Category Archives: Indoor Air Pollution

How To Choose An Air Purifier in China

Isn’t it a total hassle to buy an air purifier in China, especially for expats? Back in places like the USA, when we shop for anything we can luxuriously research using consumer magazines and other review sources, wisely shopping around online and in stores as educated consumers. Here, many of us struggle with Chinese so are at a loss on local shopping websites, forcing us to rely on word of mouth and a tiny handful of imported brand stores. This usually has meant that many of us end up buying the same legacy brands, usually quite reputable but also usually at incredibly marked up prices. Even if fluent in Chinese it’s still very difficult to research credible data. It’s a classic Economics 101 example of information asymmetry, where the typical consumer doesn’t have all the information that they need to properly research the marketplace, thus giving too much leverage to the seller — leading to a very unbalanced supply and demand curve where we pay much more than we should, for a machine that may not be exactly what we needed.

So in my continuing efforts to educate people about healthy living in China and elsewhere, I’ve written a basic how-to guide to buying an air purifier, similar to my pollution mask buyer’s guide. I won’t be discussing the need to buy one, as I’ve made it blindingly obvious in many other articles that an air purifier is essential for anyone living here, especially children.

Seal your windows and doors!

Pre-step: Hire An Independent Testing Company

I suggest that anyone with a villa, large home or a business should think about saving some time and possibly a lot of money first by hiring an environmental consultant to assess your site. You may think famous Brand X air purifier is great — but do you need one, two, three or more? And can you get equal benefits from a much cheaper model? And which speed setting do you need to use for general use? Let them figure out all the complicated cubic meters and assess how much machine(s) you need. They also do a great job looking for — and fixing — air leaks around windows and doors (which is cheap and effective advice for everyone, even in your one-window tiny dorm room). They also are experts on the brands of machines available. They should also come back for free for a specified time and double-check that all is OK. I’ve personally used two such companies, both run by expats: Environment Assured, which aim to get your indoor PM2.5 concentration under 10 ug/m3 (the WHO standard, and my goal at home), and also PureLiving China, whose indoor goals for PM2.5 follow the higher Chinese target of 35 ug/m3. I’m sure there are other companies out there, as this surely should be a booming industry. I personally think the evidence is overwhelming that your goal should be to keep your indoor PM2.5 under 10 ug/m3 all the time — even when the outside air is crazy bad. It can be done, and it doesn’t have to be as expensive as you fear.

Step 1: How much/many do you need? Let’s do some math

Kids' college funds, or air purifiers? Hm...
Kids’ college funds, or air purifiers? Hm…

If your needs are simple or small, or you’re only here a few months, you can definitely handle this on your own. First, in order to know how much or how many machines you need, you first need to do some math: you need to calculate your room’s volume. Most air purifier ads mention either recommended room size (适用面积) as square meters or airflow rate (风机空载风量), often written as CADR, and listed as cubic meters per hour. Here are the steps:

  • First, measure your room and get the area of your floor space, in square meters. For example, my dining room, living room and long hallway are collectively 84 square meters (“84 m2”).
  • You could be done already! Now you could just check the ads for the recommended room size (适用面积) and do the math. For example, I need 84 m2 of protection — a very large area. I don’t see any single machine rated that high but I see a few in the 40-70 m2 range. So in theory, two machines rated for 61 m2, with a combined area of 122 m2, should easily take care of my 84 m2.
  • That’s the easiest way — but I actually find it more accurate to use airflow and volume because it factors in the room height. A 20 m2 room with standard 2.6 meter high ceiling will need a lot less machine than the same floor space with cavernous 6 meter high ceilings, as many villa front rooms may need. So let’s do more math! Multiply your room area by the room height. For my front rooms: 84 m2 x 2.5m height = 210 cubic meters (210 m3) volume.
  • Since a proper air purifier needs to replace the entire room air at least five times an hour (5 Air Changes Per Hour, or ACH) to really be effective, you need air purifiers that collectively can cover the volume 5 times. So the total airflow (风机空载风量), or Clean Air Delivery Rate (CADR) you need = room volume x 5. For my front rooms, I need 210 x 5 = 1050 cubic meters per hour (m3/h) CADR. Since no single machine has this much power, I’ll need to buy a few machines.
  • Then you need to research air purifiers’ CADR — but here’s where it gets tricky: the published CADR applies only to max speed which is almost always not the speed you will be using 99% of the time. Most machines’ max speeds are far too noisy for normal use, and the speed you’ll actually be using may have airflow much less than the published CADR. The problem is that it’s hard to find published CADR for the lower speeds. I think as a general rule, you should cut the CADR in half to get a more accurate sense of how much coverage you’ll get.

Let’s continue with my example as a case study, now that I know I need to get to 1050 m3/h for my front rooms. Comparing a couple of very large machines:

  • Model A on max setting (5) = 783 m3/h. So I would need 1050/783 = 1.34 machines needed.
  • Model A on default setting (3) = 285 m3/h; 1050/285 = 3.7 machines needed.
  • Model B on max (6) = 510 m3/h; 1050/380 = 2.06 machines needed
  • Model B on medium (4)  = 289 m3/h; 1050/200 = 3.6 machines needed

Thus I could conclude that 2 Model A machines may be the most efficient choice for my needs, compared to 2 Model B machines, especially when it’s crazy bad outside and I want to crank up the machines to max speed for a few minutes. On the usual settings I would need an equal amount of machines, but they’re both equally expensive so I’d just start with two of Model A, take some data for a while with my Dylos, and see whether I need to buy a third one.

  • Thus armed with such information for your own rooms, now you can properly research which machines may be best for your needs. Generally you’d want a higher airflow than you really need, so you can use the machine at a quieter level and still get effective airflow. Or maybe two smaller machines combined would still be a better value than one more expensive machine.

Step 2: How Much Budget?

Only 200 RMB
Only 200 RMB (or 6 Starbucks lattes)

Not everyone can afford 12,000 RMB ($2,000 USD) for an air purifier, and the wonderful news is that nobody should be paying such exorbitant prices. Certainly when I arrived in Beijing eight years ago those very few early entry, imported brands served an invaluable service for expats and others. But now, there are many reputable brands making perfectly fine HEPA filters in China at a far more reasonable price point. For example, I’m pretty sure almost every student in China could sacrifice a few days of their Starbucks latte and buy the 200 RMB do-it-yourself air purifier. I also just published my own data proving that a slew of air purifiers under 1,000 RMB are perfectly fine for smaller rooms. I’ve tested most of the expat-famous imported models and yes, many are quite good, but I wouldn’t call most of them a good value for the money — not anymore. For example, here is a list from of HEPA machines under 3,000 RMB which are rated for room sizes over 50 m2, including models from famous international brands such as Westinghouse, Philips, and Panasonic.

Step 3: Research data

In terms of features (工作原理), I strongly feel that you do not need ions (负离子) or ozone (臭氧) which actually can cause more lung harm than help — especially in the cheaper machines. Even UV lights (UV灯) are a bit gimmicky. I also don’t care much about killing bacteria (杀菌) and in my home formaldehyde (除甲醛) isn’t a big problem. Really, all you need is an awesome HEPA filter which clearly mentions it eliminates >99% of particles (过滤灰尘/花粉 (0.3 微米)), plus a strong fan speed which reflects in a high airflow/CADR. My second rank would be an activated charcoal filter (活性炭 滤网) which absorbs the sometimes serious indoor gases called VOCs, especially formaldehyde (甲醛) and benzene (甲苯).

In terms of independent reviews, English readers should definitely start with, which collects the most reputable reviews from multiple review sites such as Consumer Reports as well as reputable independent testers and also consumer reviews from Amazon. But this is limited data for us in China as many of the brands aren’t the same here. I dearly hope there’s a Chinese version of such a useful site for Chinese consumers! In the meantime, I’m aware of a couple of independent reviews in China which I’ve blogged about here; I also found another Chinese-only review here.

For other China-specific reviewers, I’ve blogged quite extensively about air purifiers and have reviewed many, which you can read here. My other favorite tests are from Thomas Talhelm, the creator of the 200 RMB Smart Air filter who also has extensive tests on his blog here.

The next level of research are the online shopping stores. Each model is overwhelming you with pretty pictures and a hard sales pitch, but it’s here that you’ll find the details of CADR, filters, room sizes, and extras such as carbon, formaldehyde, etc.

Step 4: Buying

In China, online shopping has quickly become the most efficient way to buy almost anything, including air purifiers (空气净化器). I recommend starting with Amazon China’s Clean Air Store, partly because on the left hand side you can instantly filter by square meters, price and features like HEPA. Also, for most English-speaking expats the Amazon store is much easier to use and very familiar to the USA version; they even have an English interface. My second choice online store would be Jingdong (, especially if you click on their own distribution brands (京东配送); they also have excellent filter options. My last choices would be Taobao and Tmall. Of course there are retail stores as well but for research and convenience, not to mention price, online is great.

Step 5: Testing

Dylos 1700 particle monitor
Dylos 1700 particle monitor

What good is spending all this money on machines if your indoor PM2.5 still isn’t under the goal 10 ug/m3 all the time? I’ve saved a huge amount of money and also know that my indoor air is clean because I invested a bit in a portable particle monitor. There are many brands but my favorite is the Dylos 1700, which also seems to be developing a fan club in China, led by the helpful FAQ on the fantastic website Apparently you can buy online on Chinese sites but it’s cheaper (and probably safer) to get in the USA directly from them or Amazon. I hear that a lot of groups are pitching in to buy one and share, which is a terrific idea. The goal with this Dylos is to get the data on the left side of the screen (PM0.5 but actually more represents PM2.5) always under 3,000, which correlates to an AQI of 50 (which itself means PM2.5 concentration under 12 ug/m3).

Step 6: Maintaining

Before and after the airpocalypse
Before and after the airpocalypse

It’s quite shocking just how many people buy machines and actually forget to change the filters, essentially making them totally ineffective if the filters get too clogged. It’s not only crucial to replace the filters on time (all machine replacement schedules are different), it’s also helpful to frequently wipe or vacuum the outside of the machine to get dust off the prefilters. I’ve had a Blueair for a long time — and totally forgot about the plastic prefilter screen underneath the machine, which was totally clogged with dirt and I’m sure was dramatically lowering the airflow. That was embarrassing!

 My Bottom Line

There simply is no longer any good excuse for anybody in China, even those with limited resources, not to protect themselves with an indoor air purifier. I hope I’ve provided some helpful information for you to make informed decisions.

My Quest For The Ideal Air Purifier Is Over.

Update January 2016: My data below speaks for itself, but people should be aware of possible quality problems reported here.

I think I’m done testing air purifiers. Yes, I’ve said this before, and I admit I often enjoy crunching the numbers, but I think I mean it now. The only possible way I could be persuaded to test another model would be an amazing breakthrough in technology. But it would still have to be reasonably affordable, even under 1,000 RMB per machine. Crazy, no? Actually, it’s not, because I sold all my IQAirs and my Blueair and switched to six air purifiers which cost me a total of 5,200 RMB. That’s right; a total of 5,200 RMB, for six new machines. I hardly doubt I need to remind most of you that’s half the price of only one of the many popular machines on the market now. And as this article will show, I’ve proven that these new machines are keeping my family’s indoor PM2.5 under 10 ug/m3, which is the goal of the World Health Organization and thus should be everyone’s goal (read more about that here). So why would I even want to test anything else?

As to which machine, many readers probably already know my answer because I started to test them last year: the Xiaomi 小米空气净化器 at 899 RMB, covering 48 square meters (apparently there’s a newer, smaller model as well for only 699). I initially blogged about Xiaomi last year discussing my test results of indoor air purifiers under 1,000 RMB. After last year’s testing, I had replaced two imported models in bedrooms with the Xiaomi purifier, and my bedroom air has been just fine since the switch. Just a few weeks ago I made the really big leap, returning two large air purifiers in my front rooms with three Xiaomi purifiers (I needed one extra as the coverage was comparatively less). And for the record, I am getting no promotional money or free anything from Xiaomi.

In our nine years in China we’ve always used well known imported air purifiers to cover our front rooms, which are essentially one large open space covering 83 square meters. But the price gouging in China always galled me. So after last year’s testing of purifiers under 1,000 RMB, I was always thinking, why can’t I see if I can also replace these big machines in the front rooms with something of more value yet equally effective? The timing was perfect as we’ve had multiple December days way over 200, 300 — and the worst two days of pollution I’ve ever seen in Beijing in my nine years here, our 2015 Airpocalypse with readings over 600 and an eerily dark, yellow sky in the afternoon of December 1st.

Let’s jump right in to the only thing that matters: hard data. And maybe the most important test was Airpocalypse 2015. So here’s the bottom line: data from our Laser Egg PM2.5 monitor showed an outdoor average PM2.5 concentration of 465 ug/m3 over those couple days, and inside the front rooms only 25 ug/m3. That’s an astonishing 95% steady improvement over the worst pollution you’ll pretty much ever see here, and although it’s not at my ideal of 10 ug/m3, it’s still far below the 35 that most Chinese environmental teams shoot for anyway, and the 95% reduction is certainly way below the usually recommended 80% goal. And the machine’s weren’t even on the top, super-noisy speed but the middle speed, which is still fine for general conversation and TV watching. Here’s the graph from the Laser Egg app, with the red line showing outdoor PM2.5 and the blue line showing my living room PM2.5 (the left side axis is PM2.5 concentration in ug/m3 and the bottom axis is just the number of data entry points):

Red line: outdoor PM2.5 concentration (ug/m3). Blue line: indoor PM2.5 (Laser Egg PM2.5 monitor)
Red line: outdoor PM2.5 concentration, US Embassy (ug/m3). Blue line: indoor PM2.5 (Laser Egg PM2.5 monitor)

On more “normal” days the data was much closer to, and usually under, my aggressive goal of PM2.5 concentration under 10 ug/m3. Here’s the data from earlier November:

Red line: outdoor PM2.5 concentration, Beijing average (ug/m3). Blue line: indoor PM2.5 (Laser Egg PM2.5 monitor)

Even a quick glance at the graph above tells me everything I need to know: my living room air (the blue line) is generally exactly around 10, and always under 20 even as the outdoor air climbs up. And as Beijing’s annual average is 70-90 ug/m3, my three Xiaomi machines easily get me to my goal of 10 during those typical days — and at a nice auto setting with very comfortable sound level. For calculated averages over this week using the Excel spreadsheet data from the Laser Egg, my front room average was 6.1 ug/m3 when outdoor air was a relatively clean 61; 9.8 during heavier days averaging 136 outside; and a still respectful 14.4 when outdoor air tanked for a couple days at 220 (that’s an impressive 93% reduction).

You notice there are a few spikes on the blue line, but all of those are easily explainable as with any other air purifier I’ve had: either the windows were open while we were using our laundry dryer, or we were cooking, or the machines were accidentally off for a while and no one had noticed.

Here’s one more graph from my bedroom, again demonstrating even more clearly than last year’s test that a Xiaomi on quiet setting is perfectly effective, and easily just as effective as any machine I’ve ever tested, including all the far more expensive models. The calculated average PM2.5 concentration was 8.0 ug/m3, again nicely below my goal of 10. It was 8.6 ug/m3 when the outdoor air was a horrible 224 over two days; and 7.4 ug/m3 on two clearer days, when the outdoor air average was 69:

Bedroom data
Bedroom data, 4 nights (red line: outdoor PM2.5 concentration. Blue line: bedroom data)

So let me be perfectly blunt to those in China who say that an indoor PM2.5 of 10 ug/m3 is impossible and too expensive: I say that’s nonsense, especially now, certainly on the money side. A goal of 35 ug/m3 still gives everyone a 15% increased mortality risk, as the WHO specifically mentions in their Air Quality Guidelines. I do realize this is a sensitive concept for some readers, as many people have spent a lot of money on air purifiers, or even work for those companies. And it’s fair to point out that these Xiaomi replacement filters don’t last as long as others (three months, although they’re only 179 RMB), and their HEPA filter at 99.3% (rated H11) is not as efficient as most other higher-end machines’ filters, and maybe the machine’s solidity isn’t so great, and perhaps they may break down sooner than other machines. We simply don’t know yet, as these are new machines only a year on the market — which in itself should give people some pause. Also, the built-in sensor is quite inaccurate and pretty much useless to follow — but I’ve never found sensors ever reliable on any air purifier, and I always control all of them manually anyway (and none of the major brand machines ever had a sensor anyway). The sensor’s more concerning issue is that it thinks that a PM2.5 concentration of 75 ug/m3 and under is safe. That’s simply too high, and even the Chinese government rating of 35 would be more proper. I of course think 10 is best, but for a machine made in China I would think 35 is reasonable. So if people are using a Xiaomi and relying only on the sensor on auto speed, then indeed they still may be inhaling not great air. This again is why everyone, using any type of air purifier, should not be relying only on any machine’s auto setting ever (my personal opinion). You need to get a separate monitor and use that to test your air.) And yes, Xiaomi is mostly a cell phone company first.

But let’s be very clear here: I’ve mentioned many, many times that anyone living anywhere in polluted areas, whether China or India or Los Angeles, has an ultimate goal to get your indoor air pollution under the World Health Organization’s goal of PM2.5 under 10 ug/m3. The scientific data is very clear; any PM2.5 over 10 starts to have health effects. So people in polluted areas absolutely must have air purifiers indoors, where you spend 90% of your lives. At the very least, put one in your bedroom. And for heaven’s sake, make sure your children’s rooms are all protected.

So you’re more than welcome to get any air purifier machine you want, as long as you’re reaching your goal of PM2.5 under 10 ug/m3, and routinely monitoring your air afterwards (perhaps with the new Laser Egg, only 499 RMB) to make sure all is working fine. In terms of which air purifier is best, again I’ve mentioned often that all that anyone needs for indoor protection is a good filter attached to a good fan. I’ve tested many models over many years, in real world circumstances in my homes, and no particular model at any price point ever was so obviously better than the rest. This is why I am perfectly comfortable with my current setup.

And this is why I am done with testing. Probably…

Checking Your Indoor PM2.5, Cheaply, With The Laser Egg

UPDATE January 4, 2016: Here’s a great review article about PM2.5 monitors: click here; they praise the Laser Egg but also discuss its limitations.
laser-eggI’ve mentioned often that I feel it’s important for people who use air purifiers to make sure your investments are working well. For me, this means always keeping your indoor air PM2.5 concentration under 10 ug/m3 (read my explanation here). The only way to know this is to have a PM2.5 monitor to check your air. We can spend so much money on air purifier machines — but how are we sure they’re on the correct speeds, or in the best part of the room, or that the filters aren’t clogged and need to be replaced early? And how do you know if you need to increase their speed when the AQI is airpocalyptic?

You certainly can’t rely on the built-in PM2.5 sensors that many air purifiers have, as the vast majority I’ve seen are inaccurate and basically useless. No, the answer has always been simple — and complicated. You’ve always needed to buy a separate PM2.5 monitor, but the best ones are wildly expensive. For a few years I’ve been using a popular, more economical choice from America called Dylos, but their data was uncommonly difficult to download and convert to something we layman could understand.

Now, finally, we’re starting to see some reasonable options, and I’m thinking specifically of the new Laser Egg, an air quality monitor from Beijing-based environmental company Origins, set up by an expat couple. Many of you are already using this Laser Egg, which is available for a wonderfully reasonable price of 499 RMB. I’ve been testing a few of these Eggs for a few weeks, and I can finally say that consumers now have a real option for testing your air at home (or work, or schools, etc etc).

laser-egg-appI have some highlights about this Laser Egg:

  1. The Egg already has saved me tens of thousands of RMB in air purifier costs, as I’ve recently replaced my home’s air purifiers with a far more economical machine (the Xiaomi), and the Laser Egg data proves that these new machines are very effective. (I will blog about this soon).
  2. Their app (Breathing Space) is very useful, as you can sync all of your Eggs to your home wifi, and you can access their data anytime, anywhere in the world.
  3. The app also shows the local outdoor air, so you can immediately compare indoor versus outdoor air. Very useful.
  4. You can choose to monitor the AQI from China or from the USA, or even better, the raw concentration of ug/m3. I always use concentration because I can quickly glance and see if it’s under 10 or not. Plus it’s the most evidenced based, as 10 ug/m3 is the official recommendation by the World Health Organization, and then you don’t have to deal with the politics of AQIs, which are totally different in every country.
  5. You can also use the app to export the data to your email, and create snazzy Excel graphs like the one below. For example, the Egg’s weekly data spreadsheet has 5-minute interval data on PM2.5, PM10, humidity and temperature. Data geeks like myself will love it.
  6. Its battery charges by USB so you can unplug and stuff in your bag and walk around town with it, checking out your favorite stores and also the outdoor air anywhere.
  7. You can keep it on 24/7, in sleep mode, and always access the data via your app or the screen.
  8. The app also can send you instant messages warning you if your air is suddenly getting worse. For example, I was at work and got a cell phone pop-up warning that my front room air was worsening, which I confirmed on the app’s graphics. I called my wife at home and she realized that all our Xiaomi purifiers had been shut off after a power outage. I turned them all back on (using another app from my air purifier), and sitting in my office I watched the app to see my home’s PM2.5 improving minute by minute. How cool is that?
  9. You can also rig it to check pollution masks, sort of. See the video I made below, using the Laser Egg to test an incredible new pollution mask I’m helping to design, called Freeair. It’s a battery-powered outdoor pollution mask that delivers pure air via a tiny fan and an astonishing new filter, classified ULPA (Ultra-Low Particulate Air), far superior to N95, N99 or HEPA filters. It filters basically 100% of PM2.5 and also filters much smaller ultrafine particles by 99.99999%. As you see in the video, literally every molecule of pollution is filtered out. You’ll definitely be hearing more about Freeair soon.
  10. You can quickly check for leaks around doors and windows. You all have already sealed your windows and doors with the very inexpensive rubber strips, yes? If not, read this and then come back. Anyway, one horrible night with AQI over 400 I noticed from the Laser Egg app that my baby’s bedroom PM2.5 was strangely high. And sure enough, the rubber sealing was coming off a part of a window. I resealed it, and voila! Happy baby.
  11. The Laser Egg looks pretty cool, actually.

So how reliable is the Laser Egg, and is the price too good to be true? For accuracy, you can check out the company’s own comparison testing with the US Embassy’s monitor, but perhaps you’d rather see more independent tests such as this reviewer’s comparative testing with much more expensive sensors. It’s in Chinese but just from his photos you see they’re generally the same. I’ve discussed the Egg with some pollution experts in Beijing and they’re generally supportive but mention that humidity can mess with the numbers a bit, and also putting Eggs side to side will show slightly different numbers. But in general, on a consumer level, it certainly does what it says it does.

I personally think the Egg could be improved, but it’s accurate enough for me and very impressive for a first version. And right now I definitely have stopped using the Dylos and will only be using the Egg from now on. If you’re interested in the Egg, you can buy on their website here or also via their Wechat directly (ID: originstech). I’ve also seen it behind the cash registers at April Gourmet and Jenny Lou’s. Think Christmas stocking stuffer…

The Best Air Purifiers Under 1,000 RMB: My Test Results

(Click here to read my new guide to choosing air purifiers in China.)
Shanghai Skyline in thick Fog

Just how much, or how little, are you willing to pay for an air purifier? Here in China expats have been so accustomed to scandalously expensive air purifiers that we get suspicious if anything’s under 4,000 RMB — which still would be more costly than almost any high end model in America or Europe. But I just finished testing a half dozen models each under 1,000 RMB, and I’m now quite convinced that the lower-priced market can offer excellent value.

I’ve tested quite a few air purifiers during my eight years here in Beijing, and I honestly thought (hoped?) I was finished with the testing. But I recently was diagnosed with asthma, which really angered me as much as it scared me. In my anxious efforts to control my symptoms, I decided to ramp up my protection from air pollution everywhere:  I threw out my N95 masks and now only use N99; and I’ve totally reassessed protection at home and at work. My goal remains getting 100% of my daily exposure to PM2.5 under the healthiest target of 10 ug/m3 (AQI <50), the WHO’s official target which I blogged about last year and which I strongly feel should be everyone’s goal. I spent much of January testing more than a half dozen small air purifiers, all under 1,000 RMB, to assess which machines are best in small rooms, including my office as well as my bedrooms.

First, some background

Particles and sizes

I will repeat what I’ve said before: all you need is a good filter and a strong fan. Literally! Filter+fan. A decent filter is by far the priority, hopefully officially certified as a High Efficiency Particulate Arrestance (HEPA) filter, which in the US means 99.97% efficiency against PM0.3; the EU’s terminology is much more confusing, allowing HEPA ratings from 85% to almost 100% (more on this later). PM0.3 is short for “particulate matter of 0.3 microns size” which is essentially microscopic dust 10 times smaller than PM2.5; it’s importantly considered the most dangerous size that penetrates easiest into our lungs, causing the most harm.

While a good filter is crucial, it’s actually not super critical only to have the most awesome Hyper-Nano-Xtra-HEPA filter, with multiple 9’s that impress and overwhelm the consumer. Yes, that level of filtration is totally cool, but those extra 9’s very possibly are clinically irrelevant and only add cost, not value. Also, if the fan sucks weakly and isn’t circulating your room air at least 5 times every hour (known as ACH, “Air Changes per Hour“) then the total machine isn’t good. A machine with a 99.9999% HEPA filter but a tiny fan absolutely may clean the air less effectively than an 85% filter with a massive fan.

The same filter+fan concept applies for our ultra-expensive machines that many of us, including me, use at home for our big rooms. If they’re all humming away on a lower, noise-tolerable speed, then the air flow is much less efficient than their advertised Clean Air Delivery Rate (CADR), which refers to its max fan setting — most likely not the speed you’re always running it. Thus your rooms’ pollution may be much higher than you assume. CADR is a bit of a gimmicky test created by the industry, not a health agency, and a machine’s max setting is usually incredibly loud and which nobody uses for routine use, especially in bedrooms.  The CADR test in fact is only based on 20 minutes of data! It’s a totally impractical test for consumers in China as it tells us nothing about how we all use these machines here in China — always on, 24/7, at lower, quieter speeds which don’t impede conversations or sleeping. I completely don’t care if a machine wipes out all pollen and bacteria from the air; I want to know if a machine will protect me from dying from long-term exposure to small particles like smoke and dust — at a noise level that won’t drive me crazy. That’s the data we need — and of course don’t have, at least not from the manufacturers’ ads.

Smart Air vs. Famous Brands

Every good expensive machine is literally just an amazing HEPA filter with a strong fan. And that’s all you need — forget the rubbish about ionizers and ultraviolet light. A good machine absolutely does not have to be over 10,000 RMB, as Thomas Talhelm proved last year with his consumer-revolutionary 200 RMB do-it-yourself air purifier. His first machine was an off-the-shelf 90 RMB HEPA filter slapped with a velcro strap on top of a cheap plastic fan. And in his tests — which I confirmed on my own — it performed about as well, or better than, far more expensive machines (see his data to the right). And why wouldn’t it? His HEPA filter, if you hold a particle monitor over it, filters 100% of both PM2.5 and PM0.5; only IQAir has had similarly perfect results in my tests. You may not be ready for Thomas’ ungainly Smart Air models in your bedroom, but once you review his data you’ll certainly be much more open minded to the once heretical idea that an air purifier can be effective and affordable.

The machines

When I first arrived in Beijing over eight years ago, there were only a couple of choices for air purifiers. Now, hundreds of new models have flooded the market over the last couple of years, and I’m just as overwhelmed as you are when I search the Chinese shopping sites. There’s also very little independent testing or reviews (I wrote about one here, and here’s another). My search priorities were: price under 1,000 RMB; famous international brands > Chinese models; filter rated HEPA or darn close; fan strength flow rate (CADR) at least matching what my office needed (which is ROOM VOLUME x 5 ACH; thus 21 m3 x 5 = 105 m3/hour airflow). Another real plus was the ability to filter formaldehyde or VOCs, but that wasn’t my priority. Separate pre-filters also are helpful — but uncommon at this price point. I chose these:

midea philips panasonic air purifiers
Midea, Panasonic, Philips
  • Smart Air Original: 200 RMB; filters 99.9% of PM0.3; CADR unknown; covers 11-20 m2
  • Air-O-Swiss P320: 699 RMB; filters 99.97% of PM0.3; CADR 86 m3/h, covers 16 m2
  • Panasonic F-30C3PD-D: 599 RMB; filters 99% of PM0.3 and 89% formaldehyde; CADR 163 m3/h, covers 30 m2
  • Philips AC4025: 699 RMB; filters 91.2% of PM0.3 & 95% formaldehyde; CADR 127 m3/h
  • Midea KJ20FE-NH3: 750 RMB; filters 99.9% of PM0.3 & 87% formaldehyde; CADR 204 m3/h, covers 20 m2
  • Xiaomi: 899 RMB (or 1,000-1200 RMB on reseller sites); filters 99.3% of PM0.3 & 91% formaldehyde; CADR 406 m3/h, covers 48 m2

My Office Results

My clinic office was the main testing ground, a cozy 9.6 square meters floor space (21.3m3 volume). For such smaller spaces, a huge Blueair or similar flagship machine would be expensive overkill and take up valuable floor space; thus my urge to test smaller models. I tested these machines over my lunch breaks, at maximum speed, with the doors closed, usually after flooding the room with outdoor air and then closing the window. My key data points were at 10 minutes and one hour (50 minutes), measuring the percentage drop in PM2.5 and PM0.5 using my completely-worth-the-cost Dylos 1700 particle monitor. During the rest of the workday I would keep them on their more standard, quieter settings and do random checks. The results are all in the table below, but the take home points are:

  • All machines were generally effective at my key index of PM2.5 at 50 minutes of max speed: PM2.5 reductions ranged from 88-94%. Some got there a bit quicker, and at 10 minutes the Xiaomi’s clearly larger fan definitely shows off, already clearing 88% of PM2.5 with similar results for PM0.5. Again this shows the importance of fans as much as filters, and the Xiaomi’s CADR of 406 m3/h is double the size of the nearest competitor, the Midea.
  • The Xiaomi needs to be clarified a bit: its max speed button only lasts 15 minutes and then reverts to auto, so its comparatively lower 50 minute data is harder to compare to the others because only the first 15 minutes are on max. But considering how at 10 minutes on max it was the leader, one could assume it would do at least as well if it had continued on max.
  • The Air-O-Swiss clearly was the smallest model, with a CADR of 86 m3/h not close to the 105 that I needed, and it shows in the results. But it still wasn’t too bad. Most likely their larger 350 model would be more helpful — but it exceeded my admittedly arbitrary price rule under 1,000 RMB.
  • The Panasonic and Philips both did fairly well — it’s just really too bad the Panasonic’s garish orange color makes it totally unprofessional for me to consider using. The Philips’ green is more acceptable, especially for a kid’s room, but still it’s not ideal.
  • The Midea is a nicely handsome black and white model, something I wasn’t ashamed to show in the office, unlike the two above. Its larger size and fan speed also were borne out in the generally good results, especially with PM2.5.
  • The Smart Air Original — the do-it-yourself 200 RMB filter and fan — performed at least as well at its more expensive cohorts, although a bit less so with the PM0.5. It’s not professional-looking enough for me to use, but it’s still a great value for students and the like, and it’s far better than using nothing at all (read my earlier review).
  • Noise levels on max speed wouldn’t be tolerable all day for any of the machines, but on their quieter levels were generally all OK for an office.
Air Purifier Results — Office (21 m3)


My favorites: I decided to test Xiaomi, Midea and Philips at home for the more important 8 hour results. I really like Xiaomi’s clearly larger capability thanks to its huge fan, not to mention its much classier, sleek look as well as its small footprint. I also thought the Philips and Midea were acceptable and not too out of place for an office.

My Bedroom Results

Xiaomi, meet Blueair
Xiaomi, meet Blueair

We’ve always used overly massive machines in our bedrooms, a Blueair 503 and IQAir Pro 150, and I thought now was the perfect time to challenge them against cheaper, smaller models (similar to my Philips test last year). This test was different than my previous tests, which had always compared indoor to outdoor air. This time, I wanted to see their efficiency over 8 hours, especially after a standard challenge by filling the room with particles, closing the doors and continuously monitoring while we slept. My particle test used essential oils in a vaporizer, which can raise the entire house’s pollution levels shockingly high and quickly. A secondary test was the final hour’s raw data: did it keep my indoor air in the green zone most of the time?

All machines this time were used on the quieter settings, not maximum as in the office tests. The great majority of tests were run in the larger master bedroom, which at 40 m3 (16 m2) needs airflow of 200 m3/hour. The smaller bedroom is 23 m3 (10 m2) and needs airflow >115 m3/hour.

Just for fun, I also compared these against an Airgle AG800, a much larger machine whose massive airflow of 763 m3/h easily should clean the air in such small rooms even at lower fan speeds.

The results, please

  • At the crucial 8 hour mark, most machines performed quite well. For PM2.5, Blueair led the way at 99.9%, just edging out Xiaomi’s sleep mode at 99.8%, with most others close behind in 98-99% range. For the smaller PM0.5, Blueair again topped out at 99.9%, closely followed by Philips at 99.6 and Xiaomi at 99.5%.
  • For short-term clearing within 1 hour, the Airgle did very well as you would expect for its huge fan, designed for much larger rooms. Otherwise, the budget machines from Xiaomi on auto and the Midea on low speed also did very well. (But the Xiaomi on auto comes at a slight cost, as its higher speeds to fight the incense are a bit louder and temporarily a bit distracting. And the Midea on low is a bit too loud for sleeping; fortunately the quieter sleep mode also did fairly well.)
  • Another useful metric is the final hour’s raw data (particles per cubic foot/100): My goal for PM2.5  is to always have the Dylos data stay under 50, which is roughly equivalent to 10 ug/3 (or AQI<50). The great news is that every machine’s final ranges were far under 50, including both cheaper models. For the trickier to control smaller PM0.5, the goal is a bit less clear but Charlie Thompson from my environmental testing company Environment Assured agrees that Dylos PM0.5 under 1,000 is the best equivalent to AQI<50. Using this cutoff, both the Philips and Blueair did extremely well, followed again by Midea and then Xiaomi in third. The larger machines from Airgle and IQAir surprisingly had some data over 1,000 — the only machines to do so.
  • In overall performance, all three cheaper machines certainly held their own, and none of them were obviously inferior even to the larger machines. The Midea posted impressive data even on their very quiet sleep mode. The Xiaomi also performed very well long-term and also was the best with short-term air clearing when in auto mode. The Philips also performed quite well with some very low raw data (but the outdoor air also was quite good that night, which helps a bit). For the three larger, far more expensive machines which on paper should have blown away their cheaper competition, only Blueair’s performance seemed obviously strongest, with the lowest raw data of all the machines in both particle sizes.
  • CORRECTION March 18: new data from the incredible Dylos FAQ at suggests that the left side number, PM0.5, actually is more representative of actual PM2.5 and the right side number represents PM10. Other data suggests that my PM0.5 goal of 1,ooo equals an AQI of 18 which actually is a bit of overkill; at 3,000 it equates to AQI of 53, just above the “green zone” of 50 (which equals a PM2.5 concentration of 12 in the USA). So anything under ~2,500 seems all to be in the green zone — which is great news as all my final hour raw data for PM0.5 was under 2,00 and often much less)
Air Purifier Results — Bedroom (40 m3)

My choices

For my office, my decision was fairly easy: I decided to go with Xiaomi due to performance, price and professional look. The Midea isn’t a bad second choice.

For the bedrooms, my decisions have been a lot harder. Since I already had machines, I’d need a really good reason to replace them. For the new bedroom, I’ve decided to use the Xiaomi. For our other bedrooms, I’m still deciding but know I will replace the IQAir and possibly also the Blueair with either the Midea or the Xiaomi.

I must talk a little bit more about Xiaomi’s air purifier, which is only on the market for three months but clearly may be very disruptive to the industry due to its novel design and incredible price. One very cool thing about Xiaomi which separates it from all the other machines is their app; their machines have built-in wifi which you can link inside their smart phone app and control the speed as well as get real-time data on the pollution inside the room, compared to the pollution outside (based on China MEP hourly data). This app even sends your phone an SMS telling you “the outdoor air is now clean and better than indoor air, it’s time to open the windows”! It also gives you real-time estimate of remaining filter life. It’s extraordinarily cool, even if a bit gimmicky — but this is my first foray into the exploding “connected home” phenomenon of intelligent, wifi-enabled home appliances, and I can definitely see the value in such data.

But I do have some reservations about Xiaomi, mostly because it’s a newish company known mostly in the mobile phone market for value, not so much top quality. Their HEPA filter is advertised as H11, which they claim is 99.3% efficient, which is not the highest rated top-line HEPA of 99.97%; but clearly their massive fan more than compensates for this, which accounts for its essentially similar, sometimes superior, results to the more expensive machines. As I said before, it’s the combination of fan+filter that’s the key.

Other negatives about Xiaomi are the newness of the machine. Since their air purifier is only a few months old, there’s no track record on maintenance or reputation. It’s also extremely difficult to buy on their own website due to their very unusual direct-selling marketing techniques; you can’t just order it, you have to first click to “reserve” and then on Tuesday at noon you and thousands of other people return to the website to buy, first come first served. I tried twice and they were sold out within one minute, so I was forced to use third-party resellers on Jingdong and Tmall — marked up, of course. This limited availability really disturbs me — what if the refill filters become hard to find?

These uncertainties are the main reasons why I didn’t replace my entire house with Xiaomi; otherwise, their price point and value are exactly the kind of disruptive innovation that the air purifier market in China desperately needs.

The Bottom Line (Finally!)

To summarize, I hope I’ve demonstrated that there are indeed many air purifiers far, far less than 10,000 RMB which may be perfectly adequate for your smaller room needs. Maybe 1,000 RMB models are a bit too weak, and for 1,500-3,000 RMB you can get a stronger fan and filter, especially from international companies with long track records such as PanasonicPhilipsWestinghouse and others.

Please note that my tests should serve as a guide, not as a strict recommendation, for your own decision-making. I also hope many readers also test their machines and share results with others, as Thomas does. My main hope is that people who currently don’t use any machine, mostly due to cost, will now realize they have far more options than they previously thought.


Other versions: Translation into Portuguese

Indoor PM2.5 Under 10: A Noble Goal


window sealing PM2.5 pollution
Window sealing

I’ve recently become quite smugly satisfied that my home’s indoor air is always 80% better than the outdoor air, thanks to a quartet of air purifiers working 24/7. But I no longer think my 80% reduction is good enough, and I now have a much more ambitious goal — to keep my home’s PM2.5 concentration under 10-12 µg/m3, all the time — even when the pollution is crazy bad. This target of 10-12 µg/m3 (equivalent to an AQI of 42-50, using the US EPA AQI conversion) may very well be a tilting-at-windmills fantasy, but that is now my goal — backed up by science.

I mention this because my home’s environmental testing team has an indoor target of 10 µg/m3 which is the lowest I’ve heard. Before this, I was more familiar with an indoor air target of 35 µg/m3 (AQI of 100), which is what many testing agencies and air purifier vendors are advising. This 35 may be a fine goal for many, as long as you as an informed consumer realize that chronic exposure to 35 µg/m3 of PM2.5  still leads to long term health problems and is a compromise between economics and health, while under 10 truly is the number where health effects are approaching zero. Perhaps even more importantly, under 10 also is the official recommendation from the World Health Organization. Given all this uncertainty about ideal targets, I thought I’d try to walk my readers through the evidence, and you can come to your own conclusions as to which target you’d like to achieve.

Dirty pre-filter…

First, there actually is almost no such thing in the real world as a safe level of air pollution. Even with an extraordinarily low PM2.5 under 7 µg/m3 (AQI 30), the data shows an uptick in deaths, cancers and heart disease. As the WHO states in their 2005 WHO Air Quality Guidelines Global Update:

The risk for various outcomes has been shown to increase with exposure and there is little evidence to suggest a threshold below which no adverse health effects would be anticipated. In fact, the low end of the range of concentrations at which adverse health effects has been demonstrated is not greatly above the background concentration, which for particles smaller than 2.5 μm (PM2.5) has been estimated to be 3–5 μg/m3 in both the United States and western Europe.

The WHO updated this guideline in 2013, and with eight more years of research they are even stronger in their assertions:

Thresholds: For short-term exposure studies, there is substantial evidence on associations observed down to very low levels of PM2.5. The data clearly suggest the absence of a threshold below which no one would be affected. Likewise long-term studies give no evidence of a threshold. Some recent studies have reported effects on mortality at concentrations below an annual average of 10 µg/m3.

The WHO Guidelines for Indoor Air Quality (page 4) explain why their indoor air and outdoor air recommendations are the same:

The steering group assisting WHO in designing the indoor air quality guidelines concluded that there is no convincing evidence of a difference in the hazardous nature of particulate matter from indoor sources as compared with those from outdoors and that the indoor levels of PM10 and PM2.5, in the presence of indoor sources of PM, are usually higher than the outdoor PM levels. Therefore, the air quality guidelines for particulate matter recommended by the 2005 global update are also applicable to indoor spaces

Those italics are mine because this is very important for people to realize: your indoor air goal is the same as the outdoor air goal — and again, that means getting your PM2.5 under 10 µg/m3.

Much of the WHO’s research is based on a couple of famous, very large cohort studies involving hundreds of thousands of people, including the Harvard Six Cities Study and the American Cancer Society Cancer Prevention II Study. These studies show clear increases in death rates from all causes, as well as from heart disease and lung cancers, as air pollution rises. (It’s important to note that all of the data points in these studies, from dozens of cities, had a PM2.5 range from 10 to a maximum of 30 — far lower than most cities in developing countries across Asia now.) All make it very clear that after ~7 ug/m3, the health effects increase. Here’s the graph from the ACS Study:

Figure 2. Nonparametric Smoothed Exposure Response Relationship

Below is another graph from another famous article published in the New England Journal of Medicine in 2009, showing how life expectancy in US cities from 1997-2001 decreased with PM2.5 levels above 5 ug/m3:

Life Expectancies in 51 Metropolitan Areas in the US, 1997–2001
Life Expectancies in 51 Metropolitan Areas in the US, 1997–2001

Because of this and other data, the WHO’s Global Burden of Disease research uses a PM2.5 annual concentration of 7.5 µg/m3 as their counterfactual — the “control” number which would assume to have no health effects. All of their relative risk assessments, including their most recent reanalysis of household air pollution, use 7.5 µg/m3 as the ideal baseline — so why shouldn’t it be our personal goal as well?

Some may still argue that 35 µg/m3 is still the more reasonable goal, as even the WHO officially states that developing countries such as China could use looser guidelines, called Interim Targets. Interim Target-1 states an annual PM2.5 of 35 µg/m3 as the target for annual exposure. Also, this 35 is currently China’s target goal for urban areas (15 for rural areas). And getting under 35 is actually a significant achievement in places such as Beijing, with annual PM2.5 last year of 89.5 µg/m3. But as the WHO states in their Table 1 (below), a level of 35 is “associated with about a 15% higher long-term mortality risk relative to the AQG level” — which again is 10 µg/m3.

WHO air quality guidelines for annual PM2.5
WHO air quality guidelines for annual PM2.5

The data seems clear to me, and yet here we are in the trenches, still with many differences of opinion. I’m convinced of the science and also have no intentions of waiting years for stronger data and a more unified opinion. Besides, it’s just common sense, isn’t it? Lower is better. For the sake of my wife and new son, I want my home’s indoor PM2.5 under 10 — always. If I can get there, I can literally breathe easier.