Accuracy strips vs. drops (TDS & Salt)

[polyvue]

In recent months I've had ever-increasing levels of TDS and Salt. That is, according to the Aqua Chek Salt strips and TDS strips I've been using. My doubts prompted me to repurchase both in August, thinking that the strips may have been old (although I'm fairly scrupulous in my care of them.) When I conducted testing with the new strips I got... exactly the same results.

TDS > 5000 and Salt ranging from 3660 to 3970. My PS-8 Aqua Logic controller reports salt consistently at 2900-3000 ppm. I believe that the salt cell measures on the basis of conductivity and, per several comments I've seen, reports salt ppm based on its finding minus some variable to account for other dissolved solids. In an effort to resolve the apparent discrepancy between the strips and the SWG, I recently acquired drop tests for TDS and sodium chloride, repeating the tests with the Taylor kits and the newer Aqua Chek test strips. For a comparative analysis, please see the attached Excel spreadsheet.

My tentative conclusions:

• Observation 1: Salt cell reports ~2900, sodium chloride drop test reports 3000, Aqua Chek strips 3970
  Observation 2: TDS drop test reports 2400 as CaCO₃ but Aqua Chek shows TDS off the chart: 5000+
  Hypothesis: Aqua Chek salt strips overreport sodium chloride; the Pro II TDS strips may include salt in total

I'd love to hear what others have to say about these findings. Thanks!

 

[CaOCl2]

The Taylor Salt (Mohr) titration test is a bit tricky to perform, the endpoint is when the sample turns a pale "salmon" color, not "brick" color. That's why I prefer to use a 20 mL sample and 2 drops of R-0630. Each drop of R-0718 represents 100 ppm NaCl, for me the proper endpoint is much easier to see.

Possible that the salt cells measures conductivity and adjusts for TDS, that how my Myron-L meter works (it reports as ppm TDS and ppm NaCl). However, when testing a Standardized Solution (Taylor R-0968 aka 442) , my Myron-L meter and Taylor drop test agree.

It may be something to do with the units involved (ppm CaCO3 vs ppm TDS).

 

[Guest]

I never trust the strips, as a rule of thumb. Just seems like too many factors affect them and results are random.

Could it be that the drop test is measuring just calcium hardness and the Aqua Chek is showing all hardness minerals?

I've experienced difficulty in getting accurate salt readings as well. My IC40 (last salt system) kept giving an "add salt" message (and I foolishly did without checking it ). I finally took a test sample to my local pool place (really good folks) and was up to 6,000 ppm! I am now without a SWCG (I'm looking at the SGS one currently) since I got tired of the whole system (the IC was the second one I had installed, and I had a Polaris Dichlor G-1000 feeder with ORP prior to those) and am adding bleach only. Don't really know what route I'll ultimately end up with!

Bruce

 

[polyvue]

The Taylor Salt (Mohr) titration test is a bit tricky to perform, the endpoint is when the sample turns a pale "salmon" color, not "brick" color. That's why I prefer to use a 20 mL sample and 2 drops of R-0630. Each drop of R-0718 represents 100 ppm NaCl, for me the proper endpoint is much easier to see.

Yes, I also found that doubling the sample makes it easier to read the endpoint.

It may be something to do with the units involved (ppm CaCO3 vs ppm TDS).

I'm not entirely clear on this. Is this an apples-to-oranges comparison? The Taylor kit I purchased (K-1764) does report TDS as calcium carbonate. Have no idea what the strips represent.

Could it be that the drop test is measuring just calcium hardness and the Aqua Chek is showing all hardness minerals?

Bruce - Maybe so, but if you mean that the TDS drop test result should equal Calcium Hardness (CH), I think not. CH=275, TA=75, TDS=2400 per the drop test. So there must be other constituents represented in the TDS figure. Your second point, however, fits with my current thinking: That the Aqua Chek strip test (which includes not only TDS but CYA and Total Hardness-- both of the latter results seem proximate to that of other tests) for Total Dissolved Solids (TDS) reports sodium chloride along with other dissolved solids. The drop tests, taken together, seem to reinforce that (sodium chloride 3000 + TDS 2400 = Aqua Chek total for TDS 5000+) Of course, this convenient discovery could be merely a coincidence.

I'm still searching for a better understanding of the TDS test... but am becoming increasingly convinced that the Aqua Chek strip test for sodium chloride is overstating the salt content for my pool (due possibly to interference from high TDS.)

Thanks to both of you for your replies.

 

[duraleigh]

I can't be a good resource because I don't have a salt pool, but it has been my understanding that the Aquachek salt strips were the most accurate test of salt on the market. Earlier this summer, I seem to remember several discrepancies and the strips came out on the money.

Now, all we have to do is find a test that is known to be accurate and make the comparison.....problem is, I don't know of one that is drop dead correct.

 

[polyvue]

t has been my understanding that the Aquachek salt strips were the most accurate test of salt on the market. Earlier this summer, I seem to remember several discrepancies and the strips came out on the money.

Yes, I've read posts authored by a number of forum members stating without ambivalence that the Aqua Chek salt test is the most accurate. Conversely, I've read posts from others here that take the opposite stance. I'm not trying to bash Aqua Chek. For all I know, NASA may insist on a supply of Aqua Chek brand Salt Strips on each shuttle mission, for the astronauts' use in calibrating their $25,000 per unit on-board salt/conductivity meters!

[Don't you just know that someone is going to copy the previous sentence? An urban legend, in the making! :lol: ]

Anyway, back on earth... My theory is that high-TDS (or one or more components of TDS) is escalating the Aqua Check strip Salt Test result. What explains a 1000 ppm difference between the drop test (and concurring SWG) and the strip? Can anyone say authoritatively (meaning, with some chemical explanation or reasoning) that the Aqua Chek strip is a reliable indicator for sodium chloride (ONLY) and not for say, an aggregate of chlorides?

I guess I have the same quibble with Aqua Check (Pro II) TDS strips. Here, too, it may be just be a lack of understanding on my part. But I think I have the TDS thing figured out (if the strip test counts both salt and other dissolved solids in the TDS result.)

Saying that the Salt test strips are the most accurate is also suggesting that a.) the Taylor sodium chloride (drop) test is NOT accurate -or- that the test was done incorrectly -or- the precision of the test is not as advertised; and b.) the SWG is also in error, from 600-1000ppm LESS than the strip test.

I've a feeling that however this ends up, half the people on this board are going to be mad at me.

 

[duraleigh]

Well, I won't be in the half that's mad, for sure.

Waterbear was the testing guru and he thought the Taylor test was either difficult to perform accurately or unreliable.....I think it was difficult to perform accurately. If I remember correctly, he liked the aqua-chek strips because they were not only repeatable but fairly precise, I think.

Obviously, I simply don't have a good handle on which ones work. I think I remember Jason having some thoughts on the salt tests, as well......let's see if he chimes in.

I'm gonna' get out of the discussion because I have never used one in my life so I am an obvious poor choice for authority.

 

[chem geek]

Well, I have both the Taylor K-1766 salt test which technically is measuring chloride and reporting in units of ppm sodium chloride and I also have the AquaChek Salt test strips.

I also use a larger sample size for the Taylor test -- usually 25 ml where each drop is 80 ppm -- and this results in a slower transition end-point, but seems easier to get a more accurate number. I think the key to this test is good mixing, similar to the CH test, to ensure thorough development as you go (though I don't wait 30 seconds between drops). By the way, the tube with my test kit has a line at 20.5 ml, not 20 ml, though that's pretty close (each drop represents 97.6 ppm instead of 100 ppm at 20 ml, 80 ppm at 25 ml, 200 ppm at 10 ml).

I've found that sometimes the test strips are consistent with the Taylor test and sometimes they are not. I have no idea why, but it seems to me that a repeated consistent Taylor test is the more accurate one. My few comparative tests seemed to show more consistency between the Taylor drop test and the AquaChek test strips at lower salt levels (around 1000-1300) vs. higher salt levels in a neighbor's SWG pool (3000-4000 ppm). I don't remember the exact discrepencies, but I seem to recall the test strips were around 800 ppm higher in their salt reading in the SWG pool.

As for the TDS test, since most of TDS is sodium chloride, it is better to use ppm sodium chloride for that measurement. Generally speaking, the TDS is just a little higher than the salt level -- around 200 ppm higher for plaster pools saturated with calcium carbonate. You need to multiply a ppm calcium carbonate reading by 58.443/100.087 = 0.58 to get the equivalent ppm sodium chloride reading. That's unfortunately the opposite direction of what you were seeing with the Taylor TDS test reading 2400 and the Taylor salt test reading 3000 -- that's very inconsistent. With a salt test reading of 3000 I would expect TDS to be around 3200 ppm sodium chloride or 5480 ppm calcium carbonate.

I think the best way to determine the accuracy of the tests would be to take a very carefully measured volume of tap water (low in salt) and add a known amount of salt very carefully measured by weight to it and then do measurements with the two tests. Then, if there were no differences, one could also increase TA and CH and even CYA to see if any of these interfere with either test.

Richard

 

[Pontiac]

I guess I have the same quibble with Aqua Check (Pro II) TDS strips. Here, too, it may be just be a lack of understanding on my part. But I think I have the TDS thing figured out (if the strip test counts both salt and other dissolved solids in the TDS result.)

I have also used the AquaChek Pro ll strips on a freshly filled SWG pool and found the same "off the chart" TDS reading. It makes sence now, thanks for sharing.

 

[polyvue]

[T]he tube with my test kit has a line at 20.5 ml, not 20 ml, though that's pretty close, each drop represents 97.6 ppm instead of 100 ppm.

Good catch. Fortunately, the difference is such that it doesn't greatly affect the test result for the levels of salt we're dealing with: 2.4 X 30 drops = 72 ppm

I have no idea why, but it seems to me that a repeated consistent Taylor test is the more accurate one. My few comparative tests seemed to show more consistency between the Taylor drop test and the AquaChek test strips at lower salt levels (around 1000-1300) vs. higher salt levels in a neighbor's SWG pool (3000-4000 ppm). I don't remember the exact discrepencies, but I seem to recall the test strips were around 800 ppm higher in their salt reading in the SWG pool.

This is consistent with my findings. At higher levels of TDS and/or salt, the Salt test strips are overreporting. The question remains--for me, at least--why?

As for the TDS test, since most of TDS is sodium chloride, it is better to use ppm sodium chloride for that measurement. Generally speaking, the TDS is just a little higher than the salt level -- around 200 ppm higher for plaster pools saturated with calcium carbonate. You need to multiply a ppm calcium carbonate reading by 58.443/100.087 = 0.58 to get the equivalent ppm sodium chloride reading. That's unfortunately the opposite direction of what you were seeing with the Taylor TDS test reading 2400 and the Taylor salt test reading 3000 -- that's very inconsistent. With a salt test reading of 3000 I would expect TDS to be around 3200 ppm sodium chloride or 5480 ppm calcium carbonate

This is a bit disturbing and doesn't seem to support my theory that TDS as reported by the test strips is just a combination of sodium chloride and other components (largely calcium carbonate, as determined by drop test result of 2400 ppm TDS.)

When I tested the pool's make-up water (at the main which feeds the auto-fill), the Taylor drop tests reported Sodium Chloride 200 ppm and TDS 250 ppm. The AquaChek strips reported Sodium Chloride < 400 ppm and TDS 3000 ppm. The Aqua Chek's TDS reading seemed inordinately high... does tap water normally test out at 3000 TDS?

 

[chem geek]

When I tested the pool's make-up water (at the main which feeds the auto-fill), the Taylor drop tests reported Sodium Chloride 200 ppm and TDS 250 ppm. The AquaChek strips reported Sodium Chloride < 400 ppm and TDS 3000 ppm. The Aqua Chek's TDS reading seemed inordinately high... does tap water normally test out at 3000 TDS?

Tap water varies, but it won't have anywhere near 3000 TDS. The reservoir water that is mostly from rainfall runoff in our area (this is one of two sources of our tap water) has 30 mg/L chloride (equivalent to 49 ppm sodium chloride), 57 mg/L Total Hardness (same as ppm calcium carbonate), 54 mg/L Alkalinity (same as ppm calcium carbonate) and 120 mg/L Total Dissolved Solids. If I add all of the measured components in their mg/L units for those components I get the following:

30 mg/L chloride
16 mg/L calcium (from total hardness and assuming around hardness is 70% from calcium)
4 mg/L magnesium (from total hardness and assuming hardness is 30% from magnesium)
33 mg/L bicarbonate (from alkalinity)
5 mg/L sulfate
19 mg/L sodium
--------------
107 mg/L Total Dissolved Solids

So most of the TDS is accounted for by the above.

Richard

 

[polyvue]

Tap water varies, but it won't have anywhere near 3000 TDS [...]

Sacramento water is largely pulled in from water treatment facilities at the Sacramento and American rivers (which are fed by snow melt, springs, reservoir releases, storm run-off and reclamation sources.) I don't have their annual water quality report handy, just my own testing (see Excel attachment in first post) so can't compare notes on TDS constituents.

The Taylor drop test result for Total Dissolved Solids (TDS) for the tap water I measured reported TDS 250 ppm, which is a lot closer to the 100-120 mg/L figure than AquaChek's 3000 ppm... I'd like to get a bit more into specifics but don't want to monopolize this conversation more than I already have. Does anyone else here have a horse in this race?

 

[chem geek]

Sacramento's water quality report is here. TDS average for surface water is 100 ppm and for groundwater is 305 ppm. Since the tap water is composed of 85% river water and 15% groundwater, the tap water TDS should be around 131 ppm.

The Taylor K-1764 TDS test appears to measure mostly bicarbonate in the first part of the test (Total Alkalinity), but I am unsure what the second part of the test is measuring though it appears it might be Total Hardness (sum of calcium and magnesium). It is not clear whether the test will properly measure sodium chloride salt. Certainly the test isn't going to give you the proper units for TDS and a single conversion isn't going to be accurate if you don't at least roughly know the constituents (the same is true for a conductivity test). You might take some distilled water, add a known amount of salt, and see what the different tests measure.

 

[JasonLion]

From some time ago, the consensus was that the Taylor drop based salt test was the most reliable, but only if you know how to do it correctly and are careful. In practice it can be rather tricky to do, as discussed above. The average user does not get consistent results, so we do not recommend it for "normal" pool owners. People with extensive test experience, who are willing to take the time to understand the test, can get excellent results (and the occasional stained finger).

The AquaChek salt test strips have been very reliable/accurate/precise/repeatable in most cases, but wildly wrong every now and then (in a difficult to understand pattern). So far, I haven't been able to pin down what is going wrong, bad test technique, bad batch of strips, etc? They work correctly far more often then they fail, are reasonably easy to use, and there isn't any obviously superior alternative, so we recommend them.

TDS is total dissolved solids, so it is supposed to count everything. It should always be higher than the salt level. Since there isn't any significant value to knowing the TDS number, it doesn't tell you anything useful unless you know which particular solids are dissolved, I haven't really explored the advantages/problems with the various TDS tests.

 

[polyvue]

The AquaChek salt test strips have been very reliable/accurate/precise/repeatable in most cases, but wildly wrong every now and then (in a difficult to understand pattern).

I haven't taken a survey but I have noticed that a number of individuals who have salt water pools have complained that the AquaChek salt test strips registered a significantly higher level of salt than what their controller is telling them. I appreciate that it makes more sense pragmatically to adhere to the SWG cell's report of salt ppm unless there's an apparent problem with its output, but being of a curious nature I used the strips to confirm it and hence, this issue. Haven't read anything so far here that would dissuade me from theorizing that high-TDS (whatever its source) may have a bearing on this discrepany. Personally, I'll be using the sodium chloride drop kit in future testing (don't really think it's more difficult than adjudging the subtle shades of pink indicating pH, BTW!)

TDS is total dissolved solids, so it is supposed to count everything. It should always be higher than the salt level.

It's supposed to. But I don't think the Taylor kit is counting sodium chloride in its report of TDS. Done the test several times with consistent outcome. Thought of contacting Taylor for explanation but haven't yet done so...

Thanks very much for your comments, Jason.

 

[mas985]

You can count one more who isn't too found of the test strips.

I have found that the drop kit and my SWG read fairly close together +- 100 ppm while the strips consistently read higher by almost 500 ppm at times. They say to read to the top of the white peak, which is often hard to determine, but I have found that if you read to the point where the white is fully across the strip, the reading is a bit more accurate. Combine that with the short shelf life of the strips and I think inaccuracy outweighs the convenience. I know some swear by the strips but I don't think they are worth their weight in salt. :lol:

 

[polyvue]

Interferences with Quantab Test Strips
Problem Description:
What can interfere with the Quantab Test Strips for chloride?

Problem Solution:
The Quantab Test Strip reaction involves the formation of silver chloride on the strip from the combination of silver on the strip with chloride in the sample. Any anion (for example bromide, iodide, cyanide, and sulfide) which will react with the silver to give an insoluble precipitate may interfere with the Quantab® test and give a higher than actual result.

Strong acids may also interfere, however the strips will work in caustic or extremely high pH samples. Nitrite and nitrate have no adverse effect on the Quantab® test.

Quantabs will work in colored solutions as long as they are clear, however turbid solutions can clog the wick and cause very slow or incomplete reactions. Filter turbid samples before testing.

• Source: Product documentation by AquaChek salt test strips manufacturer: Interferences with Quantab Test Strips. Hach Company.

• 
  Bingo!
  
  The Quantab Test Strips are better known as... AquaChek Salt Test strips!
  
  I leave it to the chemical engineers (Richard, Mark ...?) to speculate as to which anion is the best candidate in this interference with the silver ions attempt to combine with chloride. But if the substance could be identified, it might also have played a contributive role toward elevating TDS.
  
  What sez you guys?
  
  
  See also http://www.hach.com/hc/view.knowledge.b ... lRFeQ==%7C
  BTW Turbidity was not a significant factor in the samples I used to perform the salt strip testing.

 

[JasonLion]

I would not expect any significant amount of iodide or cyanide in a pool. Bromide and sulfide are much more plausible, though the concentrations will never be anywhere near as high as the salt concentration.

 

[Melt In The Sun]

Wouldn't any sulfide be oxidized to sulfate in a swimming pool in fairly short order?

 

[polyvue]

Wouldn't any sulfide be oxidized to sulfate in a swimming pool in fairly short order?

That's my understanding, too. It was an epiphany of sorts when I stumbled onto the web site of the salt strip manufacturer to read about pontential interference with the mechanism that provides the test result but I still don't know what that substance could be, nor have I yet been able to identify what contributed to a rapid increase in Total Dissolved Solids over the summer in my pool. There may be a connection and if so, an explanation for all tbose folks who have likewise reported a large discrepancy between the amount of salt in their pool as reported by their SWG controller and the test result via AquaChek salt test strips; almost always, the salt test strips are showing much higher levels of sodium chloride than is known to be the case.