First of all, I’d like to express my great appreciation to all the valuable information made available both in the forum as well as in the BreathSafeAir website. I’ve been learning a lot over the past few days, and it’s been proven to be very helpful. Thank you very much for that!
Now, I’m writing as I would like to ask for some advice on choosing an air quality monitor - as I’ve been struggling to come to a decision, based on my following particular context:
My 4-yo daughter has been facing respiratory issues over a couple of years already, in particular during the night, e.g. coughing a lot for many nights in a row. After consulting with several pediatricians, the suggested root cause seems to be some kind of allergic reaction - but here I’m intrigued to understand to what exactly, and avoid always falling back to their proposed solution of anti-allergy medication. This is unfortunately a quite common problem in our region, and other parents also complain about that pattern.
Over the past year, I’ve been trying to use AC to control temperature, air purifier (Daikin) to filter particles, and dehumidifier to control humidity: we live in South of Brazil, and except for the few coldest days, we tend to have indoor RH at >= 70% (as I’ve been monitoring in our bedrooms since November last year).
Despite my efforts, her coughing episodes tend to re-occur, without any apparent link to any infection. Therefore, I decided it would be worth to invest in some [affordable] air quality monitor to try to get a bigger picture of her environment.
In Brazil we don’t have any options available, and after some reading I’d be inclined to import either a QP (Lite or Pro 2) or an Air Quality One monitor. The only caveat is that it will add 77% in taxes (over price + shipping), so e.g. a $200 monitor with $50 shipping fees already gets to almost $450 on my end, which is not cheap.
Nevertheless, I’m also concerned on what exactly to invest. I’ve been reading that measures (e.g. PM) may be affected by humidity levels, and understand that tests run by AQDM for example are under 40% RH - which wouldn’t represent my usual indoor environment to run the monitor. I wouldn’t like to invest in a monitor that may not be trustworthy in our indoor environment. And heard that overall low-cost indoor [PM] sensors may suffer from such RH influence.
So a few questions that I would have here:
Is it worth to invest in an indoor quality monitor with PM sensor or would the readings be too affected by higher RH levels (let’s consider ~70%)?
Would outdoor monitors work better for me due to the RH context (even if I want to use them indoors instead)?
As VOC values are only relative, can I still use them in case there’s a constant source of VOC inside my place? Or opening the windows for some time would already do the trick to re-calibrate them, assuming outside is absent of VOCs?
Is the CO2 measure also affected by RH? I understand that’s one thing air purifier cannot help with (and sometimes we indeed didn’t ventilate our bedrooms enough)
Based on the tax prices and my overall context, would you recommend some particular choice? We may consider following end-user costs for me: QP Lite (with AliExpress discount): $140 , QP Pro 2 (Amazon US): $300 , Air Gradient One (Kit): $340, Air Gradient One (Assembled): $440. I’m also open to other devices I have not yet considered herein.
Any advice would be greatly appreciate, and I’m happy to provide any further info.
Firstly, thank you for joining, and I’m glad to hear you’ve found the website so helpful! I hope I can help with your question. Since it’s quite long, I will do my best to answer everything one section at a time.
I’m sorry to hear this. I can understand the want to avoid medication where possible, and I agree that finding the root cause would be much better if it’s possible. Unfortunately, air quality monitors can’t identify particle type and composition, so it’s hard to determine what could be causing this, even with an air quality monitor. With that said, they could help identify the issue, and they will absolutely let you know if your air purifier is working effectively and if it’s catching airborne allergens.
Wow! Does this tax apply to everything imported, even if you purchase from a local retailer? If that’s the case, I assume it’s applied to every air quality monitor, and there aren’t any options that don’t incur this hefty tax.
This is definitely true, and humidity influences many kinds of sensors, not just particle sensors. That said, many companies and manufacturers implement correction algorithms to account (or at least partially account) for humidity and sometimes temperature. I’m in the same situation here in the Philippines, where my humidity is rarely below 70%. While it’s hard to know exactly which companies include these corrections by default (or as an option for you to enable), I know for sure that PurpleAir and AirGradient have corrections on their indoor monitors. While I’m not sure about Qingping, I am happy to reach out to them to ask.
As mentioned above, the readings will be affected, but they can be somewhat offset. Many of AQMD’s reports have a temperature and humidity combination graph, but these seem to stop at 65% RH, which is unfortunate.
Off the top of my head, the indoor monitors from PurpleAir and AirGradient have accounted for humidity (in PurpleAir’s case, if you apply a conversion on the map, such as the EPA conversion). If you’re interested in the Qingping monitor, I am happy to reach out and ask them (they usually reply quite quickly!).
I’m actually unsure about this, but they might well do so, as outdoor monitors are usually designed to handle more conditions and account for higher temperatures and humidity.
You would first need to know how often the baseline updates - for some monitors, it’s daily; for some, it’s every few days; and for some, it’s weekly. Unfortunately, this is something that manufacturers don’t often mention, so you would need to reach out to them to ask. If it’s weekly (or otherwise relatively infrequently), you could definitely identify constant sources and then re-calibrate every so often. If it’s daily, you could still do the same, but it would need to be manually calibrated quite frequently if you suspect there are constant VOC sources. Again, if there are any monitors you’re particularly interested in, I can reach out to them to ask about the details!
Yes, but these are normally accounted for with compensation algorithms so it’s not typically an issue.
Before discussing this, I do want to mention that I now work with AirGradient (I didn’t when I wrote my initial reviews), but I’m free to be as transparent as I want (and am encouraged to be!) as the monitors are open-source and we believe transparency is key.
Out of the options you mentioned, I think the QP Pro 2 or AirGradient One Kit are the best choices unless you absolutely want a warranty (in which case the assembled monitor is better, but it’s much more expensive).
I’m currently working on a review for the QP Pro 2, but so far, I think it’s a fantastic device that is incredibly user-friendly. While I think the first device is still tempting, I do appreciate the improvements of the version 2 device. Regarding our discussion here, one interesting factor about the QP Pro 2 (and I believe the other QP monitors, too, but I can’t check right now) is that you can change the eTOV reading from an index to ppb. While this isn’t ideal for reasons mentioned in this post, I believe that means it doesn’t automatically adjust - or at least not as frequently.
While I also don’t have my QP Lite on me at the moment, I believe this option is also present on that monitor. I really like the QP Lite, and it’s my favourite ‘affordable’ monitor. With that said, it doesn’t measure VOCs, which seems to be important in your case.
Since I work with AirGradient now, I know there are corrections in place to account for humidity. I also know that it’s possible to change the VOC baseline adjustment period to 30 days, which I think would be very useful for you. I will need to ask QP about if they use a correction algorithm for their PM sensor. The AirGradient devices aren’t as user-friendly as the QP monitors (I still find the dashboard somewhat confusing), but they are open-source and constantly releasing updates.
The only other monitor that I know for sure that measures VOCs (even if it’s still in beta) and that has RH compensation algorithms is the PurpleAir Touch (or Zen, but that’s pricier). Unfortunately, they lack a CO2 sensor and are pricier than the competition so I don’t think they’d be the best choices in this case.
The Atmotube Pro could also be worth looking into as it performs well in AQMD testing but lacks the CO2 sensor.
I hope this has helped and hasn’t made everything even more confusing. I will think more on this overnight and let you know if any other monitors come to mind! With that said, I think the monitors discussed here are the best in the <$200 price range.
Thank you very much for your prompt an detailed response here. It’s definitely very informative and helpful!
Regarding the additional options that lack CO2, I think it would be helpful for me to have that measurement as well, given that we work from home and might be nice to track CO2 levels throughout the day too. And I think you made a good point about the ability to measure VOCs - I agree it may be worth to have that included too. Then, I also started to consider radon monitoring too. But, except for AirThings devices, I haven’t found other options available yet. What are your thoughts here?
I have to say I’ve come to appreciate Air Gradient’s values and principles after reading them. I do like all the documentation provided in their website, how discussions held in different places are always backed by research as well as its transparency around key aspects, e.g. sensors’ accuracy and limitations.
Based on that, I’d be inclined to go for the AirGradient One Kit at this moment. I was just a bit afraid around the lack of warranty (in contrast to the assembled version) - for example due to eventual damages that it could incur either during shipping or even by eventual inspection from local authorities when it enters the country. Do you think these could pose potential damaging issues?
Also, another recent concern that I wanted to check with you: I found someone mentioning they had issues using the monitor in a bedroom due to infrared light emission (comment at bottom here: Apollo AIR-1 Air Quality Multisensor Review - SmartHomeScene). Have you seen this as well? Do you think it could disrupt sleep in a dark room?
And what are your thoughts on Apollo Air-1? I’ve been also reading about it - I think I wouldn’t go for the gas sensors, but otherwise seemed to be comparable in terms of features to Air Gradient One - even if their sensors choices differ. I couldn’t find any information about certifications they might have received though. Have you had the chance to test it and compare their readings against other devices?
For what it’s worth, I agree with your assessment of having a CO2 sensor. I am not a fan of indoor air quality monitors that lack this important sensor (unless you also have a dedicated CO2 monitor). Having a CO2 sensor is even more important if you work from home, as we both do!
I don’t think a radon sensor is essential unless you are in a radon-prone location (it’s worth checking this online if possible), and it tends to be worst in basements, but from your situation, it sounds like it will be in a bedroom or other room instead. The AirThings device is also much more expensive, and I am not a big fan of the device (it’s expensive and lacks lots of more advanced features).
Me, too, which is why I agreed to work with them. They’re extremely transparent, and they have people dedicated solely to the science side, which I think is very important. However, for me the most important aspect is that I’m free to speak my mind and even encouraged to publically highlight issues with the monitors if I find them.
I can understand your concern - they’re not cheap devices! Personally, I wouldn’t worry too much about the monitor itself, as even if there is an issue with any component, it’s all very easy to replace (and the sensors could likely be sourced locally). However, shipping and inspections are another matter altogether…
I’m unsure if a warranty would even cover these situations, however, as that would be on the shipper’s side, I imagine. If the warranty also covers shipping, then I think that is probably the safer route (and you can still replace parts if needed later down the line after the warranty expires).
I’ve been very interested in this monitor, but I haven’t yet had a chance to try it myself. However, looking at the specifications, these would be my thoughts:
Apollo has a slightly better PM sensor (SEN55) than AirGradient, but AirGradient’s is corrected, so they likely provide similar readings. I mention this issue in this post (which discusses the SPS30 but which performs very similarly to the SEN55). At the end of the day, I think the PM sensor in the Apollo is better and has a longer lifespan.
On the other hand, I prefer the CO2 sensor in the AirGradient monitor (the SenseAir S8) and still trust traditional NDIR more than photoacoustic NDIR (which the SCD40 in the Apollo Air-1 is). While both are accurate, the latter is influenced more by temperature and humidity. Regarding sleep, I don’t think it will matter as the Apollo device uses a photoacoustic CO2 sensor (fully enclosed), and the AirGradient device’s sensor sits near the top of the device and is hidden. I have an AirGradient monitor in my bedroom, and I’ve never noticed the light emissions.
I really like the Apollo Air-1’s small form factor, too. So, if that’s important to you, that could be the better choice.
Of course, I am slightly biased, but I personally would opt for the AirGradient ONE as I appreciate their science and research focus. I also like the non-Home Assistant dashboard, which I’m not sure if Apollo offers.
I was further reading about CO2 sensors, and more specifically how Aranet4 works. I’ve found this discussion on their forum, related to needed altitude compensation: Altitude compensation
It seems to be important to not only manually calibrate the sensor outdoor, but to also know the altitude. But if I understand correctly, the Air Gradient One does not have a barometric sensor included. Is the user able to enter that information somewhere to be taken into account?
I was wondering whether that could pose a potential accuracy issue on the CO2 readings, considering the city I live in is at 812m elevation. What are your thoughts?
I’ve seen in your Air Gradient review some comparisons with other sensors, including Aranet4, which seemed very reliable. But now I’m not sure at what atitude they were measured?
This is a great point, and it could be a significant issue. I’ve honestly not considered it much, as I’m almost always near sea level.
AirGradient does not have a similar compensation for altitude, and I’m not sure if SenseAir implements this by default. Here is the information I was able to find. There is no need to read all of the links, but I wanted to include them in case you or any other readers want to dig deeper.
Unfortunately, I can’t find anything on the S8 (used in the AirGradient monitors). SenseAir knows of this issue, as it’s implemented compensations on the Sunrise and Sunlight sensor, and the competition from Sensirion has it, so I wouldn’t be surprised if the S8 also has it. Still, I can’t find anything to confirm this.
All of my comparisons are carried out at low altitudes as I am normally not more than a kilometre from the ocean! I may have to start testing different pressures now, too.
You could add a barometric pressure sensor to an AirGradient monitor (I know there are a few forum posts out there from people who have), but it wouldn’t automatically compensate unless you were really comfortable modifying the device. Therefore, unless we can confirm that the S8 has such compensations, you might want to look at a device with a Sensirion or Sunrise CO2 sensor instead.
Thank you once again, Ethan, this is extremely helpful. I’ve just posted in the Air Gradient Forum to see if anyone has already implemented something similar. If so, I will share it back here for future readers, and also update the post once I finally decide which monitor(s) to purchase.
You’re welcome. I’m glad to hear that I’ve been able to help!
I hope you can get a good answer on the AirGradient forum, but I think I will also reach out to SenseAir and see if I can get a reply, as I believe they may hold the answer here.
For anyone reading this in the future, you can find the continuation of the conversation here:
@oseiler I’m not sure if you have any experience with the S8 or this particular topic, but I thought I would tag you (I hope you don’t mind!) because I know you have a lot of experience with different sensors and the conditions that impact them. Do you happen to know anything about this or how significant of an issue it would be in @Guilherme’s case?
As far as I know (from reading documentation and my own experience) while changes in air pressure impact the CO2 measurements, those changes are not huge compared to what we normally see in a normal room use. Especially when you’re not planning to move the sensor around I think the impact of living at higher altitude can be neglected, given that you do calibrate the sensor in fresh outside air (as all co2 sensors need regularly). I have combined SCD30/40 sensors with BME680 which measure pressure and feed it to the SCDxx, but stopped doing so, since it doesn’t provide any benefits.
I also believe that CO2 levels itself (unless they get very high) do not impact allergies, so I wouldn’t worry too much about high accuracy. All CO2 sensors I know have built-in humidity and temperature sensors to compensate for those impacting measurements.
From personal experience (though in a colder climate) I’d focus first and foremost on reducing humidity to 50-60%rH if at all possible, since high humidity is known to poorly effect respiratory issues. Add Hepa filters if you can to reduce any airborne particles as much as possible. Should VOC be elevated then add activated char coal filters.
Thank you Ethan! I believe it might indeed be a good idea to hear from SenseAir directly as well as from others with experience in the matter in the forum, looking forward to learning more from both sources
In the meantime, I’ve also tried to reach out to the other monitors vendors that we’ve discussed herein, namely Qingping and Apollo. I’m also adding their responses below as further reference.
Qingping:
The sensor we use has automatic calibration and manual calibration features, it can be used in high altitude areas.
(The calibration feature of “Qingping Air Monitor 2” is on its Settings page. For “Qingping Air Monitor Lite” and “Qingping CO2 & Temp & RH Monitor”, to calibrate their CO2 readings, the user needs to use the “Qingping+” or “Qingping IoT” app.)
So from Qingping I think they implied that their calibration should be enough - even though they don’t have (afaik) a barometric pressure sensor included in their devices. While Apollo confirmed to use the barometric pressure sensor for compensating the CO2 measurements.
Thank you very much for the further input here, @oseiler!
We’ve been trying to control humidity as well, but sometimes due to outdoor conditions (e.g. temperature and dew point at similar levels, and both around the comfort thresholds), we then don’t rely on our AC units to dry the air (to avoid becoming too cold), while standing dehumidifier helps but ours is not that powerful. So in our climate it is a bit of a challenge for most of the year. This week in particular is an exception, as it’s always belos 8°C outside, thus keeping AC in heating mode already does the trick for us.
As for the HEPA filter, we have an air purifier in her bedroom with HEPA and carbon filters. But still I thought it would be useful to have some way to measure all these components, also closer to her bed, to monitor air quality as well as air purifier efficiency.
And it’s also interesting to hear and learn this:
I have combined SCD30/40 sensors with BME680 which measure pressure and feed it to the SCDxx, but stopped doing so, since it doesn’t provide any benefits.
Just to confirm - you tested that also in higher altitudes and observed to not yield significant changes in the CO2 readings after controlling for barometric pressure?
I’ve no experience with that myself, my background was the Aranet4 home forum aforementioned, where the other user had expected quite some different levels due to his altitude. And then Ethan also found the S8 specs stating the 1.6% deviation per kPA, which I think translate to 16% deviation at 800m altitude (if the kPA calculator I found online is correct).
Thank you again for also contributing in the discussion!
Firstly, @oseiler, thanks once again for sharing your knowledge. I would also be curious to hear if you’ve tried at higher altitudes (or otherwise increased air pressure when testing) to see how the sensors perform.
@Guilherme, I will try to get in contact with SenseAir to ask this question. They’ve been quite responsive in the past, so I hope they will have a relatively quick answer!
Regarding Qingping, if the original thread you shared is anything to go by, I don’t think a simple offset can account for the differences. If an offset can account for this, then any CO2 monitor should be fine if it’s calibrated.
The Apollo answer seems much better, but it’s really hard to tell without knowing exactly what the compensation formula is. This actually leads me back to the documentation for the SCDx sensors, particularly this article: Sensirion SCD41 and SCD40 CO2 Sensors - eMariete. If it’s this easy yet makes such a big difference, I am surprised that more air quality monitor manufacturers don’t allow users to manually set their altitude (on sensors that lack an atmospheric pressure sensor).
Considering the extra cost, even with the 16% difference, I would personally still choose an all-in-one monitor such as the AirGradient, Qingping, or Apollo over two separate monitors (one of which likely is an Aranet4). Obviously, a monitor that can do both would be best, but there are very few all-in-one monitors that use a Sunrise or SCD30 (I believe my previous post about the SCD4x sensors compensating for pressure by default was incorrect).
I think there’s a misunderstanding here - the deviation is not absolute (e.g. the sensor is not generally 16% off at a higher altitude), but it defines the error you can get when the pressure changes from the pressure at the time you calibrated the sensor to a specific CO2 measurement (usually 420 outside air).
So as long as you calibrate your sensor at home (altitude), and leave it at that altitude, the deviation from pressure will only be that introduced by changes from weather (which is less than moving 800m up/down).
As opposed to calibrating the sensor at sea level, then moving it to 800m (which will get you into the 16% off error band).
When I set up my sensors to take pressure input readings then that improves their accuracy for changes in weather induced pressure changes, and taking them around the region (usually about 100m up/down, sometimes more). The difference of feeding live pressure data into the co2 sensor or not is minimal in that case, and it’s much more important to ensure that the sensor get calibrated regularly (manually, or relying on self-calibration => do check the manual for how to do that properly).
Hi @oseiler , thank you again for the further contribution here. As I mentioned, my background for this discussion was based on the Aranet4 Forum, in particular the portion quoted below:
For example, let’s say the sensor is factory calibrated at standard sea level pressure & temperature and then placed in service in Denver (5,430 ft elevation). If outside temperature is, say, 30F and pressure is nominal for Denver, the monitor should read 360 ppm prior to calibration (because there are fewer CO2 molecules at 5,430 ft). After calibration, the monitor reads 360 + 60 offset = 420 ppm. Let’s say actual indoor CO2 concentration is 1,800 ppm at 75F. The NDIR sensor will only detect 1,438 ppm. The calibration adjustment will add 60 ppm so the reading will be 1,498 ppm. That means the altitude error in this case is -302 ppm!
So my understanding was that the calibration outside, if solely resulting on applying an overall offset (e.g. +x) to future CO2 readings, would not be enough to properly account for the impact of the barometric pressure, which seems to rather vary proportionally to measured CO2.
There, they reference the ideal gas formula, as follows:
ppm CO2 corrected = ppm CO2 measured * ((Tmeasured * pref) / (pmeasured * Tref)) , where:
pmeasured = Current pressure, in the same units as reference pressure (not corrected to sea level)
Tref = reference temperature, usually 25°C, 77°F, converted to absolute (298.15 for °C, 536.67 for °F)
Tmeasured = Current absolute temperature, °C + 273.15, °F +459.67
pref = reference Barometric Pressure, usually sea level, 29.92 in Hg, 760 mm Hg, 1013.207 hPa or 14.6959 psi
But if we now rather assume that the calibration process is rather trying to identify “pmeasured” in the above formula, given that all other variables should be known, then it should indeed be enough? I’m not familiar how sensors calibrate and what equation they use though.
To complement the previous post, below a chart to illustrate this, as obtained from the Sunrise and Sunlight specs (pdf previously shared by Ethan in this thread):
As promised, I would like to share my decision here.
Initially, I was inclined to purchase the Apollo Air-1 as they do have an air pressure sensor that inputs to their CO2 sensor, which would then solve the issue above. However, I was also not so confident with to potential heating issue due to small enclosure, and how that could potentially impact the sensors measurements.
After that, I started considering investing in the Aranet4 and/or the Purple Air PA-II , as I seem them as the most promising for reliability on CO2 and PM2.5 readings, respectively. But it would be quite pricey in the end, and in particular PurpleAir has a few limitations on how we can use the data, which I was not so sure about.
But then I remembered one point that I like a lot about AirGradient: it is open source. So I can potentially adjust the firmware code towards my needs. Then I confirmed with Senseair that the math behind the adjustment is rather simple, and thus doable. And I could also adjust the current PM output to an alternative formula (e.g. ALT-CF3.4, an alternative proposed in the PurpleAir forum).
So AirGradient One it is! In addition to being open source and also sharing data with OpenAQ, I think everything is truly well thought and considered regarding sensors models, their placement, enclosure, etc. And I like the active community and the research support. Hopefully I can also contribute somehow with my own findings.
Now, I think I would be very glad to also have an outdoor monitor to better understand external influences, also because there is no outdoor monitor (reference grade nor low cost) near me - I think the closest one is 200km away. But I thought it would be quite expensive to add that to the order just yet (also due to expected taxes).
Thank you again for the fruitful discussions here. It has been very helpful for me, and I learned a lot about air quality components and sensors in the past weeks.
I’ve been following the conversation on the AirGradient forum, and your findings about the new SenseAir S88 intrigue me. I wonder if it’s something that can be swapped in the AG monitor since it appears to use the same form factor.
In my own review (https://breathesafeair.com/purpleair-flex-and-purpleair-zen-review/), I was quite surprised by how much the Plantowers in the PurpleAir overreport concentrations by default. Of course, AirGradient uses the same (or almost the same in the case of the PMS5003t vs PMS6003) sensors and suffers from the same issues. Luckily, both of them allow the use of the EPA formula (this will soon be the default shown by AG monitors), and you can also implement the ALT-CF3.4 formula either through the PA map or through the open-source aspect of the AG monitors. I don’t think there is such a thing as an ‘accurate’ low-cost PM sensor out of the box. Instead, it seems that the biggest improvements come from corrections and algorithms implemented on sensors that show systematic errors - as is the case with the Plantower sensors.
I would love to hear about your experiences with the AirGradient ONE! Also, I would love to discuss something with you. If you don’t mind, please reach out to me at ethan@breathesafeair.com.
