You may know this already, but both Olin Foundations are good examples, actually. I believe the John M. Olin foundation dissolution plans were specifically in response to the Ford foundation's drift. The F. W. Olin foundation (John's Father) coincidentally dissolved in the same year, but that was due to largely accomplishing their original goal of endowing engineering buildings at colleges, and pivoting to founding a new engineering college entirely.
I don't really know the details but an organization I was/am involved with did get money from the "Olin Foundation" but didn't know specifics beyond that. Yeah, one of my fellow board members observed that Olin was pretty much the canonical example of a foundation that set itself up to be dissolved.
This reference instrument variability is especially true of the BAM, which due to its measurement technology has a high "noise floor" at low concentrations, while remaining a fairly good measurement at high concentrations. I believe two identical BAMs next to each other can easily have more than 25% disagreement at the lower end of the concentration range.
The T640 also recently had its approved calibration algorithm(s) updated, to a new set which often reads about 7% lower. What does that mean for T640 reference data collected in the past? I don't know, and AFAIK neither does the EPA really.
All that to say that reference instruments are still instruments with limitations, even if from a regulatory perspective we treat them as "truth".
I link some sources in a separate comment up-thread, but the short answer is two stroke engines run orders of magnitude "dirtier" than a four stroke in a car, primarily due to incomplete combustion of fuel. (Some of this is fundamental to the physics of their operation, and some of this is they aren't subject to the same strict automotive emissions standards, so less engineering goes into reducing their emissions)
Intuitively, this is why you can sometimes taste a lawnmower running the next block over in the air, but the same isn't true for a modern car idling, even if they were consuming fuel at the same rate.
First off, kudos for putting in the effort to critically think through the implications of stats you see!
Actually, the counterintuitive truth is that two-stroke engines produce ~300-500x more PM/hydrocarbon emissions from the same amount of fuel use than a four-stroke engine, due to the fundamental nature of incomplete combustion and the oil-fuel mixture used in two stroke engines, and the efficacy of emissions controls on modern cars vs common leaf blowers. Your calculations are correct, it's just this difference is so big it seems crazy when you first learn about it. (Which is what Nicole's art is trying to communicate!) This is cited in the NYT article from the tweet you link: https://www.edmunds.com/about/press/leaf-blowers-emissions-d... , although for a more scientific source you can see https://publications.jrc.ec.europa.eu/repository/bitstream/J... .
While technically, carbureted two-stroke engines can sometimes produce less NOx than four-stroke fuel-injected counterparts for equivalent fuel use, incomplete combustion means leaf-blowers and the like have a massively outsized impact on air quality, especially for the operator. The good news is that electric systems are now cheap and lightweight enough for most applications, which is fantastic.
From that article you linked to, while it's true that the figure for "NMHC" is higher (0.005g/min to 1.495g/min, yes this is almost 300x), this is just the worse case result - look at the other figures too: NOx is 0.005g/min to 0.010g/min (a doubling) and CO is 0.276g/min to 6.445g/min (about 25x). Of course, all of these figures will be dwarfed by CO2 emissions which will be roughly proportional to the amount of fuel used.
Now consider the time spent doing these activities - 30 minutes compared to 47 hours (according to Google), so multiply all those per minute figures for the truck by 100 to make a fair comparison. Now, the blower is triple the "NMHC", 2% of the NOx, 25% (EDIT: I made a mistake before) of the CO and about 0.2% of the CO2 (100 gallons vs 0.2L).
Honestly, I don't know what "NMHC" includes - as presumably it doesn't include NOx as that is called out separately, but whatever, it's just part of the emissions and so the claim that the image makes of "fewer hydrocarbon emissions" is clearly false. I'll agree that it seems to be a bad summary of the article which was "the emissions are dirtier".
EDIT: Actually, I should add that the image has done a good of raising awareness that leaf blowers are clearly pretty dirty even despite not using much fuel, so I guess it served its purpose even if after looking at the sources, I'm still not really sure by how much in actual terms, because the things I know to look out for - CO and NOx - aren't as bad as claimed, only NMHC, something I've never heard of before, and I've no idea how bad that is.
That's a fair point, although I'm not sure I'd personally categorize it as "actively deceptive", I agree it could be a lot clearer about "Emissions", and empirically you're clearly right that is has some folks confused. To be charitable, like all science communication it's trying to simplify a very nuanced topic and probably could be improved. I do think the illustration of how leafblowers pollute at ~300x the rate a car does is largely true, and the GHG/environmental impact is a lot messier than just primary CO2 emissions.
(Disclaimer: I've never met or interacted with Nicole, but I know people that have, so I'm likely biased to assume good intent.)
I'm not sure about actively deceptive. What else does it mean? Hydrocarbon emissions are what you get when you burn hydrocarbons - H2O, CO2 and CO. If it's referring to a different thing, it's not a hydrocarbon emission.
The image, and the claim within it, is actively deceptive because it's not true.
It's a little messy, and it's not clearly disambiguated what concept the artist is referring to. I agree "Hydrocarbon emissions" could reasonably mean what you get when you burn hydrocarbons, but that definition can also include literal "Hydrocarbon emissions": uncombusted or partially combusted hydrocarbon fuel being emitted out of the exhaust after ignition due to non-ideal combustion. Like you mention, this is alongside CO, as well as elemental black carbon and other trace combustion products like NOx etc...
You're right that there isn't orders of magnitude more CO2 from a two stroke, but there are orders magnitude more CO, as well as these literal "emitted hydrocarbons", which I think it what the direct claim the art intends to illustrate with either dinfinition. But the fact we're having this conversation is evidence the claim at the very least isn't clear.
Relevant disclaimer: I work for QuantAQ, another distributed Air Quality Monitoring company, my opinions are my own.
With proper calibration, the PMS5003 sensor AirGradient uses is a real and fairly good measurement of PM1 (<1 micron), which is often fairly correlated with (and makes up much of) PM2.5. However, the sensor cannot really see particles larger than 1 micron, which sometimes matters for PM2.5 (when the PM1:PM2.5 ratio changes, like wildfires), and very much matters for PM10. That's why QuantAQ uses a PMS5003 combined with a different sensing technology (an optical particle counter) to actually measure PM2.5 and PM10, not just extrapolate from a PM1 measurement.
I love the work that AirGradient is doing making AQ sensing technology open and accessible, especially to hardware hackers. The BOM cost of the OPC sensor we use alongside the PMS5003 is more than the cost of the entire AirGradient kit, so it is valuable to have options in the market depending on application.
Excellent overview; I think the space for air monitoring is only going to grow as time goes on, especially after COVID and with increased awareness of the downsides of airborne particulates and high CO2 exposure.
I've been setting up PM sensors in my home since installing a log-burning stove last winter; I've found a couple of interesting things thus far; Notably, despite the FUD, air quality _in_ my home isn't affected by it.
What does affect air quality in my home massively, are the people parking their vehicles surrounding my home morning and afternoon as they drop off and pick up their children from the nearby school. I see a 10x ambient measure at these times. I was starting to think about how I could use BLE monitoring to track this somehow (the other thing my home is surrounded with at these times is devices broadcasting Bluetooth "beacons"(?)); could be an interesting experiment.
The next step for me is really knowing what and how something can be done about the pollution (generally speaking); Monitoring is great, but what's the next step?
How much would it cost me to install a modulair-PM on my rooftop, assuming I did the work? I’m just an individual, i just want transparent information. I don’t want to “contact sales” as it says on the link from that blog post.
Short answer, $1500 + $300 annually (includes cellular, realtime data calibration in our cloud, and full API access).
Relevant to the topic above, we are building for a slightly different market than AirGradient, and want to make sure we're an appropriate solution for potential customers -- you might genuinely be better served by a different product! As such, we're currently not optimized for low touch single sales to individuals, but it's something we'd like to improve in the future.
If you do contact sales, and we end up being a good fit for what you're looking for, we'd love to help!
I’d be interested in a prosumer version of this that ditches the LTE and cloud system for wall power + WiFi or (ideally) PoE with a fully local backend via home assistant
The BME 680 does temperature, humidity, barometric pressure, and VOC gas sensing. Maybe there's something reasonably-priced that does particulates, but this doesn't seem to be it. What's a good alternative?
That looks awesome, thanks! I already have some ESP32 devices acting as weather sensors that log to a Raspberry Pi over MQTT. I might have to add this too.
I'm just going to take a minute to be happy to see this friendly exchange of information in public between two people who are both knowledgeable about the subject and competitors.
I'm a big fan of QuantAQ. Speaking from a PM10 perspective, your monitors perform on-par with high-cost regulatory monitors in a much smaller package and far lower cost. I also love the open API. That said, I would love to see a locally-hosted variety for environments where LTE/cloud is not accessible/preferred.
At QuantAQ, we build air quality sensors and software to help customers measure, understand and act upon the air around them. In other words, digital twins for air quality. We're a very small team that is growing, profitable, and doing genuinely fun and impactful cleantech work.
We are looking to hire a software engineer to help us build out our cloud platform for air quality sensor networks: you'll have to be comfortable full stack, and hopefully you're a better front-end engineer than any of us are. Our web stack is mainly Python (flask), Postgres and Docker, with more buzzwords in the job ad.
At QuantAQ, we build air quality sensors and software to help customers measure, understand and act upon the air around them. In other words, digital twins for air quality. We're a very small team that is growing, profitable, and doing genuinely fun and impactful cleantech work.
We are looking to hire a software engineer to help us build out our cloud platform for air quality sensor networks: you'll have to be comfortable full stack, and hopefully you're a better front-end engineer than any of us are. Our web stack is mainly Python (flask), Postgres and Docker, with more buzzwords in the job ad.
I've been fortunate enough to take a class with Sanjoy (The author) covering this book. It was an incredible experience, and he's one of the best mentors I've ever met.
One of the things I found most interesting from the class that you couldn't get from the book was a particular method of assessment. He would pose a question, and allow you to give a weighted answer across a few options, to better understand your uncertainty. It was a very useful pedagogical technique, and allowed me to quantify my understanding in a very tangible way.
I highly recommend the whole book, but especially the first section, on breaking things down.
To further the aside, there are many Olin buildings on college campuses, all due to Franklin W. Olin (the founder of the Olin corporation). There is also now an entirely standalone Olin College, which is the biggest shared name with this project in my opinion.
