Public Lab Research note


Dissecting Humidity from Particle Count in Frac Sand Sampling

by Jdahlen | June 24, 2014 19:32 | 3,134 views | 17 comments | #10611 | 3,134 views | 17 comments | #10611 24 Jun 19:32

Read more: unstable.publiclab.org/n/10611


What I want to do

Dr. Crispin PIerce (UW-Eau Claire) and I, Jonathan Dahlen (UW-Stout), are researching into particulate matter 2.5 and 10 around frac sand mines. What we are looking to do is create a device that can sample both PM 2.5 and 10, but also collect measurements of temperature and humidity.

My attempt and results

We are currently using direct reading instruments (Dusttrak 1 and 2, Dylos), but humidity is playing a factor in our results and we would like to eliminate that problem so we are able to just collect particles.

Questions and next steps

We are wondering what some of the people from Public Lab have for thoughts or ideas on our project.

Why I'm interested

We would like to add to the array of inexpensive tools in monitoring around frac sand facilities.


17 Comments

Cool!

Does the dusttrack or dylos have a sensor-out port? I'm thinking we could hook it up to another data logger and collect humidity data on the same time stamp. I have an SHT 15 temp and humidity sensor here and an arduino-based datalogger here that writes to an SD card-- if we could get it to read the Dusttrack then you'd be set.

can you find any info on the Dusttrack's communications protocol or output?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Also, I noticed your two Dusttrack monitors you're using had different air intakes (pic below). is the cyclone really important? would it improve the data from the dylos to have a cyclone?

talkin' frac sand with Crispin Pierce UW Eau Claire

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Hi, thanks for the post -- are the Dusttrak & Dylos optical devices, like the one the DustDuino project's using? http://publiclab.org/tag/dustduino

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


yes, they are also optical, like the Dustduino and Speck, but they use lasers instead of focused LEDs.

Reply to this comment...


Sorry for the late reply, but the Dusttraks and Dylos, to my knowledge, do not have sensor out-ports. If I would be able to see what the connector for the sensor out-port would look like, I could look into the Dusttrak for the same port. We're in the process of buying an environmental enclosure for the Dust trak II which will be very similar to the Dusttrak I environmental enclosure, seen above. Another improvement is that we are planning to buy tripods. As for the cyclone on the Dylos, I'm not sure if that will be much of an improvement since it already sorts out the PM2.5 and PM10 particles.

Reply to this comment...


How does data come off? USB? Serial? There's got to be something

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


We usually just record the data right after our 24-hour sampling, but the Tech team at UWEC is helping us figure out how to retrieve our continuous data from the Dusttrak 1 and Dylos. The Dusttrak only has an ethernet cable, so we plan to get an ethernet to USB cable and for the Dylos, I think it's a 9 pin serial port.

Reply to this comment...


Oh Ethernet! Cool that probably isn't too hard to talk to but a computer of some sort will be needed.

Serial is great! We can probably get it to talk to a datalogger/microcontroller

Reply to this comment...


Did you see the article AmySoyka posted about the Dylos? http://publiclab.org/notes/Schroyer/06-26-2014/broadcast-dust-readings-from-a-smartphone#c9306

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


Yes, we're hoping we can get the Ethernet and USB connection to work for the Dusttrak. As for the Dylos, the only information we are able to take off it is the hourly readings. I read the article and that is great! That will hopefully help for experiments in the future.

Reply to this comment...


@donblair-- the author doesn't seem to understand the terminology in use. The conclusions are based on a misunderstanding of what PM2.5 actually means I posted this comment:

PM2.5 is not the same as 2.5µ. it means particles with a mean aerodynamic diameter of 2.5µ. So a PM2.5 measurement is 50% above 2.5µ and 50% below. it includes 0.5µ particles. The same goes for PM10, it includes 2.5µ particles. https://en.wikipedia.org/wiki/Aerosol http://publiclab.org/wiki/particle-sensing

Reply to this comment...


sorry for the late reply but as you probably discovered as the air sample relative humidity increases, the moisture increasingly interferes with the optics. a few possible solutions:

  1. heat the air sample to reduce the relative humidity. Ideally you'd control the temp to maintain a constant RH. this is a battery killer though and typically only works when connected to the grid.

  2. you can dry the air sample:

2a) mix with dry air--easier said than done!

2b) dry with a nafion dryer (from permapure.com ). They can be used a few different ways but essentially you need heat, compressed air, or a strong air pump (differential pressure) to drive out the moisture.

note that whenever you heat or dry the sample you're changing the particle size so the challenge is to reduce the RH just enough to minimize water interference but not dramatically change the sample.

Reply to this comment...


Hank Boschen's Dylos monitoring system in Chippewa Falls records readings from a Dylos every 10 minutes and posts to a website. It also includes weather data from the Eau Claire-Chippewa Falls regional airport. Why not talk to him and find out how his system is set-up? Matthew's post on frac sand has an image of a typical graph. It should be simple to find a temp-humidity logger if you wanted local values. Try Onset data Loggers. Why try to reinvent?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


I'll connect with Hank on his Dylos-- we'll see what he's doing. there are a million ways to do this that have already been done-- so yeah, we should pursue those.

Reply to this comment...


After reviewing these posts it became apparent that there are two questions here. One is about "eliminating" humidity, possibly by mechanical means; the other would be about taking humidity into consideration, for example, in the calibration or validation regression. In fact, I will be brazen and ask how either of the monitors is calibrated and converts counts to concentrations. If any use is made of comparisons to a Wisconsin DNR monitor, it would matter if that monitor makes corrections for humidity.

Reply to this comment...


@jefffalk TSI explains the instrument and its calibration in documents on its web site:

http://www.tsi.com/DUSTTRAK-Aerosol-Monitor-8520/

see "theory of operation" and also the "Rationale... for Photo Calibration Factor"

But in short they emit some dust (but just not any dust, it's Arizona Road Dust, with its very own ISO standard) and calibrate it to a mass concentration:

"The DUSTTRAK™ monitor is calibrated against a gravimetric reference using the respirable fraction of standard ISO 12103-1, A1 test dust (Arizona Test Dust). This test dust has a wide size distribution covering the entire size range of the DUSTTRAK™ monitor and is representative of a wide variety of ambient aerosols. The wide range of particle sizes averages the effect of particle size dependence on the measured signal.The sensing volume of the DUSTTRAK™ monitor is constant and is defined by the intersection of the aerosol stream and the laser beam. Mass is determined from the intensity of light scattered by the aerosol within the fixed sensing volume. Since the sensing volume is known, the information can be easily converted by the DUSTTRAK™ microprocessor to units of mass per unit volume (mg/m3)."

if you look at a TSI calibration certificate you'll see they calibrate four points, just found this googling, see http://xtralis.com/misc/VIPAC%20Test%20Report%20-%20ASD%20Performance%20and%20Reliability%20Testing%20in%20Dust%20Laden%20Environment.pdf

It looks like the cal point targets are 0, 1, 5, and 25 mg/m3 and the acceptable tolerance is +/- 10%

they also note that above 70% RH the instrument reading is "over-estimating" values.

I don't think a water correction is going to be fruitful. It might work up to a certain RH but once the RH goes too high it's going to saturate the response. It depends on your location but it's not uncommon to have fog at day break where I live and that'll flood the detector. Theremo/MIE actually sells a heated inlet with their DataRAM instrument to reduce water interference but it requires power from the grid.

The OnSet Hobos are great. I'd use one just for QC and then if you see high RH you might deem the PM invalid.

Reply to this comment...


Thanks for the info on the DUSTTRAK. It's good to remember everything comes with advantages and disadvantages. Adjusting for or eliminating humidity carries its own problems. The simplest thing as suggested is discarding readings at high humidity times. A comparison at higher humidity times with the Wisconsin DNR monitor in EauClaire would seem worthwhile for comparative information.

Reply to this comment...


Login to comment.