Testing sample size variation with Clear Choice Labs kits - should be 10ml

[AusPhil]

Firstly I would like to thank Brett from Clear Choice Labs (CCL) for answering my questions both promptly and honestly. This is excellent customer service and the obvious choice for Australian members for their test kit resources.

Whilst the following specifically applies to my personal experience with the CCL test tube sample volume sizing there is of course potential other kits may see variations in sample size i suppose.

Last summer as part of putting up the 12x24 Bestway AGP and being keen to follow TFP process I ordered a fresh water testing kit from CCL that came with all the relevant tests. In addition we adapted a generic speed stir from ebay to handle the test tubes from the kit.
I had ordered five magnetic stirrers, on for each of the tests and sitting on my desk was a jewelers scale with 200gm total with 1/100th gram resolution that i used a lot weighing out SMD components. Ideal for weighing out sample water volumes.
I couldn't find it documented in any of the CCL doco what the volume was but a check against the lines on a couple of the tubes showed that like the taylor/tft kits 10ml was the sample volume being used. The SG of pool water for this purpose is close enough to 1.00 so that 10ml = 10grams.
All good this let me fill each test tube quickly and consistently.

Fast forward to Nov 2019 and we are getting ready to install a SWCG so I ordered and received my Salt Water test kit. Got all the bits out, filled the tube with 10gms (10ml) of water and noticed the water was above the line .... mmmm ... tip sample out and fill to the line, 9ml .... ok assumed the test used a different sample size.

Added more salt based on the test and retested on Wed 6th .... for giggles i tested using 9ml and 10ml samples, tested twice and got consistent 400ppm difference with the 10ml being 400ppm higher.
Checked the other test tubes and the CH tube had the line a little low ... around 9.5ml.

Fired off an email to CCL and Brett replied that the sample size for all their tests is 10ml and that the labels are currently applied by hand and may have small variations.
These variations generally won't be of significance for pool owners and they are working to improve the accuracy of the label placements.
However if you take in any variations reading the meniscus curve and i'm personally not great at that you could have up to 15% variation on your sample volume .... is this significant enough to be an issue ... maybe ....

My salt test tube however resulted in a 10% variation in sample size from expected and this is enough to provide a difference in the test result of reading low, this in turn could have meant I added more salt than the SWCG requires. fortunately I am sneaking up on the final value.

TLDR: If your chasing consistent same size for testing results from your test kits it's worth checking the sample volume indicating by the markings on the tube by weighing the water volume or weighing the sample size every time.

Notice i said consistent sample size, this being only one factoring many such as sample temp and drop size that lead to accurate, consistent results.

I'm sure that many will just think it's being a little OCD and it may well be but maybe the info is useful to the users here of the CCL kits.

Cheers

 

[kimkats]

VERY useful info and timely with your season starting up there! Glad to hear CCL is still a stand up company and are taking measures to make sure the labels are being placed correctly!

 

[Brad_C]

I noticed errors with the CCL label positions also. I use a volumetric pipette to fill the sample bottles. Firstly it gives a consistent sample size, but it also pulls the water from about 14" below the surface without me getting wet.

I generally only use a 5ml sample for chlorine (so 1ppm per drop resolution), and that makes it easy too.

 

[UND312DOG]

I noticed errors with the CCL label positions also. I use a volumetric pipette to fill the sample bottles. Firstly it gives a consistent sample size, but it also pulls the water from about 14" below the surface without me getting wet.

I generally only use a 5ml sample for chlorine (so 1ppm per drop resolution), and that makes it easy too.

I also have the clear labs test kit. So if I reduce the sample size to half (5ml) do I need to decrease the amount of detection powder?

Also by halving the sample size, the titrating drops will be 1ppm per drop instead of 0.5ppm per drop, correct?

Can this 5ml sample size be applied to all tests?

Shane.

 

[Brad_C]

if I reduce the sample size to half (5ml) do I need to decrease the amount of detection powder?

You don't have to, but I use roughly one level spoonful. I've experimented with a few variants, and frankly anything that turns the sample pink seems to be enough.

titrating drops will be 1ppm per drop instead of 0.5ppm per drop, correct?

Correct.

Can this 5ml sample size be applied to all tests?

Probably, but you'd need to scale all the reagents accordingly. I only do it on the Chlorine test as that's the one I use the most.

I probably only do CH & TA maybe 4 times a year. I use a meter for salt and pH, so the pH never gets used and I do the salt drop test maybe every 6 months at most.

The CYA test uses a 5ml sample in any case, and the resulting 10ml is just enough to fill the viewing tube, so you won't be reducing that one.

 

[BrettClearChoiceLabs]

Hello there!
Yes, labels are applied by hand. Oh my god such a long story but it basically boils down to the labelling machine I wanted works so well because it "sees" the edges of the label and so knows how far to advance the bottle and the rollers. But, this doesn't work when you have clear labels. So, back to the drawing board. Hopefully a new sensor I've ordered will actually be able to read clear labels and apply them properly. Otherwise I'll be building my own labelling machine from scratch which would be HILARIOUSLY PAINFUL but would finally justify to myself the 3d printer I bought last year.

I'd love to blame the label discrepancy on a few too many apple ciders but I'm always sober when working in the lab which does not help explain things ?? I love that Phil has done his own measuring, and totally correct all water sample sizes are calibrated for 10ml. Feel free to sharpie on your own line if mine are too far out for your liking.

Oh I totally forgot to reply to you Phil about your 3d printouts. I've been telling all my friends about them, they are amazing. *coughcoughwonder if the forum would like to see them coughcough not subtle coughing*

 

[AusPhil]

I also have the clear labs test kit. So if I reduce the sample size to half (5ml) do I need to decrease the amount of detection powder?

Also by halving the sample size, the titrating drops will be 1ppm per drop instead of 0.5ppm per drop, correct?

Can this 5ml sample size be applied to all tests?
Shane.

Hi Shane,

Whilst i can see that using 5ml for FC saves on some reagents especially if you doing daily checks the other tests are not run anywhere near as often and at least for me the TA and CH test need at least the 10ml testing resolution.

cheers

 

[AusPhil]

Oh I totally forgot to reply to you Phil about your 3d printouts. I've been telling all my friends about them, they are amazing. *coughcoughwonder if the forum would like to see them coughcough not subtle coughing*

No worries Brett .... maybe it is really to many Ciders

anyway you have prompted me to take photos on the new design for holding all the CCL testing parts I decided to do after getting the salt test.
I broke it into separates for each test that can be screwed down to a base if desired.

Personally i keep the sample bottle and the FC test sitting on the desk the rest are on the shelf

Test tube holder
3D printed Clear Choice Labs test kit holders by Phil Gartner, on Flickr

All six tests in five holders, i combined the salt and ph tests into the one holder including a stand for the electronic ph tester
3D printed Clear Choice Labs test kit holders by Phil Gartner, on Flickr

Spoon holder incorporated
3D printed Clear Choice Labs test kit holders by Phil Gartner, on Flickr

Sample bottle holder, sort of helps ensure i don't knock it over
3D printed Clear Choice Labs test kit holders by Phil Gartner, on Flickr

I'm sure i've posted this before but this is the cheap ebay stirrer (speed stir) with printed test tube holder to centre things.
_Z086892 by Phil Gartner, on Flickr

 

[imagineero]

Nice work on the 3D printing - I was looking through the garage at some toolbox foam after my own CCL test kit arrived today. I also noticed the discrepancy with the labels after using a syringe type dispenser for the water. I don't know enough about pools or chemistry to know how sensitive the results are to discrepancy, but I know how sensitive my own OCD is to such matters. After having a cup of coffee to carefully consider the potential effects on the universe of the water sample size variation between having the meniscus (wish I hadn't learnt that word) just touch the bottom of the line, the middle, or the top of the line, I had to measure the line height and thickness.

I came to the conclusion that for my own purposes it makes little difference so long as it's consistent. It took a second cup of coffee for me to arrive at this conclusion, at which point, for my own sanity and my eyesight I decided to just use a volumetric dispenser. I hate myself.

 

[AUSpool]

Geday AusPhil,

I think your discrepancy is dew to the specific gravity assumption. You cant accurately weigh a volumetric sample without knowing the conversion factor for the specific gravity. In order to find the specific gravity you need to know what’s in the water and thats what the tests are for and even then we only test for some of what is in the water, not everything. The specific gravity of pure water at 30degC is ~0.996 but then there is all the other stuff in it making it heavier. I would assume the s.g. of the average pool water sample would be closer to 1.1 meaning that somewhere between 10.5 - 11g of water would be equal to a 10ml sample but you’d be better just using a volumetric measure. A 10ml volumetric pipette or filling the CCL vial to the line and be done with it.

Try going the other way and weigh a 10ml sample from a volumetric pipette and remember a little will remain in the pipette so the weight will be slightly under.

Love the 3D printed stands and good work on the home made speedstir.

 

[AusPhil]

Geday AusPhil,
I would assume the s.g. of the average pool water sample would be closer to 1.1 meaning that somewhere between 10.5 - 11g of water would be equal to a 10ml sample but you’d be better just using a volumetric measure. A 10ml volumetric pipette or filling the CCL vial to the line and be done with it.

Thanks for the kind words on the 3D printed stuff but i do think the SG "assumption" needs to be dealt with........ Considering that the published average for Seawater is itself 1.025 ie 1025g per liter and that the average salt water pool (yes here's an assumption) should be somewhere between that and 1.000 (0.996) then 10gm of pool water should be as near as 10ml..... respectfully it's going to be closer then meniscus viewing for a lot of people.

I just filled a CCL test tube to the very top with distilled water and got a weight of 37.7gms ...... filling the same tube to the top with a pool sample without looking at the weight during filling i got 37.6gm to the same point ... repeated the same and got the same.

Happy to conclude that assuming a SG of 1.000 and weighing the samples is still better than assuming the label is in the right place ........

My real reason to weigh the samples is i can leave the magnetic stirrers in the test tubes and just tare the weight before filling..... keeps it quite simple really

Cheers

 

[AUSpool]

I’ve always used a s.g. of 1.0264 for natural sea water but thats close enough to a s.g. of 1.025 which is 1025kg/m3. Since pure water is 1000kg/m3 @ 4degC with a s.g, of 1 @ 4degC, pool water must have a s.g. somewhere between 1 and 1.025 and be heavier then pure water. I believe the s.g. for pool water at 3300ppm would be ~1.002. 37.7gms for distilled and 37.6gms for pool water gives a sg of 0.997 but indicates that something is not right since the pool water should be heavier.

I recorded the same error, or discrepancy, with 10ml of tap water and 10ml of pool water. I didn’t have any distilled on hand. I got 9.84g for the tap water and 9.79g for pool water. I suspect we’re seeing a temperature compensation interference. I think the samples need to weighed at 4degC to be the most accurate.

I know we’re really close on these numbers but my point is that when weighing the sample your just swapping one interference factor for another. For the convenience of leaving the stirrer in and reading to the line you’d be better off using a 10mm volumetric pipette which you can get off ebay for $15 - $20. A volumetric pipette would do a similar job as the Taylor SampleSizer.

 

[AUSpool]

I’ve had my two samples in the beer fridge overnight and weighed them out this morning to get 9.70g for the tap water and 9.81g for the pool water. At least the pool water is heavier this time. The s.g. for the pool water sample is 1.011 @ 4degC and given the s.g. from the original results was 0.995 @ ~25degC the margin of error is just under 2%. My original estimate for the pool water s.g. was a long way off the mark. I now believe pool water s.g. is somewhere between 1.002 and 1.011.

While weighing volumetric samples is unusual approach it is not going to make any significant difference to the outcome.

I think too that given the tolerance of the tests the concern with labels and sample size is not going to have any great impact on the end result for most users. The salt test has the broadest range of variance at +/- 200ppm and in most cases, regardless of your test result, your not going to add any salt unless your SWG displays a low level indication.

 

[AusPhil]

I’ve had my two samples in the beer fridge overnight and weighed them out this morning to get 9.70g for the tap water and 9.81g for the pool water. At least the pool water is heavier this time. The s.g. for the pool water sample is 1.011 @ 4degC and given the s.g. from the original results was 0.995 @ ~25degC the margin of error is just under 2%. My original estimate for the pool water s.g. was a long way off the mark. I now believe pool water s.g. is somewhere between 1.002 and 1.011.

While weighing volumetric samples is unusual approach it is not going to make any significant difference to the outcome.

I think too that given the tolerance of the tests the concern with labels and sample size is not going to have any great impact on the end result for most users. The salt test has the broadest range of variance at +/- 200ppm and in most cases, regardless of your test result, your not going to add any salt unless your SWG displays a low level indication.

Hiya Steve,

This is actually very interesting to get educated on..... a bit more reading tells me that water is considered at it's most dense at 4c and is normally considered to be 1.0000 at 20c/20c but taking from the Wikipedia article we get....
Taking into account different sample and reference temperatures, we note that, while SGH2O = 1.000000 (20 °C/20 °C), it is also the case that SGH2O = 0.998203⁄0.999840 = 0.998363 (20 °C/4 °C).
Given that it would seem even your volumetric sample is off as it should weigh 9.99 gms at 4c.

Given that the margin of error quoted by Taylor (and by inference CCL) for just about all tests is well above any sample size error from volumetric or weighted sample sizing then you and i are well into theoretical differences.
However a 10% variance caused by incorrect label placement .... the issue that started this thread ..... is significant enough to want to get consistent sample sizes regardless of the label, meniscus read errors and even parallax errors.

I have a beer brewing hydrometer (a decent one) buried in a toolbox somewhere that i meant to find over the weekend, given it is meant to measure the subtle variations of SG caused by alcohol it should give a good indication on pool water .....

 

[BrettClearChoiceLabs]

I'll "weigh in" on this educational debate by saying my main skillset is puns, so that's my only contribution I'm afraid

More seriously though, I'm fascinated by this discussion because the whole aim of the Labs and any test kit really should be to get the most accurate results easily and quickly. Any time I do reading/research like you guys are doing I've always got the business side of things in the back of my head which obviously adds a different slant again. So I love it when people get pure science on pool testing.

 

[AUSpool]

Thanks for ‘weighing in’ on this Brett, I guess you’ve attempted to source a cost effective, relatively accurate pre graduated or printed cylinder.

Just for the fun of it (actually I’m procrastinating and should re-stumping) I used my little 10ml pippette and filled the 25ml measuring cylinder, an old glass Sera cylinder and a Taylor 9198 and all came right to the 10ml mark. An interesting observation though, I cant really see a meniscus in all three cylinders, just a water line. The volumetric pipette is narrower where I can see a meniscus.

“Given that it would seem even your volumetric sample is off as it should weigh 9.99 gms at 4c.” Given it was tap water it should have probably been a little closer to 10.0gms but some fluid will always be left in the pipette. Given an expected of 9.99gms and an actual of 9.70gms my margin of error is close to 3%, slightly worse than weighing...

I think its a bit lame that 0.999840gms @ 4degC is used for the max density of pure water. Thats extrapolated from actual max density of pure water which is 1.00gms @ 3.98degC and 1 atmosphere, but some twit has said ‘let’s just call it 0.99984 @ 4degC, like 1 @ 3.98degC would be too easy right. There must be a bigger picture I cant see. And we probably should add the equation; SG = density(sample) / density(pure H2O @ 4degC).

I had good success with a hydrometer in an aquarium, it was quick, accurate enough and had a much broader scale than the other more expensive toys I played with. But an aquarium (salt) is inherently different from a pool. I persisted with a Watermaid pool hydrometer a while ago I don’t know why I’ve still kept it. Its calibrated at 20degC and the instructions give a correction factor of +/- 400ppm per 1degC. Ouch, that means that with a reading of 6000ppm @ 30degC I could be anywhere between 10,000ppm and 2,000ppm.

The Watermaid hydrometer has a SG/ppm salt scale where a SG of 1.000 = ~800ppm salt, 1.002 = ~3200ppm and 1.004 = 6000ppm. The SG of 1.002 = 3200ppm salt is kind of in the ball park.

All this kind of makes me sound like I’m a bit retentive about accuracy and I have had my moments, I’ve probably been wrong more that I’ve been right, but nowadays I use the Taylor cylinder poolside, rinse it three times, plung it down to my elbow, flick out some water till I’m kind of close and wack it on the speedstir.

 

[AusPhil]

Thanks for ‘weighing in’ on this Brett, I guess you’ve attempted to source a cost effective, relatively accurate pre graduated or printed cylinder.

All this kind of makes me sound like I’m a bit retentive about accuracy and I have had my moments, I’ve probably been wrong more that I’ve been right, but nowadays I use the Taylor cylinder poolside, rinse it three times, plung it down to my elbow, flick out some water till I’m kind of close and wack it on the speedstir.

Brett "weighing in" should only be taken seriously if he is balancing the scales!

I thought I was sounding retentive but honestly this imo has been a great discussion/thread ......

Me i just drop each test tube onto the 200gm scales, squirt the water in till i typically get from 10.0 to 10.2 grams and call it close enough and then do the next tube ..... i've got better at getting closer to 10

 

[Costas > DownUnder]

I suppose it raises the question as to 'How accurate' do we need to be when measuring our pool water....

A factor of 10 or better than test strips and i'll be happy..

 

[AusPhil]

I suppose it raises the question as to 'How accurate' do we need to be when measuring our pool water....

A factor of 10 or better than test strips and i'll be happy..

Considering the Taylor technologies heavily market their own test strips and in proper use they advise that the accuracy is half the distance between two colour standards they are likely good enough most of the time if used as per instructions.
In reading some Taylor literature on strips it seems the greater issue is that they are nearly always used incorrectly and different brands have differing requirements.
At least the drop tests are easy to follow.
One thing i just seen on Taylor's website is
False-high reading due to cyanuric acid
Cyanuric acid (CYA) titrates as total alkalinity when CYA is greater than 30 ppm. Adjust for this by subtracting 1/3 of your CYA reading from your total alkalinity reading to get the correct alkalinity value.

To me the question even in this thread is not how accurate but how consistent we can be .... counting more than 20 drops really sucks and i'd prefer to know how many drops to a ml and maybe use a 1ml syringe to measure the reagent ... Taylor spec 25 drops to the ml .... wonder if CCL is the same.

It's like digital PH meters get pretty badly beat up on this site but i'll take a digital meter reading over a phenol red drop test any day as I personal suck reading the colour values (shades). I calibrate my cheapy every few weeks and it's never been out by more then 0.1 .... i'd call that more than good enough

 

[AUSpool]

CYA doesn’t interfere to cause a high reading on the TA test, it is part of the total alkalinity and should be accounted for. CYA, borate and carbonate are the major species that interact to stabilise the pH and are all part of the total alkalinity. If present, all three are accounted for as part of the total alkalinity test. Removing the CYA portion is an ‘old school’ pool industry thing, TFP does not remove the CYA portion, that would lead you to over dose with baking Soda and artificially raise your TA too high. Removing the CYA borate portion leaves you with carbonate alkalinity and is done automatically to calculate CSI.

Drop sizes are a function of the tip on the reagent bottle. Taylor and CCL use different bottles and have different drop sizes, CCL has compensated for this by altering the reagent concentrations.

It occurred to me that I wasn’t being much help weighing out 10mls @ 4degC. No one’s going to bother chilling down a water sample. So before I packed away my glass wear I weighed a 10ml pool sample from the volumetric pipette (Ex20degC) at 27degC and got 9.7gms. Next I weight 10gms of pool water @27degC in a 25ml graduated cylinder (Ex20degC +/-0.5ml) and it was right on 10mls. 11mls was 10.59gms. Given that I reckon that if you weigh out anywhere between 9.7gms and 10gms you’d be good to go.