The commercial air flow meters are easy to use and come with NIST traceable certificates but you can make your own air flow meter for lab use that rivals the $1k meter performance.

There are many variations to this method but essentially you need a graduated cylinder column with a hose barb fitting. Attach the air intake line to the column then time how long it takes a soap bubble to travel a portion of the column. Calculate the flow rate as volume / time.

For the column, you can buy a cheap plastic graduated cylinder and then drill a hole in the base, tap the hole, then screw in a plastic barb fitting and goop on some silicone water sealant. OR, you can use a graduated burette tube (like a pipette tube). Wet the inside of the column with soapy water and then mount it in a lab stand with the open end facing down and the air hose fitting facing up. Attach the air intake tube from your pump sampling apparatus to the column. With air flowing in the open end, take a cup of soapy water and briefly submerge the open end of the column in the soapy water to make a soap bubble (alternatively, for higher flows you can blow bubbles in the cup with a straw and just touch the bubbles to the column). With a stop watch, time how long it takes the soap bubble to pass a certain distance (volume) in the graduated cylinder. The flow rate is then calculated as ( [end vol.] - [start vol.] ) / [time]

You may want to add a water knock out to prevent any soapy water from entering the tube and fowling anything upstream. For a knock out you just need any air tight container (e.g. mason jar, tupperware) with a couple hose barbs.

You also need to size your column appropriately for your flow rates. If measuring 0.1 to 1.0 lpm, I would consider using a 500 ml column. If the bubble travels 400 ml in the column then it would take 240 to 24 seconds at 0.1 and 1.0 lpm respectively.

I would check the flow of the system at at least four points and then make a correction curve if necessary--and it will probably be necessary.

let me know if you have questions.

That sounds like a great idea. The flow meter remains the most expensive component so reducing this cost is clutch. This sounds like something that we should experiment with at the barnraising in a couple week's time. Thanks very much for the suggestion!

Exciting stuff! Do you think this method could work for other pump based tubes like these http://www.buydraegertubes.com/hydrogensulfide02b.aspx

I think so. Those tubes require one pump on the Draeger Tube proprietary pump. So if one were to rent their proprietary pump for a single day and attach a flow meter to it (preferably one of the higher end ones with a large monitoring range), then you see what the flow rate through the tube is and that could likely be replicated by an inexpensive pump-turned-vacuum.

Their sampling seems to only take 55 seconds, so perhaps the flow rate is much higher than in the method above. If that is the case it, and a cheap aquarium pump could not muster that much pulling power then one could experiment to see if a longer sampling time with a equal total volume pulled through the tube would yield the same results. (when conversing the the lab that made the tubes I used, they were not positive that such a method change would yield precise results but it sounds worth some experimentation!)

Yesterday I was at APHA and heard mention of H2S biomonitoring in urine (but I can't quite remember the metabolite that would be assayed) . Would be interesting if pH like test strips could be made for H2S biomonitoring.

I just bought the air pump components and a graduated cylinder to try and compare @davidmack's cheapo air meter... can't wait. I'm going to try using this to control airflow to a shinyei dust sensor.

Hi, Nick - I'm working on some sketches of the kit; do you have any photos of the entire thing, just a bit more zoomed out? Thanks!

On reversing the aquarium pump: Be sure and reverse both diaphragms!

@danbeavers were you using the whisper pump or another pump? I think the whisper pump can handle one pump and one vacuum if need be

Oh, I was making assumptions. I had not seen a pump that was set up to have 2 different attach points. So never mind about my comment!

following up on our discussion about trying to calibrate the pump without the flow meter-- could we calibrate the pump at our office, and then have it run on a timer? that would probably make the test easier-- we could do automatic shutoff. I think that makes the process even easier.

While regular wall switch timers are affordable, I can't find one with one-minute accuracy that isn't a wall plate-- so we'd have to wire our own electrical box, which I don't want to do. They're usually in 15 minute increments, which won't be precise enough:

I think the best interface is an analog dial. I have two darkroom timers here, and vintage timers are 10-40 bucks on ebay, but new timers are too expensive. Dunno if we want to deal with the headache of sourcing a bunch of different things.

This is great. The vintage timers are pretty amazing although they add a lot of weight and thus money and are not tamper proof. We could just include one of these 4 dollar timers. as the next best thing. http://restauranttory.com/taylor-58-5842-58-mini-digital-timer-36200-minimum-order.asp?gclid=CJWEj-fTusUCFUwXHwodMU0A-Q

I bet our calibration could last for at least 5 tests. having people pass them along to the next person (for a set batch) as opposed to sending them back and forth would save a ton of money on postage and personal shipping time.

when we use the mini buck to calibrate we can see how many tests each calibration can hold. setting a time/test duration is a pretty standard.

if in the long run we could hardwire our own timer that would be ideal

The vintage timers are probably far too massive. I'm not as worried about tampering as I am about good faith forgetfulness-- I don't always hear the washing machine buzz in my house-- I'm glad when things turn themselves off.

Gotcha. In my testing in 2011 i did some of the test and mailed test kits to those i couldn't meet in person. altho it wasn't statistically significant the results of the self test were lower, probably meaning that they didn't prepare their homes (close windows and door for 24 hours prior to testing) as thoroughly as they did when someone was showing up or they turned it off too early. So i'm not so worried personally with forgetting the test is happening (leaving it on too long would result in a higher result which didn't appear to be the case), i say "tampering" because our case will be stronger if we can claim it to have an un-mess-with-able sampling time.

thinking speculatively, could we have a system that you call when you start the test and it calls that number back after the allotted time and then they confirm into a voicemail that the test has been stopped?

@danbeavers just sent me a great email about cheap flow meters. he told me I could repost it, so am putting it here:

"I have been thinking/researching a cheap flow meter. I sent a couple of e-mails to vendors that advertise seemingly appropriate meters for under $30 and have not received a reply yet.

I have also been thinking on how to construct a variable volume flow meter with the least amount of tools and cheaply.

Calibration: I don't remember who and can't find the e-mail but someone suggested the bubble in the graduated cylinder method. That seems to me to be the way to go. Use this to calibrate the homemade flow meter. The cheapest way to go here is probably a disposable plastic pipette. I have not priced or sourced any yet but I can't believe that they would be very expensive. They come in lots of sizes so some appropriate size should be easy to obtain. I got a free sample of several sizes at the Pittcon conference in March so I should be able to determine an appropriate size fairly easily.

Flow adjustment: A simple screw clamp or similar device on an aquarium hose should be easy to construct or find.

Flow meter: The most accurate measurement at 0.3 lpm is most likely the variable volume flow meter or rotometer. The problem with constructing such a device is how to make the conical bore. Drawing a melted glass tube is probably not easy enough to accomplish. I attempted a similar process with placing aquarium tube in boiling water and stretching it. That may be a viable solution but more research needs to be done but I don't have much confidence in the process.

My neighbor and I were talking about this and we came up with an idea to make a stepped bore. Since we only want one flow rate (perhaps 2 if measuring a different chemical also) then when obtaining the correct flow rate the ball would be no lower than one step in the bore and no higher than the other step in the bore. To construct this would require a standard ball, a set of drills, and a hunk of clear plastic. Drills may not be accurate enough so I would expect reamers would work in that case.

Determining the correct size of the holes and ball can perhaps be calculated to reduce the amount of experimentation required. I have not researched the formulas involved but expect that they are readily available."

This might be one way of getting our per-unit cost down. any thoughts on this @mathew

Thanks Nate for posting my thoughts.

I wanted to comment on Mathew's idea of calibrating the pump before shipment. I assume what is meant is that a fixed restriction is attached to the pump to get the proper flow rate for the formaldehyde tube. The problem with this idea appears to be in preparing the measuring tube. When you score and break both ends of the tube the resulting two orifices will be different sizes each time this procedure is done. They will be scored on a different point on the tapered glass seal. Thus the flow restrictions from the tube will vary. I do not know if it will add sufficient enough variation though. Perhaps a way around this is some fixture to use so that a consistent diameter is scored thus resulting in a standard flow thru the tube.

Another approach to this problem would be essentially what Drager does. We could construct a manual vacuum pump. It would essentially be like a bicycle pump but with the valves reversed and a spring to return the plunger. We know the Kitagawa 710 Formaldehyde tube requires 300ml/min×30min which is a volume of 9,000ml or 9 liters. If the vacuum pump moves 1 liter per stroke then it takes 9 strokes over 30 minutes. The vacuum pump could have an orifice so the spring would slowly push the plunger out. Clear indications that the plunger has been fully pressed and fully returned would be needed. A mechanical counter could be used to count the strokes or a place on the DIY Formaldehyde Test Photo Card could be used to make tick counts. Some big advantages are that both the electrical pump and the flow meter are not needed.

I just found an Instructable that is SIMILAR to what I am describing: http://www.instructables.com/id/Make-a-High-Volume-Manual-Vacuum-Pump/

@danbeavers hmm... the variability in the break and its effect on flow could be quantified with a more accurate flow meter. I wonder how variable it would be-- they provide that little bauble of glass on the end to try to make it break in the same place.

@davidmack suggested the graduated cylinder method in the first comment of this note.

@nshapiro I wouldn't feel comfortable with a stepped, DIY bore to an air meter- I'd seriously doubt its accuracy. On making the system tamper-proof- we could use a "time delay relay" which is a standard piece of automation/security/HVAC equipment ($8-25). We could then set the timer before the pump leaves our office, and the user would just hit a push button to start the process. Time delay relays generally reset the counter with every button press though-- so we'd probably want to get a two-channel relay and use the second relay for turning off the button as soon as the timer starts EDIT-- a quick search reveals there are also "self-locking" time-delay relays. Someone would have to take the whole monitor apart to reset the pump timing.

Mathew: Why would you doubt the accuracy of a stepped bore? A reamer will provide very accurate hole size and ball bearings have very accurate dimensions. The combination of the two make a very accurate orifice. But I think I like the manual volumetric vacuum pump best of all. Also, I wonder if there is much variation in the flow restriction from one packing of chemicals in the tube to the next.

+1 to the time relay delay it could be adhered to the belly of the pump.

Digging the flow meter debate!

@Danbeavers If you ream it you're right-- I was thinking about just a drill hole and that note. Vibrations on either a cheap drill or a poorly secured work piece will introduce significant wobble in drill holes.

I really like the idea of a vacuum pump, but even with sensors watching the pump stroke, someone could still cheat or just miss on the process. I'd like any new hardware we introduce to make the process more fool proof.

I'm trying to think about how everything might go together for shipping-- I'd like to put the monitor in a plastic suitcase with a top panel cut to give storage spaces for parts, and a folder for all the photo cards on the inside of the lid. It might also be nice to put a standard place for the tube to go:

I think this looks great!

The air intake is in one end of the tube. so we could have one less inlet to cut in the top panel and we could have it so that the tube can flip up vertically so that it can sample without pulling air over the foam of the packing material or the lid that may be off gassing formaldehyde.

Mathew: Perhaps I am not understanding some objectives here. Is the idea to be able to use one of our measurements in court? I doubt that that is the purpose. Is it then more about concerns for PL's reputation or is it negligent operators? Accomplishing a "fool" proof system sounds rather expensive to me. Probably the objective is reasonably compromising between cost and automation. This http://www.komyokk.co.jp/kweb/pbtop.do?pbno=1002&je=1 is more like what I am thinking would be best. Has anyone received a quote on this? Would it work with the formaldehyde tube?

Answering my own question: "The KITAGAWA air sampling pump reproducibly draws 100cc of sample air" So that would be 90 pump cycles thus makiing it impractical. If we can accurately design a similar pump with a liter volume then it becomes 9 strokes.

i actually have three of those pumps on my desk. they have a counter up to 10 and a thermal ring (that doesn't appear to work) for temp and a tube tip breaker (a glass tip is stuck in one of them). I think they work best at high levels. The tubes that go with those pumps have a higher (10x higher than our tubes). I think the larger volumes of air work better for lower levels.

I totally agree that we are looking at a balance of cost an automation. we also want to make sure people are getting the most robust data that we can give them and to make every test as similar to the next test as possible. how to we demonstrate trust in the capacity of non-scientists that the chain of custody system so thoroughly disregards while also encouraging consistency and validity? we could have some kits with the auto shut off and some with just a timer and see how their levels compare?

I think people will be pretty good at unplugging their kit right on time, its something that you don't do everyday (like cooking or running the dishwasher) so I think people are pretty attentive to it. We can also get very bold and underline-y in the instructions about the timing.

@danbeavers I am interested in making our data usable in court, but I'm also trying to be budget-sensitive. we're looking to mail these around from house to house, and over the long-term our biggest costs will be related to logistics, not the upfront costs of hardware. The automation option I'm looking at adds probably ~20 bucks to the finished cost. A solid case will add another $30. While optional, these things potentially make packing, shipping, and operations more seamless.

While a solid case and automation will add about a $50 to the equipment, if it prevents 10 or so mistakes over the pump's lifetime we'll have saved money: Every time someone screws up and decides to either re-run a test with a tube reserved for the next user, or holds onto the pump waiting to re-run an test is going to result in a bunch of logistical rejiggering over here. It is also easy for someone to miss-pack a box and break something, or disconnect stuff that will need to be plugged in, potentially introducing more logistical issues.

I'm not thinking about negligence so much as everyday carelessness-- no matter what we do even the best of users are going to screw up sooner or later and give us a headache, but we can design to minimize the chances of that happening. If it makes our data look more legally admissible, all the better.

Hey @mathew, do you think the goose neck thing I put together is worth pursuing, or is compatible with the box idea you posted? Just want to know how to proceed on visual design on the video I'm working on.

I think for the neck unit-- we need to have a flow meter in the flow. That is a perfect configuration to use in the buy your own kit, whereas we're trying to design the boxed kit as a thing we'd ship around in a library. very different design considerations between those two.

@Danbeavers this bubble flow meter is super simple

OK, cool. Are there options for the flow meter Nick's used above, in terms of size, etc? It has to be upright, right?

it has to be upright, yes. what other options are you thinking of?

Just size, really. I have to order one of these too so I can use it in the video I'm working on.

1 VWR FLOW METER ACR 2IN.1-1LPM ($74.62 + shipping).

Is that supposed to be 1-10 LPM? Thanks!

.1-1 is correct. I have three right now. i can send you one. but you would need to get the valves. my supplier just ran out.

Oh I see! I had read it as 2IN. My brains a little slow right now, sorry. I need to be watered.

I have small plastic valves, had to order 5. I'll send one out.

Actually that'd be super great. If they'd arrive before next Thursday, Nick and Mathew, could you send them to Propeller?

Propeller c/o Public Lab 4035 Washington New Orleans, LA 70125

Mathew: The "super simple" flow meter you linked to is what DavidMack was describing in the first post comment. That is real good to calibrate a meter or measure a small volume and not so good to control a flow rate by.

Luc L. and I did some brainstorming this evening at the Picayune Make meeting. We went over all the ideas I had and he came up with some more. The least expensive and practical we ended up with was a manometer attached to two tees with a needle valve (or similar) between the two tees. You essentially end up with this configuration: https://en.wikipedia.org/wiki/Flow_measurement#Orifice_plate The way you operate it is to place some colored water in the manometer (U shaped tube) and move the measuring stick so the 0 is on the meniscus. After you prepare and attach the formaldehyde tube you turn on the vacuum pump and adjust the valve so the meniscus is at the previously calibrated 0.3 lpm value. When you are done with the measurement you disconnect the manometer and plug both ends so that the water does not leak in the shipping container.

The cheapest flow meters in the appropriate range that I could find:

$18.90 http://www.ebay.com/itm/0-1-1-5L-Oxygen-Air-Flow-Meter-Gas-Flow-Meter-With-Copper-Connector-measurement-/331371524139?pt=LH_DefaultDomain_0&hash=item4d2748302b

$27.00 http://www.dwyer-inst.com/configurator/index.cfm?Group_ID=326

I'm arriving tomorrow -- if you could order those I'd buy them off of you when I arrive, or we could just order tomorrow night when I get in. I just need to go to sleep really badly right now :-)

Perhaps [@Stevie](/profile/Stevie) could order them.

@Danbeavers what if we put a standardized, small orifice on the broken end of the glass tube that is always smaller than a hole broken into the tube-- then we'd have a standardized airflow (provided the flow through tubes is consistent.

Here is a photo of the tip breaker that is on some of the analog pumps. We could also pre-score them in a standardized way to create consistent openings.

Mathew: That sounds like a mechanical engineer (which I am not) question. I SUSPECT that it will be the sum of the restrictions because of the low pressure we need.

Nick: I am also not a glass expert. My thinking is that the breaking of glass should happen soon after scoring. The reason is that glass heals. Aged scores would probably result in breaks that would be either hard to accomplish or not consistent in shape. But again, I am not a glass expert.

@Mathew The following is from a mechanical engineer friend of mine:

Assuming you could pull a reasonably good vacuum on the orifice, you could get a fairly constant flow. What you would need to do is ensure the flow through the orifice is sonic, or "choked" as it's often called. The rule of thumb for choked flow is that the absolute pressure of the inlet is 2 times that of the outlet. Since atmospheric pressure is about 14.7 psia, the vacuum pump would only need to pull the outlet side down below 7 psia, which I don't think would be much of a task at .3 lpm. The advantage of using choked flow is that once the orifice is flowing sonic velocity, the downstream side of the process no longer affects the upstream side. In other words, you could go from a perfect vacuum to a marginal one and as long as the pressure is below 7 psia, it won't change the flow rate. The bad news from a practical standpoint is that the required orifice size for .3 lpm at 14.7 psia is down below .01" according to my calculations (using two different tools), which is about as small a hole as you can drill. You could probably find some kind of laser-cut orifice smaller than that. As far as the upstream flow resistance, you would have to have some extreme restrictions to give you much effect at .3 lpm - the hose ends shape would be negligible. If extreme accuracy is not an issue and the orifice size is obtainable, I think the vacuum method is reasonable. In that case, you'd also want to use some kind of filter with a smaller rating than the orifice size to preclude clogging such a small hole, which would give you a false negative reading due to lack of flow.

Another observation: It would be good to have a check list for the measurement. For example:

1. Verify that the humidity is above ?? and below ??
2. Record the 6 digit number above the word Japan
3. Score and break both of the tapered glass ends
4. Nam sit amet purus sit amet eros blandit ultrices.
5. Fusce viverra odio quis est elementum convallis.

This is just an example.

Our portland office has .026ppm (26ppb) of formaldehyde when measured over by our packing and shipping area. That's over 3 times the chronic danger level. fun!

I ran a formaldehyde test over by our shipping station, now that the flow on our lending kit is calibrated. We got 26ppb, or about 3.5x the chronic exposure safety level. Fun!

beginning:

end:

Just repeated this test today at 26 degrees, and temperature corrected we got 25ppb. So good news: consistency. Bad news: consistently hazardous conditions in our office!

It appears you need some more ventilation or perhaps start using the remediation process. http://publiclab.org/notes/nshapiro/10-20-2014/diy-indoor-air-quality-remediation It would be interesting to see how much reduction you can get.

Any thoughts on using a smaller pump? If you're only trying to draw 0.3 SLPM, there are a couple of pumps (see the attached datasheet and catalog) used for a similar application that may be able to draw enough air through the draeger tube. There are probably more out there like these as well.

I'm unsure of the cost of these, but I can request a quote and find out. If they perform as well as the larger aquarium pumps, it may be good for reducing the overall footprint of the device (the smallest for air is the SP 200 EC-LC, at 33 x 25 x 16 mm) and significantly reducing shipping costs. A handheld formaldehyde monitor would be really cool.

SP_PUMPS_DataSheet_ClassEC_Series200_Type200-250-270EC.pdf

SP_PUMPS_ProductBrochure.pdf

@kkoerner Hi Karl! Great find! It would be fantastic if you could request a quote on these. We are planning on have a lending library of these tools that would be shipped around between people needing formaldehyde testing so over time the shipping costs will really add up and weight saving would save the community a good amount of money.

@mathew might also be interested in Karl's pumps as he has really taken the lead on the lending kits.

@kkoerner, I tried a similar pump when I was researching this and it created too much turbulent/pulsed of the air stream to get a good measurement when restricted to low flow levels. I ended up using a 3 diaphragm low voltage 12VDC pump, I did a writeup with low voltage DC pumps that work: http://lukeskaff.com/diy-formaldehyde-testing

Great writeup, Luke! cool work.

LukeS good to know, thanks for testing out those pumps and posting your results!

Thanks, @Dave-O ! We did some side by side testing of this method next to an EPA approved method and we found these tubes to be cross contaminated by acetone and acetaldehyde, and these Draeger tubes may as well. Here's the note: https://publiclab.org/notes/gretchengehrke/10-07-2015/formaldehyde-measurement-testing-public-lab-s-kit-with-doh-s-equipment and since then we've started development on a new system https://publiclab.org/notes/nshapiro/10-17-2016/public-lab-awarded-nsf-funding-to-help-develop-low-cost-community-formaldehyde-monitoring-technique Jeff just wrote up a new study design for HCHO if you've got some HCHO that you're interested in assessing https://publiclab.org/notes/nshapiro/10-17-2016/public-lab-awarded-nsf-funding-to-help-develop-low-cost-community-formaldehyde-monitoring-technique

You're welcome, @nshapiro ! Is this the same EPA approved method that you used? https://www3.epa.gov/ttnamti1/files/ambient/airtox/to-11ar.pdf And I do have some HCHO

Oh, also just for ref, the idea i posted is here: https://publiclab.org/notes/warren/02-09-2017/a-kit-to-design-your-own-formaldehyde-experiment

Hi, my search for formaldehyde testing brought me to this article. I'm wondering if you could help me with some advice - I'd appreciate it. I'm very suspicious of the floor we got installed over a year ago in our house - suspicion triggered by health issues. Just after we got it installed, I heard about the scandal involving Lumber Liquidators. I asked the installing person about his product but never got a proper answer. I asked about laboratory tests but I was quoted over $600 which I'm not able to pay. Please let me know what would be the best way to test our flooring and air quality. Also, is it too late to test this? The floor was installed 1.5 years ago. Thank you very much

Hi @indy if you're based in Indiana, as your name hints, you can report your symptoms to the stat DOH and they might come out and test for free. Indiana is the only state in the country that does that! Otherwise, you could try: http://homeaircheck.com/products/ or http://www.acsbadge.com/formaldehyde.shtml if the results are high we can talk about how to put together a low-cost formaldehyde remediation kit https://publiclab.org/notes/nshapiro/09-26-2016/build-a-plant-based-air-purifier (sorry for the delay!)