Two CYA kits consistently disagree.

[lalittle]

I have two different CYA test kits -- one is the Taylor test that comes with the K-2006 (the small tube), while the other is a Blue Devil kit -- the larger tube with the "slider" that is sold by various companies, including Taylor. Both kits are recommended by various people.

I'm finding that the two kits are constantly giving me what I would call substantially different results. The Taylor kit gives me around 90 ppm CYA, while the "slider" kit shows much lower levels -- more like around 60 or so. I'm being VERY careful to follow all instructions, and to accurately use the correct amounts of water and reagent. I get consistent results on each kit -- i.e. successive tests show me the same results -- but the kits disagree by about 30ppm.

What could be causing this? Which kit should I trust?

Note that I ordered the Taylor K-1720 kit, which seems to be a larger and easier to read version of the CYA kit in the K-2006. I'm hoping that this will be a tie-breaker, but I won't have it for a little while still, so I won't be able to post the results for a bit.

Thanks for any feedback on this,

Larry

 

[Bama Rambler]

Re: Two CYA kits consistantly disagree.

Are you doing the tests with your back to the sun?

 

[stev32k]

Re: Two CYA kits consistantly disagree.

I would suggest you buy a standard CYA solution and test both kits against the standard. TFtestkits sells a 50 ppm standard solution of CYA here: http://tftestkits.net/splash-page.html. Run each test kit using the standard and see what the results are. Then you can do additional testing to see if the problem is the reagents or the tubes.

P.S. I would be very interested to know what your results are - keep us informed.

 

[lalittle]

Bama - I've tried looking at the results a few different ways (i.e. in the sun, in a shadow, etc.), but I find that this does not change my results.

stev32k -- I had no idea that standard solutions were available for this. This should give me exactly the information I'm looking for. I'll give it a try and report back.

Thanks,

Larry

 

[duraleigh]

Re: Two CYA kits consistantly disagree.

If you will PM your name and address to me, I will get you some of the sample solution so you can report the results.

The test is not terribly accurate but that's a little farther off than I think it should be.....it'll be interesting. :lol:

 

[Patrick_B]

Was hoping to see an update on this. Curiosity is up.

 

[CaOCl2]

What could be causing this?

Kit and reagent quality, origin, labware accuracy, different and/or errors in the testing protocol.

Which kit should I trust?

Taylor. They have traceable reagents, stuff that's all made in the US, quality management that meets ISO 9001, and you can get a Certificate of Analysis for that particular reagent and that particular lot number. If there was a problem they would know immediately.

Note that I ordered the Taylor K-1720 kit, which seems to be a larger and easier to read version of the CYA kit in the K-2006. I'm hoping that this will be a tie-breaker, but I won't have it for a little while still, so I won't be able to post the results for a bit.

You'll probably find, just as I did, that both Taylor tests give the same results.

 

[lalittle]

Okay -- I got the standard solution from Dave (thanks Dave!) and finally had a chance to do a few tests. Sorry it took me so long to do a write up.

My overall reaction is that I find these "black dot" CYA tests to simply be VERY inaccurate. Regardless of which test I used (Taylor or Blue Devil), I found that there was simply no obvious point at which the black dot became obscured by the "correct amount." Using the standard solution I was able to see exactly what the dot "should" look like at 50ppm. The problem was that this was just not a recognizable point that I could reliably repeat from day to day. At the "correct" point I could still barely see the dot, so it was not "totally" gone, but it was very difficult to quantify this "just visible" point. I was eventually able to "tentatively" hit the correct result after several subsequent tries with the same sample, but only when doing one test right after another. If I waited a little while and then did a new test again, I found myself unable to hit the correct point, and my margin of error was significant. The transition from "barely visible" to what I would call "not visible" was just too gradual to get consistent results, and having to hit an arbitrary "not yet fully obscured" point was just really hard.

In other words, it was really hard to remember the correct amount of "black dot visibility." It's interesting to note that I seemed to be able to get more consistent (albeit incorrect) results if I went all the way to FULLY obscuring the black dot to MY eye. I thought that I might be able to use this result and apply some form of "correction" to get the real results. Unfortunately, the more I tested, the more I realized that my results were not as consistent as I had originally thought. This was revealed when using the standard solution, where I knew the correct point ahead of time.

Note that I tried the tests in various light, but could not find any situation that would give me consistent results.

After playing with the standard solution, it would not surprise me at all if some people were over 20ppm off in their tests. Also, since the graduations get much closer together at higher CYA levels, it strikes me that higher CYA levels can potentially lead to much lower accuracy, where a small difference in fluid level gives a large error.

With regard to the Taylor and Blue Devil tests giving different results, I think that it's the nature of the two different testing methods. The Taylor tests have you add drops, so you need to stop at the correct point. The BD test (and others like it sold under numerous brand names) allows you to "bounce" up and down in smaller and smaller increments until you find the correct point. The different methods simply color the results a bit. In the end, I find the BD test easier to use since the "bouncing" allows you to zero in on the correct point a bit easier, but the Taylor test has much finer graduations, which is also an advantage. The larger Taylor kit is easier to use than the smaller one, so it most likely offers slightly more consistent results, but this does not change the fact that the test itself is highly inaccurate.

I should add that I believe that over time, one might be able to "learn" to get more consistent results by getting a more accurate feel for the test, but I don't know how long this would take, and it would certainly take some effort.

All in all, this was an interesting experience, and I once again want to thank everybody for their help here (and another big thanks to Dave for the solution.)

Larry

 

[mynewpool]

I feel the CYA test is the hardest one to get correct. It seems to be all about perception and how you feel when you think the black dot is no longer visible.

 

[CaOCl2]

Consistent lighting makes for consistent results on this test.

cya-testing-t34850.html

 

[Beamup]

The fundamental problem with the CYA test is that you have two degrees of freedom (attenuation length and initial light level), but one measurement. That makes it impossible to disentangle the two. If you can get a consistent initial light level, and calibrate that effectively, then you can accurately measure the attenuation length (which then gives you CYA) - but that's not really possible.

The better approach would be to have TWO dots, of differing brightness, and do the measurement with both. Then you could use the ratio of the amount of solution required to obscure each, cancel out the effect of the initial light level, and have a much more robust measurement. It would be a more complex process and would probably end up needing a table to look up the two measurements and get the CYA, but it ought to be much more accurate.

The subjectivity of when the dot disappears would ALSO cancel out, as long as both dots were judged the same way.

I initially was going to show the math, but since this isn't the Deep End I'll skip that.

 

[lalittle]

Consistent lighting makes for consistent results on this test.

cya-testing-t34850.html

I can only speak for myself, but I found that my results were far from consistent even when using the exact same lighting for subsequent tests (i.e. no change to light or body position.) I prefer to be in sunlight with the cylinder in the shadow of my body, but I even tried other lighting situations, and got similarly inconsistent results.

As mentioned above, by using the standard solution I discovered that I could still just barely see the black dot when the fluid level was correct. The difficult part was quantifying this point so that I could repeat it, which I found very difficult to do without a large margin of error. I also found that my results changed from one day to the next, as if my "opinion" about the appearance of the black dot had changed even though it was a totally clear day at the same general time.

As others on this forum have noted, I find this test to be the most difficult test to perform accurately, and I have no confidence in being anywhere close to accurate. My only hope is that I'll get better over time, or that somebody will come up with a better way of testing for CYA.

Larry

 

[stev32k]

For me, what works best is to fill the tube without stopping until I believe the dot is not visible. If I slow down or pause when putting the mixture in the tube my perception of the end point changes. It may have to do with the particles settling to the bottom and the upper portions of the liquid clearing up, but if I slow down or pause my reading are always very inaccurate compared to the 50 ppm standard. If I fill the tube at a fairly constant rate and stop the instant I think the dot is gone my readings are about +/- 5 ppm from the standard. If I fill the tube slowly - just a few drops at a time - then pause - then a few more drops - the readings are all over the map.

The reason I bought the standard was to calibrate my eyeballs and it seems to have helped. I ran about 15 or 20 tests by mixing the reagent take a reading - then pour the solution back in the mixing bottle and repeat. I sometimes do that with a pool sample just to check the repeatability and it's usually pretty good. At least now I feel fairly confidant that my results are good enough to base chemical additions on them.

 

[chem geek]

With regard to the Taylor and Blue Devil tests giving different results, I think that it's the nature of the two different testing methods. The Taylor tests have you add drops, so you need to stop at the correct point. The BD test (and others like it sold under numerous brand names) allows you to "bounce" up and down in smaller and smaller increments until you find the correct point. The different methods simply color the results a bit. In the end, I find the BD test easier to use since the "bouncing" allows you to zero in on the correct point a bit easier, but the Taylor test has much finer graduations, which is also an advantage. The larger Taylor kit is easier to use than the smaller one, so it most likely offers slightly more consistent results, but this does not change the fact that the test itself is highly inaccurate.

With the Taylor test, you can always pour the solution back into the bottle to repeat the test. I find that lets me zero in on the "just disappearing" point better. As for more error at higher CYA, it doesn't work that way because the turbidity increases logarithmically so the dot gets obscured more rapidly (i.e. with fewer drops) at the higher CYA level. So the error in the test is still on the order of 10-15 ppm if one is able to be consistent relative to the standard. If you look at some Taylor photos of the test, the idea isn't to completely have the dot disappear with absolute certainty. It's instead to get to a point where you really can't easily tell that a dot is there. There will be a point at which you can clearly distinguish a dot and another point where you'll wonder to yourself if the dot is there and think you see it faintly, but that doesn't really count as seeing the dot.

I prefer to do the test outside with my back to the sun shading the tube and looking down straight into it -- that's the approach Taylor recommended. It's always measured for me pretty much as predicted, within around 10 ppm or even better than that, but I usually have low CYA levels at < 20 ppm at the start of the season raised to around 40 ppm that then slowly drops (over the winter I dilute the water with winter rains which accounts for at least some of the drop). For me, a 10 ppm difference in the tube results in a fairly clear distinction between seeing a dot vs. wondering if that's really a dot I see. There is no question that if I keep going I can get to a point of more certainty of there being no dot, but that is really going too far.

Thank you for doing the tests and using the standard solution. There's no question it's the most difficult and least accurate test. It's unfortunate there isn't a "count the drops until the color disappears or stops changing" sort of test with a better transition point, but there isn't.

 

[duraleigh]

As usual, chem geek's precise post is written far better than I can do but my thoughts are identical.

Over the years, I feel I have learned to get to within 10ppm or just a little less and my results seem to bear that out.

While it certainly remains the hardest test to master, it also is the "ingredient" in our pools that has enough leeway to make the test meaningful and precise enough for what we need to know.

 

[Jersey Devil]

I had all the same concerns when I switched to the BBB method. I was so concerned about the apparent subjectivity of the test that I was ready to go out and buy an expensive laboratory instrument to measure turbidity. Then, as I became more experienced I found my results to become more consistent, until now I can arrive at the same reading week after week with virtually no variance (Dave's standard solution helped me to get there.) In actuality, I don't care if my estimate (40 ppm) is accurate or not. What I care about is consistent results and that my current CYA provides for adequate FC stability in pool ( as checked by daily readings and occasional am/pm tests to see how much FC the sun is burning off). For me, 2 ppm FC loss on a sunny day is acceptable (with no bather load). As long as I stay above the minimum effective FC level advised by Pool School I can maintain a crystal clear pool throughout the summer. (Oh, and to be safe, I try to stay at the higher end of the recommended FC range in case my CYA reading is a bit lower than actual - 1 ppm FC accounts for 10 ppm higher CYA, so it is pretty painless to do so).

 

[frustratedpoolmom]

As long as I stay above the minimum effective FC level advised by Pool School I can maintain a crystal clear pool throughout the summer. (Oh, and to be safe, I try to stay at the higher end of the recommended FC range in case my CYA reading is a bit lower than actual - 1 ppm FC accounts for 10 ppm higher CYA, so it is pretty painless to do so).

I agree wholeheartedly - thats the beauty of our methods - and I'm constantly telling my friends/clients - just keep the FC at the appropriate level and you will be Trouble-Free. I think the recommend levels give us breathing room - when it comes to the CYA test even if we aren't 100% precise that breathing room protects our pool water. I'm always telling folks - yes it's ok to swim if your FC is at a 5. You're not gonna melt. LOL The Pool Store has folks so misled and confused/overwhelmed when it's really pretty simple.

 

[anonapersona]

OK, looks like I need to get a standard solution. I have tried to tell myself that the test is done when "you can't tell if the dot is an oval or a circle" in that it is there but not with much certainty. I find that it is too easy to end up focusing on the sticker itself, which is much larger than the black dot, when in sunlight that reflects from the ground.

 

[rme]

So I just received my TF-100 kit in the mail today and feel much better about the CYA test after reading this post. I figured I just needed practice. My pool was filled with well water and I used BBB method from the start so when the 1st results read over 100 I knew something was wrong since I've only added about 3 lbs of stabilizer so far. Second test read between 20-30 which seems more likely.