将快 上 升。脏空气会持续进入房间，尽管我们用肉眼看不到。测试的结果显示，80分钟 后房间里的污染就达到了原来的水平。
As always, I’m publishing the raw data and more details on the methods below. I’ll also be publishing data from similar tests in a much larger room (30.5m2) and for tests when I’m at home and moving in the room.
Methods: I used a regular timer to have the Cannon turn on between 3pm and 5pm every day. I didn’t choose those two hours for any particular reason.
Room: The room is a 13.5 m2 bedroom in the Chaoyangmen neighborhood of Beijing. The doors and windows were closed at all times.
I did the tests while I was on vacation, so it is a very controlled test, although reductions may take long if people are opening doors and coming in and out of the room. For that reason, I’ve repeated the tests while I was at home, and I’ll be publishing that data soon.
Data: Because the raw data file is very big, I’m making it available as a download here.
How loud is it?
Original: 52dB. Cannon: 56db.
The Cannon kicks butt (scientific definition of kicking butt), but it’s noisier than the Original DIY. How noisy is it? As is my habit, I wanted to answer this question scientifically.
So I bought a decibel meter:
And I tested the Cannon, Original DIY, Blue Air 203/270E, and IQ Health Pro Plus on their highest settings from 1.95 meters away. Here are the results:
The cannon is noisier than I’d like, but it’s similar to the Blue Air on the high setting. To give you an idea of how loud that is, this decibel chart says that’s between “conversation at home” and “conversation in restaurant.”
It’s still louder than I’d like, but fortunately I’ve found that the Cannon is still very effective on the lower settings:
So I recommend running the cannon on a lower setting if you find it noisy.
1. Cannon-owners can use the lower settings without sacrificing much performance.
2. For people who are particularly sensitive to noise, the Original may be a better choice.
3. For people who are VERY sensitive to noise, the Philips AC4072 is expensive (2,700 RMB), but it’s quite quiet on the low setting.
I’m posting the raw data here for the medium and low settings. The raw data for the tests on the high setting are here. All of the methods are identical to my other tests and are described in (excruciating) detail there.
The Cannon and HEPA used in these tests are identical to the kits we are shipping from Smart Air.
根据世界卫生组织的报告，空气污染是世界最大杀手之一，它每年导致约200 万未成年人死亡。许多死亡案例发生在发展中国家(单是在印度每年就有超过50 万人死亡)，然而富裕的发达国家同样也深受其害:例如，据推断美国每年约有 41,000人因空气污染而过早死亡。试想一下，如果200万人(大约相当于美国得克 萨斯州的休斯顿或者英国西米德兰卫星城的总人口)在一次恐怖袭击中或是地震中 丧生，媒体将会如何进行报道。仅仅因为空气污染杀人于无声和持续的过程中，很 难被察觉，它很少得到人们的注意或是关注。
How much does outside air affect indoor air?
I’ve always wondered: how much does outdoor air pollution affect the air in my home? Even if I keep my windows closed, if the air outside gets really polluted, does air in my home get a lot worse?
To answer that question, my Smart Air collaborator Anna took particle counts in her Beijing apartment everyday for two months and compared the counts to outside pollution. Here’s how she did it:
When she got home from work (and before she turned on her DIY filter), she used a particle counter to measure how much particulate matter was in her room and compared it to outside air pollution from the US Embassy Twitter feed. Her windows were always closed, and her home is near Chaoyangmen, which is in central Beijing.
Here’s the data from 41 days:
With this data, we can actually start to predict how polluted the air is inside based on how polluted it is outside. To do that, I removed a few outlying datapoints and plotted a regression line:
So when does indoor air pollution get dangerous?
To answer that question, we first have to define “dangerous.” I use the WHO limit of concentration of 25 micrograms per meter cubed.
Then we need to convert the PM .5 measurements from my Dylos machine into official concentration numbers. That’s tough, but we can get a rough estimate based on my prior tests. That data shows that the WHO limit of 25 micrograms is equivalent to about 4,000 on the Dylos. The graph above shows that the air in Anna’s home is frequently over the WHO pollution limit.
If we round that number up to a more conservative estimate of 5,000, we can reach a rough conclusion: if outside concentration is above 40, the indoor counts are consistently over the 5,000 limit:
A concentration of 40 is an AQI of 112. (Remember than governments convert particle concentrations into AQI numbers.) Therefore, we can conclude that, if the AQI is above about 110, the air inside Anna’s bedroom is more polluted than the WHO limit.
Conclusion: Outdoor air pollution strongly affects indoor air pollution, and indoor air is often more polluted than the WHO limit — even with the windows closed.
Now, remember this data is from one apartment in Beijing. The numbers probably vary a lot between apartments that have better or worse seals around the windows. The numbers probably also vary between downtown and farther out in the suburbs. That said, this data convinces me that indoor air in China can be dangerous.
As always, I’m posting the data for fellow nerds below.
Here’s the raw data:
I’ve highlighted the outlying datapoints in orange. “Concentration” and “AQI” are measurements of outside air from the US Embassy.
All tests were in Anna’s bedroom, which is 15 meters squared. Her apartment is near Chaoyangmen, central Beijing. The outside AQI and concentration come from the US Embassy Twitter feed from the closest hourly timepoint to when Anna took the measurement of indoor air. The US Embassy measures air pollution from their location near Liangmaqiao, in northeastern Beijing.
The windows were always closed, so this represents the effect of outside air on inside air without opening the windows. Anna’s apartment is not particularly modern, so I’m guessing the seal around the windows isn’t great. Numbers might be lower if we tested in a more modern apartment.
Is smoking worse than outdoor air pollution?
A while back, I posted data I collected from places around Beijing showing that indoor air is consistently cleaner than outdoor air. When I analyzed that data, I excluded places that allow smoking, but I’m posting the data here now:
Even on days where outside air was bad (AQI ~ 180), the air was even worse in the cafes that allow smoking. This is even more surprising because:
1. I was not sitting in the smoking section.
2. The smoke was not very noticeable. (The air seemed good enough to me that I sat there and worked — and I hate smoking.)
My guess is things are much worse in smoky bars and clubs, where the smoke is so thick my clothes smell like smoke the next day.
Conclusion: Indoor air is better than outdoor air in China, but you lose any advantage once people start smoking — even if you’re in the non-smoking section.
Is air pollution just a Beijing problem?
Beijing’s air pollution is the most famous in China, but that can make people in some other cities think their air is good — at least, not as bad as Beijing’s. But is the air in other cities safe?
In 2013, Shanghai had newsworthy air pollution that convinced many people in Shanghai that air pollution was a problem there too. But what about Guangzhou? Guangzhou doesn’t have winter heating. Does that mean its air is safe?
To answer this question, I analyzed all of the hourly data from 2014 from the American consulate in Guangzhou. Year to date, that data covers 4,572 datapoints. Then I calculated what percentage of those had PM 2.5 readings about the WHO 24-hour upper limit of 25 micrograms (which is about 77 on the US AQI scale).
To date, 76.9% of the readings exceeded the WHO upper limit. The average reading was 53.4 micrograms — over two times the WHO limit.
Breathe safe, Guangzhou!
How does air pollution change throughout the day?
Usually lowest in the afternoon, highest at night.
My intuition has always been that air pollution is lowest at night because there are fewer cars on the road and fewer factories humming. Apparently I’m not alone: 139 voted for their guess about what time of day has the lowest PM 2.5, and night time came in first:
I also know people who arrange their schedules to work out in the morning to avoid the worst pollution. But how accurate are our intuitions?
To get to the bottom of it, I analyzed thousands of hours of PM 2.5 data from the US Embassy in Beijing. When I finally got the answer, I was surprised:
Instead of during the quiet of night, it’s the afternoon — right around rush hour — that PM 2.5 is the lowest. So if you’re planning a picnic or insist on exercising outside, you’re usually best off between noon and 6pm.
What about other cities?
Is that how PM 2.5 generally works, or is it unique to Beijing’s activity or climate? Fortunately, US consulates in several other cities publish their historical data.
In terms of climate, Shenyang is pretty similar to Beijing, and its daily PM 2.5 patterns are very similar (although night time isn’t quite as bad as Beijing):
The pattern in Chengdu is similar. PM 2.5 is lowest in the afternoon and highest in the morning.
The picture starts changing when you get to the south. In Shanghai, PM 2.5 is lowest in the afternoon (like the north), but there’s also a dip in the very early morning:
Guangzhou’s pattern was virtually identical to Shanghai’s: a dip in the very early morning and a dip in the afternoon.
Bottom Line: When is it safest to be outside?
In all five cities, the afternoon had the lowest PM 2.5 levels. And in contrast to many people’s intuition, the night time had the worst air in several cities. Thus, you’re usually best off organizing your picnic or tai chi in the afternoon.
But keep in mind that the air is NEVER safe on average in any of the five cities at any time of day. So take “safest” with a grain of salt!
Does Chinese New Year affect inside air quality?
I’ve posted data before showing that outdoor air quality is strongly correlated with indoor particle counts (r = .71), but Chinese New Year gives nerds like me a great chance to see what happens when we get a momentary shock to air quality.
The media made a big deal about people cutting back on fireworks this year out of a concern for air quality, and that may be true, but you can still see a strong spike in PM 2.5 as Beijingers rang in the year of the horse:
Not all that surprising. But what’s more interesting is that you can see a corresponding increase in the particle counts in my collaborator Gus’s bedroom:
These indoor counts are without a purifier running, so they demonstrate how quickly outdoor air pollution can find its way indoors and how variable indoor air quality can be in a single room over time. Simply put: the worse the air is outside, the worse it is inside.
A couple of notes for fellow nerds:
1. The indoor particle counts are not precisely on the hours, so the apparent time lag between indoor and outdoor counts may be exaggerated.
2. The early spike in indoor 2.5 micron particles may be because people were moving around the house at that time, which affects the larger 2.5 microns more than the smaller .5.
What happens to a room with no purifier in it?
I usually test air purifiers by taking a baseline measurement of particulate pollution in a room, and then turning on the purifier and testing whether the counts drop. My collaborator Gus once suggested another method: run one particle counter in the bedroom that has the purifier, and run another particle counter in a different room that does NOT have a purifier. The benefit of this method is that the control room represents the counterfactual — what would have happened if we hadn’t turned on the air purifier.
Thus, if a northwest wind hits Beijing and makes the outdoor air a lot cleaner, we can separate the effect of the outdoor air fluctuations from the effect of the purifier. In that situation, my old method would artificially raise our estimates of effectiveness. Changes in outdoor air can also artificially lower our estimates of effectiveness if the outdoor air gets dirtier after we turn on the purifier.
In previous tests, I corrected for this by averaging over multiple tests. I also analyzed the data after removing days in which outdoor air pollution fluctuated a lot (for example, I do that sort of analysis in the extra nerd notes here).
But it’s always nice to use different types of tests to make sure an effect is real, so Gus did this experiment. He set up one particle counter in his room and one in his kitchen:
He let the particle counters run for several hours, and then a timer turned on the Original DIY in his room. (The kitchen had no air purifier.) Here’s what happened:
The difference between the bedroom and the kitchen air quality can approximate the effect of the air purifier. It looks like Gus would have been breathing 16,000 .5 micron air in his bedroom if he hadn’t turned on his DIY purifier.
And it’s pretty clear that the kitchen air quality (where we don’t have a purifier running) is following outdoor air quality:
(Be aware that I’m overlaying these two lines on the same graph, but the Y-axes are different. This is NOT saying that indoor air is as bad as outdoor air. Indoor air is usually cleaner than outdoor air.)
Conclusion: Similar to earlier tests, the double particle counter test shows that the DIY purifier is removing particulate pollution from the air.
What about 2.5 micron particles? In the main text, I present the .5 micron results, but what about the larger particles? The results there are similar:
Remember that because these particles are larger, they are more affected by people moving around. That’s probably why there are large spikes from 7 to 9pm, and why counts level off after midnight. However, even after the numbers level off, the purifier is reducing the counts in bedroom even further.
Timing: Careful readers will notice a slight lag between the bedroom and kitchen numbers. That’s because after entering its hourly mode, the Dylos particle counter takes its first measurement after a random number of minutes. (After that, it’s every 60 minutes.)
Therefore, we cannot guarantee the kitchen and bedroom numbers are taken at the exact same time. However, after the fact, we can tell when the measurements were taken. Thus, we know that the bedroom numbers were taken at 27 minutes past every hour, and the kitchen numbers were taken at 36 minutes past the hour. Those numbers are reflected on the X-axis.
Percentage effectiveness: I also computed the percentage reductions in particulates using my regular method of averaging the last four hours before the DIY was turned off. Then I compared that as a percentage of particulate in the kitchen.
According to that calculation, the DIY removed 86% of 2.5 micron particles and 87% of .5 micron. Compare that to 92% of 2.5 micron and 84% of .5 micron in my earlier tests. However, I am cautious about drawing strong conclusions from one day’s data. It would be best to average the results of several testing days.
Has anyone else tested the DIY purifier?
Beijing-based Dr. Saint Cyr’s tests of air purifiers were one of my original inspirations for the whole DIY project, so I was happy when the DIY became a part of those tests:
These are the first independent tests of the DIY, and the results parallel mine. You could also include the tests by doctors at the University of Michigan as independent “proof of principle,” although they used a different fan and filter. The commonality is that all of the tests have shown that a simple filter and a fan can reduce particulate pollution in the home.
Dr. Saint Cyr’s review isn’t all glowing. He rightly notes that the cannon is noisy, which I’ve also written about (decibel counts and comparisons here). Tests show the cannon is still very effective on the lower settings, so I recommend running the cannon on the somewhat quieter settings. And for people who are sensitive to noise, I recommend the quieter Original.
Are all DIY purifiers the same?
When I did my first experiments, several people told me not to publish the data. “Don’t give it away for free,” they told me. “Use it to make money!”
I decided then that my main goal wasn’t to make money. I almost got tricked into paying $1,000 for clean air, and I wanted to help people avoid getting tricked too, so I published the data anyway.
Of course, publishing the instructions online has made it easy for people to copy the idea. 还我蓝天 (Huanwo Lantian) was one of the first to follow in our footsteps, selling a DIY filter a few months back. They even use a screen capture of Gus’s appearance on Chinese TV on their shop:
I was curious to see how their filter works, so I decided to order one off of Taobao and put it to the test.
Now I’m in an awkward position because I found that their HEPA was not working nearly as well as my Smart Air HEPAs. It’s awkward because, if I publish the data, will people think I’m just trying to attack a competitor?
In the end, I think it’s better to publish the results and be honest about my conflict of interest. At the very least, I think people have an interest in knowing how well other DIYs work — especially when some of those websites use graphs that are lifted from my site, which can mislead people into thinking the test results are from their machines.
And as always, I’m publishing my raw data and testing methods at the end of this post, so fellow nerds can replicate my studies.
Method: My collaborator Anna used the same methods as our earlier tests in her 15m2 room. Anna did five overnight tests with the same Dylos Pro particle counter, and I calculated effectiveness as the percent reduction of particles in the room air, averaging the last three hours (more info). Then I compared the results to my earlier tests in the same room.
Here’s what I found:
Results: The 还我蓝天 DIY removed 21% fewer particles .5 microns and above and 11% less 2.5 micron particles than the Original.
Is it the fan? The 还我蓝天 fan is slightly smaller than the Smart Air Original, so one explanation could be that the 还我蓝天 fan is just moving less air. Anna tested that by strapping the 还我蓝天 HEPA onto our Original fan.
Here’s what I found:
Result: There wasn’t much difference. With the new fan, it was getting 4% more PM .5 and 2% less PM 2.5. Thus, the fan doesn’t seem to be the reason.
Is it the HEPA? The second possibility is that the 还我蓝天 HEPA isn’t as good. Anna tested the HEPA by doing air outlet tests with a Met One GT-521, which measures down to .3 microns. Anna tested the air coming out of the HEPA for 10 seconds, and I averaged the results from three tests. (More details at the end of the post.)
Here’s what I found:
Results: The 还我蓝天 HEPA performed worse, about 7% lower than the HEPA standard. The major source of the 还我蓝天’s poor performance seems to be the quality of the filter.
Conclusion: In room tests, the 还我蓝天 DIY removes about 21% fewer particles than the Original DIY, and the data suggests that the reason is that 还我蓝天 HEPA is lower quality.
Now, don’t get me wrong. The 还我蓝天 DIY is making the room air cleaner. I’d rather have a 还我蓝天 than nothing. But the results show that this DIY copycat is cutting corners by using cheap HEPAs.
For the methodology of the overnight room tests, see my earlier post. Anna ran five tests of the 还我蓝天 DIY and six tests of the 还我蓝天 HEPA + Original fan. The raw data is below.
1. The Met One gives counts for particles .3 microns and above, which is important for HEPA tests because .3 microns is the particle size that is hardest for HEPAs to capture.
2. The Met One uses a pump, which is better for regulating air flow than the fan in the Dylos. I think the fan is adequate for most tests, but pumps help keep a steady air flow when the particle counter is in a stream of quickly moving air (which is what happens during HEPA tests).
Cumulative numbers: All of the particle counters I’ve used display the current number of particles, but the GT-521 displays the cumulative number of particles, so the numbers keep increasing, even if the room air is getting cleaner.
To deal with that, I divided the cumulative number of particles by the number of seconds in the test.
Baseline: To get a baseline number, Anna held the particle counter about a foot away from the purifier while the purifier was off. Anna ran the particle counter for 10 seconds and divided the cumulative number of particles by 10.
Timing: The particle counter starts immediately after hitting the start button, but starting the test right away is a bad idea because we want to avoid moving during the test. Therefore, Anna hit start, moved the counter so that the inlet was about half an inch from the center of the HEPA, and then waited until the particle counter collected a fresh 10 seconds of data.
Anna set the particle counter to collect data for 10 seconds, pause for 10 seconds, and then collect data for another 10 seconds. Anna used the latter 10 seconds so that we could avoid the effect of moving the particle counter into place.
Calculation: Anna repeated all of these steps three times for each HEPA, and I averaged the results of the three tests.
Room Test Raw Data
Here is the raw data for the 还我蓝天 plus Original fan:
Fluctuations in Outdoor Air Pollution
The outdoor air seemed to fluctuate more widely in these tests than during most of my earlier tests. So to test for the effect of outdoor air fluctuations, I analyzed days where the outdoor air was above the WHO limit (25 micrograms) and didn’t fluctuate more than 15 micrograms. Data from 3/16 met those criteria, starting at 149 micrograms and fluctuating a maximum of 6 micrograms.
On that day, the 还我蓝天 removed 66% on PM .5 and 77% of PM 2.5. That’s within 2% of the overall average of 67% and 79%.
I’ve had similar conclusions when I did this type of analysis in my previous tests: changes in outdoor air pollution can strongly affect data on any single day, but averaging the data over multiple days leads to surprisingly robust estimates.
HEPA Test Raw Data
Are DIY purifiers still cheap in the long run?
When I published tests of the DIY compared to the expensive machines, a couple people on Zhihu (China’s Quora) asked if you really save money in the long run with the DIY once you calculate in the cost of changing the HEPA. After all, the IQ Air HEPA costs 1,782 RMB, but you can use it for more than a year.
Long-Term DIY HEPA Costs
At that time I didn’t have an answer. To get an honest answer, I needed to do tests in the real world, but that took almost a year to complete. (That’s more than I can say for the numbers IQ Air and Blue Air give. If filters last 6 months in Sweden, are they going to last 6 months in Beijing?)
Now it’s a year later, and I have that data. Smart Air co-founder Gus ran his Original DIY for eight hours a day and tracked what percentage of particulate it removed from the air each day with a Dylos particle counter.
Based on that data, I recommend changing HEPAs after 140 days at 8 hours per day (about 1,000 hours of use). Since that HEPA costs 80 RMB, that averages to .57 RMB per day and 208.6 RMB per year.
To get an idea of what that means, we can compare that to the cost of drinking a bottle of water a day:
Big Brand HEPA Costs
I’m highly skeptical that HEPA recommendations for Sweden can be mapped onto China. But to be conservative, I’ll assume in my calculations that their numbers work the same in China.
I calculated costs for the Blue Air 203 and IQ Air Health Pro Plus using the same conditions I used for the Original DIY longevity test:
A. Highest setting
B. 8 hours of use per day
C. Extend those costs over a year
D. Include the cost of pre-filters but not carbon filters
Blue Air’s HEPA costs 300 RMB and lasts 6 months. That works out to 1.67 RMB per day and 609.6 RMB per year.
IQ Air’s HEPA is more expensive: 1,782 RMB. On the sixth setting, it lasts 4,968 hours. If you use it 8 hours a day, that works out to 2.87 RMB per day and 1,047.4 RMB per year.
However, we have to add the cost of the pre-filter because the HEPA lifespan depends on the pre-filter. The pre-filter costs 645 RMB and lasts 2,016 hours on the sixth setting. That works out to 2.56 RMB per day and 934.3 RMB per year.
IQ Air grand total: 1,982 RMB per year.
The long-term HEPA cost for the Original DIY is 66-89% less than the Blue Air and IQ Air.
It should be noted that these results may not map onto the Cannon. I’m still working on the Cannon longevity test, so we’ll need to wait until that test is done.
DIY HEPA Costs
DIY HEPA costs are based on the 200-day longevity test in real Beijing air. All 200 days of original data are available in that post.
IQ Air Costs
One thing to note: the calculations above are based on the sixth setting. If the IQ Air runs on setting three, the effectiveness will be lower, but the HEPA will last longer. In my calculations, I totaled the costs for one year one setting three: 901 RMB. That’s lower than the calculation above (because it’s processing less air), but it’s still more than four times the cost of the DIY HEPA.
Blue Air Costs
Calculating costs for the Blue Air is more difficult. That’s because Blue Air doesn’t seem to publish longevity recommendations for different settings or how many hours you can use the HEPA. Instead, Blue Air says you should replace the HEPA after six months.
Thus, I sent Blue Air an email asking how many days I should replace the HEPA if I run my machine eight hours a day. A Blue Air representative told me six months and told me I would void the warranty if I used it more than six months. Thus, I used their recommendation in my calculations (original email).
If Blue Air is wrong and the HEPA lasts long at eight hours per day, the calculation changes. I have alternative calculations here.
Are cheaper HEPAs possible?
Yes, but at a cost.
After I published directions for how to make your own purifier, people asked me: “Which HEPAs should I use? Is this one trustworthy?”
That’s the type of question that makes a nerd like me happy because it means we need to get more data. So I ordered HEPAs from every manufacturer I could find, and I tested them all. After all the testing was done, I found I could ship the HEPAs that worked best for 80 RMB, which was cheaper than the 110 RMB HEPAs I was buying — quality and price!
Can HEPAs be even Cheaper?
Later I found HEPAs for 20 RMB wholesale. I was excited. If HEPAs are that cheap, we can make the DIY even cheaper!
But the test data was terrible. These HEPAs weren’t anywhere close to getting 99% of particles, so I passed on them. If didn’t want to use it in my home, why would I want to ship them to other people?
A 99.97% HEPA for 30 RMB?
Thus I wasn’t surprised a couple weeks ago when I saw a store on Taobao selling HEPAs for 30 RMB and claiming that my HEPAs are 暴力 (aggressively overpriced). They claim that their HEPAs get 99.97% of particles, and if that’s true, maybe these were the holy grail of HEPAs I was looking for all along!
So I ordered two online and put them to the test. The first shock was seeing that it doesn’t have a frame:
That makes it harder retain its shape, but it might still be effective without a frame, so I put it through the tests.
Air Outlet Test
First, I tested it by putting it on the Cannon and testing the air coming out of the HEPA with a Met One Aerocet 531S. (The Met One is useful because it has a pump to regulate airflow. In air outlet tests, the particle counter is sitting in a stream of air, so using a pump maintains constant readings.)
The results weren’t pretty. Smart Air HEPAs got over 99.9% of particles, but the 30 RMB HEPA was below 90% — far below their claim of 99.97%.
But particle effectiveness isn’t everything. A HEPA in the 80% range might work better if it has better air flow. In that case, maybe the HEPA could process the air more times and clean the room air as well as a real HEPA.
To test that possibility, I put each HEPA on the Cannon and used a tool to measure air speed (fancy name “anemometer“). I placed the anemometer on the HEPA at four locations (left, right, top, bottom) and took the average air speed.
Again, the results weren’t pretty. So not only was the 30 RMB HEPA getting far fewer particles, it was letting much less air through.
Quality HEPAs for 30 RMB are still a dream. They’re not useless, but using this 30 RMB HEPA would expose people to significantly more particles.
I still hold out hope that manufacturers will be able to innovate cheaper HEPAs without sacrificing quality, but I haven’t seen those HEPAs yet.
Is the Taobao Store Owner Being Dishonest?
The 30 RMB HEPA store makes claims that their HEPAs get 99.97% of particles, and the data clearly contradicts that, so it’s tempting to think that they’re lying.
But are they? I don’t know what’s in their mind, but my guess is that they simply didn’t go through the hassle of buying a particle counter and testing the HEPAs. I suspect that half of what seems like dishonesty on Taobao is actually just sloppiness.
Air Outlet Test
For the air outlet test, I used the Met One Aerocet 531S. Similar to my test of another sub-par imitation DIY, I took readings of the air at the inlet as a baseline, then sampled air coming out of the HEPA for one minute. I tested the HEPAs on the Cannon on its high setting.
The Met One measures particles .5 microns and above. Ideally the test should use a particle counter that measures down to .3 microns because that is the hardest particle size for HEPAs to capture. However, I am still in the process of getting a particle counter that measures to .3 microns (they’re not cheap!). But given the fact that .5 microns is a slightly lower bar, it’s hard to imagine that the 30 RMB HEPA would score better on even smaller particles.
I used the particle count mode rather than the microgram-estimation mode. Micrograms are far too coarse to detect differences between HEPAs. Microgram modes will often read zero even when placed in front of a poor-quality HEPA.
Does that mean microgram particle counters are wrong? Not quite. A reading of zero micrograms does NOT mean the machine has detected zero particles. If you switch it to particle count mode, you’ll see that the machine is still detecting plenty of particles. It reads zero micrograms because the algorithm it’s using to convert to micrograms is below the threshold for micrograms.
Here is the raw data for the air outlet test:
Air Speed Test
In the air speed test, I placed the HEPAs on the Cannon on its high setting. I put the anemometer directly on the HEPA and let it remain until the reading was stable (about 15-30 seconds). I oriented the Cannon as horizontally as possible (air speed readings are slightly lower if the Cannon is pointed directly up).
I tested the lower left, lower right, upper left, and upper right. Then I averaged the values. There is slight variation in air speed between the different areas of the HEPA.
Does carbon really work?
HEPAs do an amazing job at removing particulates, but particulates aren’t everything. Gases like volatile organic compounds (VOCs) are so small that they slip through HEPAs. “VOC” is a big category, including things like benzene and formaldehyde. VOCs can cause cancer, throat irritation, dizziness, and other not-fun side effects.
And for homes that have VOCs problems, we’re supposed to use activated carbon filters. But do they actually work?
When I published the instructions for how to build your own purifier, I wasn’t comfortable recommending activated carbon because I hadn’t tested it, and I didn’t want to just trust what the big filter companies say. So this year, I set out on a journey to test whether carbon actually works.
I soon learned that gas testing is not easy. First off, “harmful gases” is not a natural category. You can buy a particle counter that will detect all particles of a certain size, but there is no detector that will detect all gases. Instead, you need one for each type of gas, and that is not cheap.
In this case, my scientific curiosity cost me $3,542 for this Industrial Scientific Ibrid MX6. It detects VOCs, carbon monoxide, sulfur dioxide, hydrogen sulfide, and nitrogen dioxide. It uses a photo-ionization detector to measure VOCs from 0-2,000 parts per million with a resolution of .1.
Next, I needed a source of gas pollution. Interestingly enough, my apartment didn’t have enough VOCs to register on the MX6 — nor did 8 other Beijing apartments I tested. (That speaks to whether purifier companies should scare people into thinking that everyone needs carbon.)
I ran four control tests with a fan but no filter. That way the room still has air flow, but no carbon. The cigarettes burned out after about 15 minutes, and I left the fans on for another 30 minutes.
Looking at the Cannon + carbon alone after the cigarette extinguished, the VOCs dropped. Here is the data from one test:
From there we can zoom out to include the time the cigarette was burning and the control fan-only condition. From there, it becomes clearer that the carbon was removing VOCs above and beyond just having a fan on.
Averaging across all of the tests, the VOCs reached a maximum of about 1 ppm while the carbon was on. Without the carbon, VOCs reached 1.5 ppm.
After the cigarette burned out, the Cannon cleared the air of VOCs in 5 minutes to just over 15 minutes on average. Without carbon, the air still had VOCs after 30 minutes.
Averaging over all of the tests, the carbon removed 38% of VOCs by the time the cigarettes burned out compared to the fan-only condition. The Cannon removed 68% after another 5 minutes and 100% by 20 minutes after the cigarettes burned out.
Does everyone need carbon?
Purifier companies have a financial incentive to convince people to buy carbon filters. They make more money if they can scare people into buying carbon. But does everyone need carbon?
I don’t want to scare people into buying carbon. Why? In most homes, my MX6 detected zero VOCs. I only found VOCs in homes that had an obvious source of pollution such as remodeling or smoke. And in all of the places where the MX6 detected VOCs, I was able to notice a chemical smell. That makes sense, since lots of VOCs have noticeable smells, like benzene, toluene, and formaldehyde.
Now, I don’t want to say that VOCs are never a problem. Photo ionization detectors like the MX6 are not the most sensitive test type out there. I bet I’d pick up small amounts of VOCs if I sent air samples to a laboratory.
However, scientists have used fancier methods and found similar results. For example, scientists in Hong Kong tested homes and found that most non-smoking homes did not have un-safe levels of VOCs. We need more tests like this.
For now, I will not be using carbon in my home, but I think it is right for people whose homes have:
1. Recent remodeling
2. Recent painting
5. Nearby sources of gas pollutants (such as living near a factory)
6. Symptoms such as inflammation and asthma
Do I still need the HEPA?
Yes. Carbon is designed to get organic compounds, not particulate in general. I wouldn’t have even tested this, but Anna accidentally forgot to attach the HEPA and unknowingly ran a regular particulate room test with carbon only.
The results weren’t pretty — far below the 95% reduction with the added HEPA. Thus, I do not recommend using carbon only.
Gas Test Methods
During the gas test, all of the doors and windows were closed. Because the test was in an enclosed porch, there were several windows, which means the room was less well sealed than an interior room.
During the test, I burned three Marlboro cigarettes pointing straight up in an ash tray on the floor. The gas detector was on a small ledge at just above waist height, only a few feet laterally away from the cigarettes. The purifier was placed a couple feet from the cigarettes in the center of the porch with the air stream pointing upward.
Burning three cigarettes in a small room is extreme. With the exception of airport smoking lounges, few places will have this much smoke in one place. However, the test does demonstrate that carbon reduces VOCs.
Many people have asked me how long carbon filters last. Unfortunately I don’t have a good answer to that. The carbon filter I used was still effective after several of these powerful smoke tests, which should be far more VOCs than a normal person would face, but I cannot say more than that.
VOC Info: The PID lamp has an eV rating of 10.6. eV ratings determine what types of VOCs PID machines can detect. With a rating 10.6 eV, this machine was probably reliably detecting benzene, which has an ionization potential of 9.25 and perhaps toluene, which is at 8.82.
This PID lamp is not designed to detect formaldehyde, which has an ionization potential of 10.88. However, I later conducted tests with formaldehyde samples and found that the MX6 could detect it, and an engineer at Industrial Scientific told me he thought it was plausible that the machine would pick up on formaldehyde, although not 100% reliably.
Detector Info: The MX6 came with a pump to draw in air (as opposed to versions that passively sample air). During the test, it was outfitted with the filter designed to prevent particles from entering the machine. I zeroed the machine if necessary before the test.
Early Tests: As I started doing the tests, I was still figuring out how to do the tests — it seems simple when you look at the results, but tests like this are not easy! For the first three control tests, I had not yet decided to let the MX6 run for 30 minutes after the cigarettes burned out, so data is limited for those three occasions. However, the conclusions are similar with those three tests excluded.
Control Condition: In the control condition, I used a Honeywell fan on the low setting to approximate the air flow of a Cannon on high with a filter on it. In other words, a filter lowers the air flow, so the Honeywell may be a better approximation of the amount of air flow during the experimental condition.
In addition, I ran tests with an IQ Air Health Pro Plus (1) with the filters removed and (2) with the pre-filter only as additional control tests. The results were similar to the Honeywell fan-only tests. I’ll use the IQ Air tests as the control condition in tests of the IQ Air’s ability to remove VOCs in the future.
Gas Test Data
Below is a summary of the gas tests. Because the raw data file is large, I’m making it available as a download here.
Data: Carbon-only test
Anna only ran the carbon-only test for one day because it was an accident. More tests would give a more reliable estimate of the effectiveness of using carbon only, but because carbon isn’t designed to get particulates in general, I’m hesitant to subject anyone to that much dirty air. On the upside, the outdoor PM 2.5 was quite stable that day, only moving up 4 micrograms from start to finish.
Does adding a carbon layer reduce particlate effectiveness?
Not significantly on the Cannon, significantly on the Original.
I recently published tests showing that carbon actually removes VOCs. But adding that carbon layer raises a question: adding carbon means the fan has to power through yet another layer of resistance. Does that make purifier less effective at removing particulates?
I measured how much particulate it removed with a Dylos particle counter and compared the particle counts (1) before I turned on the purifier at night and (2) the average of the last four hours before I woke up in the morning.
With the additional carbon, the Cannon particulate effectiveness dropped 1-2%. Thus carbon has a very minor negative effect on the Cannon.
However, the Original DIY had a harder time powering through the extra layer. Its 0.5 micron effectiveness dropped 19%, and 2.5 micron effectiveness dropped 15%.
Over the 10 testing days, outdoor air pollution went down significantly on 6 days — poor luck! To test whether that affected the overall estimate, I isolated the three testing days with changes in outdoor PM 2.5 concentration of less than 10 micrograms.
For those three days, the 0.5 micron results were still 95%. The 2.5 micron results were 93%, which is 2% lower than in the overall tests. Thus, the overall estimate seems to be little affected by outdoor fluctuations when averaged over all of the tests.
Data: Original DIY
Below is the data for 10 tests with the Original DIY.
One day (12/21) was an outlier because the outdoor AQI almost tripled during the test. Thus, I calculated the average with and without that outlier removed. With the outlier included, .5 microns effectiveness was 5% lower and 2.5 micron effectiveness was 2% lower.
Can carbon remove formaldehyde?
Why are people in China so concerned with formaldehyde?
Formaldehyde causes scary health problems like ulcers and cancer, yet it’s common in construction materials and new furniture. The people who should worry the most are people in new or recently renovated homes. Tests of new and renovated homes routinely find high levels of formaldehyde.
Why formaldehyde is so hard to remove
Formaldehyde is a tricky problem because it escapes from materials as a gas, and HEPA filters aren’t made to capture gases. I’ve heard companies claim that activated carbon can remove formaldehyde, but I’ve also seen companies selling sprays that claim to clear formaldehyde from your home. That sounds a lot like snake oil to me.
Since there’s a profit motivation for companies to claim they can get rid of formaldehyde, I wanted to empirically test whether carbon actually works.
In my quest for an answer, the first stumbling block was detecting formaldehyde. Many companies on Taobao will sell you a “formaldehyde detector”, but they’re actually general VOC detectors. (There are lots of volatile organic compounds; formaldehyde is one type.) So even if the machine tells you it’s detecting formaldehyde, you have no way of knowing whether it’s formaldehyde or some other VOC.
Thus, to be absolutely sure we were detecting formaldehyde, my collaborator Anna bought bottles of liquid formaldehyde — risking our health for science!
To spread it in the room, we put it in a rice cooker along with 250 ml of water in a 4.14m2 porch (volume 10.35m3). When the rice cooker heats up, it releases formaldehyde as a gas into the air.
The Industrial Scientific MX6 detects different types of VOCs, not formaldehyde specifically. But because we released formaldehyde in the room, we can be sure that VOC is formaldehyde.
Anna turned the cooker and the fan on at the same time and let them run until the VOC level fell back to zero. We also ran a control test with a fan only. We ran a total of three carbon tests and two fan-only tests.
Here’s what one of the carbon tests looked like, starting from the peak formaldehyde level:
But we need to be sure that’s the effect of the carbon, not just the formaldehyde dispersing over time. To do that, we need to compare those results to the fan-only condition. Here’s what the two tests look like side by side:
The formaldehyde levels stayed higher for longer in the fan-only condition, but the levels dropped much quicker when we used carbon.
I averaged across all three carbon tests and compared the average reduction compared to the fan-only condition. On average, the carbon reduced formaldehyde levels to 50% within 15 minutes of the peak formaldehyde levels and then down 0% by 25 minutes.
These composite activated carbon filters removed formaldehyde from the air. My earlier tests show that these filters remove other types of VOCs too.
Does everyone need carbon?
Studies have found that formaldehyde is much more common in new and recently remodeled homes, so people in new or remodeled homes probably need carbon. However, my MX6 found zero VOCs in ALL apartments I tested except for places that had recent renovation or smoking.
Now, there could be VOCs at levels lower than the MX6 can detect, so I’m not confident to say there are ZERO VOCs in most houses. But I think it’s reasonable to say that carbon filters are not mandatory for homes without obvious sources of formaldehyde or other VOCs.
Do I still need a HEPA?
Activated carbon is made to get smells and gases like VOCs. It is not designed to get particulate in general. In fact, activated carbon is made to be as porous as possible to get as much air into contact with the carbon. And my tests with a carbon filter alone show that it is does not remove high amounts of small particles. Thus, we still need a HEPA.
The raw data is a large file, so I’m making it available as a download. Here is the summary data:
Check out my earlier VOC tests for more details on the MX6 detector, placement of the detector, and the fan-only control condition. Conditions were identical except for the pollution source.
How do we know the detector was actually detecting formaldehyde?
I wanted to test whether the MX6 was detecting the formaldehyde and not the heat and humidity coming from the electric cooker. To test that, I also ran a condition where I filled the rice cooker with water, but no formaldehyde. In that condition, the MX6 read zero:
That tells us that the MX6 wasn’t mistaking heat or water for formaldehyde.
Similar to my earlier VOC tests, one limitation is that the formaldehyde here was not naturally occurring. It would be ideal to find a house that was recently remodeled and already has formaldehyde in it because that would be a closer approximation of how most people would use it.
However, one difficulty of that sort of test is that the VOC detector wouldn’t be able to tell us if we’re detecting formaldehyde or other VOCs. In one sense, that’s not important — we want to get rid of all of them. But it would also be interesting to know if we’re getting formaldehyde specifically. To do that, we would need to take gas samples and have them sent to a lab (although if anyone knows of an easier way to detect formaldehyde and only formaldehyde, let me know).
Does air conditioning bring in dirty outside air?
A question that I get asked often (and that I have always wondered about) is whether my wall-mounted air conditioner is bringing in dirty air from outside. If so, it’d be safer not to use it, especially on really bad days.
My short answer is no. To explain why, I’ve got three points of evidence:
1. How air conditioners work. Regular wall-mounted air conditioners in China do have a unit outside connected with tubes to the inside, but that tube is not bringing in outside air. It’s passing coolant, and letting heat escape outside.
So where does the air it’s blowing come from? If you look around your air conditioner, you’ll probably discover that it works like mine: it brings air from the top, runs it over the cooling coils, and blows it out the front. It’s recycling indoor air, not bringing in outdoor air.
2. Tests of the air coming out of the air conditioner. (See a live test here.) I’ve held my particle counter up into the air coming out of my AC unit, and it’s no different from the ambient room air. I’ve also compared that air to outside air on very dirty days, and the air coming out of the AC is nowhere near as dirty as outside air.
(I did this test when I had just turned on my AC. If the AC were bringing in dirty air and I were to test the exhaust after I had been running the AC for a long time, then my whole room would be dirty, not just the exhaust.)
3. Tests of the ambient room air before and after turning the AC on. Results? AC makes basically no difference. Here are the results of 7 tests I did in my bedroom.
In each test, I ran my particle counter for 30 minutes to get a baseline. Then I turned on the AC for 30 minutes. Here I’m comparing the numbers just before I turned the AC on and 30 minutes later. As you can see, there’s basically no effect. If anything, 2.5 micron particles go down slightly. This could be because of the coarse filter in the AC unit. Or it could be random variation.
Conclusion: If it’s hot outside, don’t sweat it. Use your AC.
Central AC: I should note that these tests are of wall-mounted AC units in China. Central air conditioning may work differently.
Here’s an Excel chart with my original data, including test times, outside AQI as measured by the US embassy’s Twitter feed, and (at times) outside measurements with my particle counter. Remember that the scales between these two types of counters are completely different.
If the AC were bringing in outside air, I would expect the largest effects on really bad days. But if you look at the tests on the very bad days (8-15 was the worst), the numbers are no different.
Are our particle counters the same as government air monitors?
My particle counter is a beast — I’ve loved it. But I’ve always wondered how the counts of laser particle counters like mine:
…compare to the measurements of the huge stationary air quality monitoring stations that governments use, like this one in New Zealand:
If you look at the US Embassy’s Twitter feed, you’ll notice that its raw numbers are “concentration,” which it explains are micograms per cubic meter (µg/m³). The way government (BAM) machines work is that they use a source of carbon 14 that emits beta particles and then measure how many of those beta particles make it through to a detector. They then use those numbers to estimate the weight of those particles (micrograms).
In contrast, laser particle counters like mine use a laser and a photo diode sensor to estimate the number of particles in the air. I don’t see why the weight should be any more important than the number — they’re both telling you how much particulate pollution is in the air.
As an analogy, if we want to understand the crowd at a basketball game, we could count the number of people, or we could weigh those people. Of course, the more people, the heavier the total weight. And of course the two numbers won’t correlate perfectly if we have more heavy people on some days and more children on other days. But the weight and the total number should correlate highly.
The other major difference is that laser particle counters give the number of particles at that particle size and above. Government machines give the number of particles at that size and below.
To see how the two numbers compare, I put my particle counter outside my second-story window 70 times (that’s nerd dedication!) and compared my numbers to the US Embassy’s Twitter feed at the same time. Here’s what they look like:
They correlate at r = .89, meaning the two numbers are very strongly related (remember the highest possible correlation is 1). That high correlation is especially impressive given that my house is near Gulou, and the US Embassy is out in Liangmaqiao — about 7 kilometers away.
The difference between the readings was particularly noticeable on days where a strong wind moved through Beijing. I noticed several times that my particle counts would drop before the embassy’s counts as the wind moved in from the west (where my house was) to the east (where the embassy is). (Remember, Beijing’s air gets a lot cleaner when we get winds from the west.)
If we want to get a rough conversion between the numbers, we can remove a few of the outliers and compute a regression line:
For example, a government concentration of 100 micrograms (four times the WHO limit!) is approximately 25,000 on the Dylos particle counter:
And the 24-hour WHO standard of 25 µg/m³ is about 3,000-4,000 on the Dylos.
During my home tests (before turning my filter on), the air inside my home was very often above 3,000 (even though it was still much cleaner than outside).
Conclusion: My particle counter is giving measurements that are highly related to the much larger air monitoring stations. The scale is different, but the two can be roughly converted.