Quantifying TDS constituents affect on freezing point of H20

[polyvue]

In a recent post I speculated on what impact salt and other TDS components might have on the freezing point of pool water. Does anyone here have a formula or explanation as to how much a given quantity of salt reduces the freezing point?

The presence of chlorine, calcium and other constituents in pool water should lower the freezing point to below 32°F, the commonly accepted point at which water freezes at sea level (0' elevation.) Those who add salt to their pools (whether or not they have a SWG) or have high levels of TDS have even less to worry about regarding freezing pool water. My [very modest] understanding of this process is that the salts interfere with the conjoining of hydrogen and oxygen; I believe the extended quote (see below) also states this. What I don't know is how much these salts affect the freezing point.

The following explanation, from a grad student writing in 1998, is expressed very clearly. (Admittedly, Ms Benison has dropped a zero from the stated salinity of ocean water--which I understand to be roughly 35,000 ppm, but I think her points are still valid.)

Salt lowers the temperature at which water freezes and melts. Pure water, H2O, freezes (and melts) at 0°C (32°F). But the more "salt" (any elements or compounds carried by the water in solution, such as Na (sodium), Ca (calcium), Cl (chloride), and SO4 (sulfate)) in water, the lower its freezing point. For example, seawater, which has approximately 3500 parts per million "salt" (including Na, Ca, Cl, SO4, Mg, K, and CO3), will freeze (and melt) at -2.2°C. A water with extreme salinity such as very salty lake waters at Death Valley, California (approximately 300,000 parts per million "salt") may freeze and melt at temperatures as low as -20 - -30°C! Because salt lowers the freezing point depression, it is added to icy roads in order to melt the ice.

• Source: http://www.madsci.org/posts/archives/ma ... .Ch.r.html

If anyone here has a firmer, or different, understanding of this I hope they will post it.

 

[chem geek]

I answered this question in this thread. The effect is negligible at around 0.3ºF for 3000 ppm salt.

 

[waste]

Poly, I would go with Richard's assessment (Hi Richard, it's been a while - I hope that you, your wife and your pool are all well )

R.E. your source's quote that sea water is "3500" ppm salt - I think she left out a "0" --- it's my understanding that sea water is ~ 36,000 ppm salt (Grad students tend to be ~ younger and we can understand a typo here or there - now I just need a reason to excuse my own typos :lol: )

 

[polyvue]

OK. So if one includes the other salts she refers to (calcium, sulfur, etc.) that contribute to a high TDS (in my pool, 6000+ inclusive of sodium chloride) --- even then, the difference is perhaps only 1-2ºF?

So, not near enough to disregard freeze protection in pipes/pools that are expected to have an air temperature in the low 20s or below for an extended period but enough that I shouldn't worry too much about an internal cell temperature (or other water temp) of 35º F, if I know this to be the cellar. This information is probably most valuable in the South and on the west coast here in the US. I worried myself a bit when I cut off the pump during last week's cold spell, but knowing that I have a one (or two) degree buffer probably means that I won't ever have to initiate protective measures for my pool (in Sacramento.) thanks.

 

[chem geek]

Nearly all of the TDS in a saltwater pool is from the sodium chloride salt. A fresh non-salt pool that has saturated calcium carbonate from adding calcium chloride for calcium and sodium bicarbonate for bicarbonate will have a TDS of around 525 ppm and have around 350 ppm salt (chloride as ppm sodium chloride). That is, the TDS of calcium and bicarbonate is not very high. Where do you get that your pool has 6000 ppm TDS but only 3000 ppm salt? That is very, very unlikely.

Even with 6000 ppm TDS, the freezing point depression would only be around 0.6ºF since it's roughly linear.

 

[polyvue]

Nearly all of the TDS in a saltwater pool is from the sodium chloride salt. A fresh non-salt pool that has saturated calcium carbonate from adding calcium chloride for calcium and sodium bicarbonate for bicarbonate will have a TDS of around 525 ppm and have around 350 ppm salt (chloride as ppm sodium chloride). That is, the TDS of calcium and bicarbonate is not very high.

Where do you get that your pool has 6000 ppm TDS but only 3000 ppm salt? That is very, very unlikely.

Using test strips, results were consistent but steadily increased, from, in May:

TDS 3500 - ~5000
Salt 3660

to (by August):

TDS 5000++ (very pale green, 5000 is top of scale; I estimated 6000)
Salt 3970

In September, I began using drops tests for sodium chloride and TDS. The results were:

TDS 2400
NaCl 3000

After a bit of rain overflow, my last measurement, taken in October, was:

TDS 2100
NaCl 2800

Note: The conductivity measured by the Aqua Logic has pretty consistently reported in the 2900-3000 range since early summer. I'm out of salt strips but the fresh bottle of TDS strips I got (at end of August and last used in late October) show a TDS level of 5000+

 

[mas985]

The TDS measurement is inclusive of salt, so it should always read higher than the salt level. If it is reading less than salt, one of the measurements is in error. But if you just add the salt level to the calcium level, you should be fairly close to TDS.

 

[Durk]

Enough Calcium Hardness and you never have to worry about freezing. We put CaCl in our tractor tires and they are good to around -50°F. Of course, that takes 4-5 pounds per gallon--that would be about 60 tons or 3 big dump trucks for my pool. It would also corrode any air exposed metal it touches, let alone what it would do to a swimmer. (It's 30% by weight, I don't know the ppm equivalent but I am sure it has five zeros.)

As salt concentration rises, the "slush free" and "solid freeze" temps become further apart. (For salt-free H2O, they are both 32°F.) For the loaded tires above, "slush free" is -12°F and "solid freeze" is -52°F.

http://www.agf.gov.bc.ca/resmgmt/publist/200Series/210104-1.pdf

 

[polyvue]

The TDS measurement is inclusive of salt, so it should always read higher than the salt level. If it is reading less than salt, one of the measurements is in error. But if you just add the salt level to the calcium level, you should be fairly close to TDS.

I have no reason to disbelieve you... but I do have test results, taken from pool samples and fill water that would seem to refute this. I use the following kits. The TDS is reported as calcium carbonate.

K-1764 http://www.taylortechnologies.com/produ ... KitID=2215
K-1766 http://www.taylortechnologies.com/produ ... KitID=2176

 

[mas985]

The conductivity TDS performed by some pool stores, is inclusive of all elements of TDS including salt. However, the Taylor test kit you are using seems to be only inclusive of calcium carbonate which is basically a calcium hardness test so it would only results in a true TDS measurement if there were no other solids such as salt.

Also, your SWG is basically a TDS measurement so that would include all solids (salt, calcium, magnesium, etc). So unless your SWG is shutting off because of high TDS/salt, I don't think TDS is really that high.

 

[polyvue]

The conductivity TDS performed by some pool stores, is inclusive of all elements of TDS including salt. However, the Taylor test kit you are using seems to be only inclusive of calcium carbonate which is basically a calcium hardness test so it would only results in a true TDS measurement if there were no other solids such as salt.

Also, your SWG is basically a TDS measurement so that would include all solids (salt, calcium, magnesium, etc). So unless your SWG is shutting off because of high TDS/salt, I don't think TDS is really that high.

I wish I could get a handle on all of this... Taylor insisted when I spoke to them that the TDS test I purchased was the right way to account for total dissolved solids. Neither does the (TDS drop test) result seem to be equivalent to Calcium Hardness since my CH result is significantly less than the TDS. Here's the full results from the last time I used K2006 tests along with sodium chloride and TDS drop tests:

Water temp 64 degrees F
SWG Salt reading 2900 ppm
FC 2.8
TC 3.0
CC 0.2
pH 7.6
TA 65
CH 275
CYA 50
TDS 2100*
NaCl 2800

[*TDS detail: 2 drops sulfuric acid, 40 drops sodium carbonate]

There's a small amount of borax in the water, but this was not tested.
No other metals or salts (iron, copper, magnesium) were tested.

 

[mas985]

I read through the instructions again, and I think the test kit is measuring both calcium carbonate and sodium carbonate. The kit web page is confusing as it mentions only calcium carbonate but in the instructions it also mentions sodium carbonate so I think it does both.

However, the second kit measures sodium chloride so I am not entirely sure that sodium carbonate will have the same ppm concentration as sodium chloride so that may explain why the numbers don't quite add up.

The TDS should be approximately = 2800 (salt) + 275 (CH) = 3075

Now in the TDS instructions, there are actually two test. Do you have the results for each half of the test?

TOTAL DISSOLVED SOLIDS (TDS) (1 drop = 50 ppm)
COMPONENTS:
1 x 4026 Dipper, plastic, large 1 x 5074 Instruction 2 x 9198 Sample Tubes, Graduated, 25 mL, plastic w/cap 1 x R-0007 Thiosulfate N/10, DB 1 x R-0645 Total Alkalinity Indicator, DB 1 x R-0809 TDS Resin 1 x R-0810 Sodium Carbonate .24N, DB 1 x R-0811 Sulfuric Acid .24N, DB
TO ORDER REPLACEMENT PARTS AND REAGENTS CALL TOLL-FREE800-TEST KIT (800-837-8548).
PROCEDURE:
CAREFULLY READ AND FOLLOW PRECAUTIONS ON REAGENTLABELS. KEEP REAGENTS AWAY FROM CHILDREN.
1. Rinse and fill 25 mL sample tube (#9198) to 10 mL mark with water to be tested.
2. Add 3 drops R-0007 Thiosulfate N/10. Swirl to mix.
3. Add 5 drops R-0645 Total Alkalinity Indicator. Swirl to mix. Sample should turn green.
4. Add R-0811 Sulfuric Acid .24N dropwise, swirling and counting after each drop, until color changes from green to red. Record drops of R-0811 Sulfuric Acid .24N used. Always hold bottle in vertical position. Discard sample.
5. Rinse and fill 25 mL sample tube to 25 mL mark with water to be tested.
6. Add 3 drops R-0007 Thiosulfate N/10. Swirl to mix.
7. Using a large, plastic dipper (#4026), add 2 level dippers R-0809 TDS Resin. Cap and shake vigorously for 1 minute. ALLOW TO STAND UNTIL RESIN SETTLES. Carefully decant 10 mL into second sample tube.
8. Add 5 drops R-0645 Total Alkalinity Indicator to second sample tube. Swirl to mix. Sample should turn red.
9. Add R-0810 Sodium Carbonate .24N dropwise, swirling and counting after each drop, until color changes from red to green. Record drops of R-0810 Sodium Carbonate .24N used. Always hold bottle in vertical position.
10. To calculate parts per million (ppm) total dissolved solids as calcium carbonate add drops of R-0811 Sulfuric Acid .24N and drops of R-0810 Sodium Carbonate .24N (Steps 4 & 9) used. Multiply by 50.

 

[polyvue]

Now in the TDS instructions, there are actually two test. Do you have the results for each half of the test?

10. To calculate parts per million (ppm) total dissolved solids as calcium carbonate add drops of R-0811 Sulfuric Acid .24N and drops of R-0810 Sodium Carbonate .24N (Steps 4 & 9) used. Multiply by 50.

Yes, I listed each in my last post:

2 drops sulfuric acid, 40 drops sodium carbonate [R-0811, R-0810]

2 + 40 = 42 * 50 = 2100 ppm

 

[chem geek]

The TDS measurement is pointless and I don't understand why you are trying to measure it. If you want to know the conductivity of your water for your SWG, then do a conductivity test since that is all it really cares about. If you want to know the saturation index to prevent scaling and to protect plaster, then measure pH, TA, CH, CYA and temperature.

In the TDS test Taylor is using (an acid-base titration), the first part is like a standard TA test that is mostly measuring bicarbonate ion. It uses a Total Alkalinity indicator that changes from green to red when the pH goes from above to below 4.5 as acid is added. It will NOT include dissolved carbon dioxide, but that probably isn't normally counted in TDS tests anyway as it would not remain if the water were evaporated (of course, some of the bicarbonate wouldn't remain either as carbon dioxide would tend to outgas having the pH rise). The second part of the Taylor TDS test appears to be using a TDS Resin that is in essence an ion exchange resin that exchanges all positive ions (sodium, calcium, magnesium) for hydrogen ions. The (first) test appears to not count chloride ions but charge balance requires charge pairing but their simple sum calculation would appear to be leaving out chloride. [EDIT] The first test accounts for the pH buffering so measures what the acid in the second test did to get the pH down to 4.5 so when added to the second test that measures what it takes to get the pH up to 4.5, the entire range is counted. [END-EDIT]

Then there is the issue of units. The test is measuring in units of ppm calcium carbonate but in fact it has no way of knowing the actual substances in the water so cannot give an accurate weight-based TDS (that is, ppm). It can only give you a molar amount that is artificially scaled in ppm calcium carbonate units. 1000 ppm calcium carbonate is the same as 584 ppm sodium chloride in equivalent molarity, but accounting for the factor of 2 charge for calcium and carbonate, these are close where 1168 ppm sodium chloride would theoretically end up getting measured in this TDS test as 1000 ppm calcium carbonate equivalent IF chloride were actually getting counted (which it isn't, as far as I can tell [EDIT] actually it is indirectly as all cations are counted [END-EDIT] ).

What is interesting is why Taylor bothers with the first part of the test since the second part should measure virtually all cations (positively charged ions) so with charge balance one could assume a TDS based on that. Perhaps they assume that the bicarbonate measured in the first part of the test comes from carbon dioxide, but in fact it often comes from dissolved calcium carbonate from rock as well. [EDIT] I figured out that the first part of the test is used to measure the amount of acid from the starting pH down to 4.5 while the second part of test measures from having all cations replaced with hydrogen ions and then going up to a pH of 4.5. [END-EDIT]

Why don't you try measuring the TDS of the tap water and then add a known amount of sodium chloride table salt to another sample and then measure it using the salt (chloride) test and the TDS test. The salt (chloride) test will tell you the true amount of chloride, though it is expressed in sodium chloride units. Though in any water the chloride portion would be correct (and you could scale the salt test by 35.453/58.443 to get the TDS of chloride alone), some of the cations will be calcium and magnesium though most will be sodium in most cases (when salt is added to the pool or if the pool water is older and chlorine has been added over time).

Richard

 

[polyvue]

Whew! Your text evidences lots of knowledge in this area, but the heavy use of modifiers (some, most of the time, in most cases, etc.) leaves me a bit perplexed as to how to resolve this. Could you explain what you mean by your suggestion to "scale the salt test by 35.453/58.443 [...]"?

Are you discounting the entire Taylor TDS drop test for purposes of gauging TDS constituents in pool water? You said the TDS measurement was "pointless" but I'm just trying to gain some knowledge of my pool water and a deeper understanding of what matters. My motivation is partly to help my own maintenance efforts and partly to be better able to assist others with questions they may have or problems they encounter. Surely, you understand and appreciate this.

 

[chem geek]

Sorry if I sounded frustrated, but I was a bit. You really, really, really do not need to worry about the TDS. Honestly. It's what makes up the TDS that is by far more important and the Taylor test does not tell you that at all. "35.453/58.443" means multiply by 35.453 and then divide by 58.443.

I think I figured out the Taylor test. It's basically a measure of all cations (positively charged ions) with the assumption of charge balance for the negative ions and then scaled to be in units of ppm calcium carbonate. The first part of the test is standard TA and is used to account for the fact that the second part of the test produces acid that will be counteracted by the pH buffering so the first part of the test measures what it takes to get down to a pH of 4.5. The second part of test measures what it takes to get up to a pH of 4.5 since it's lower in pH due to the hydrogen ions replacing all other cations.

So the Taylor TDS test is truly not very meaningful since its units of ppm calcium carbonate do not reflect what is actually in the water, which in your case is mostly sodium chloride salt. Nevertheless, we can compare your numbers with what is expected as follows.

Your TA of 65 ppm showed up as 2 drops in the TDS test since it basically rounded up and results in 100 ppm TDS though it's really only 65 ppm (as calcium carbonate). The test is really mostly measuring bicarbonate ion.

The CH of 275 ppm is also in units of calcium carbonate but is really just measuring calcium.

The CYA of 50 ppm is in units of ppm cyanuric acid.

The NaCl salt of 2800 is in units of ppm sodium chloride.

If we convert the salt to calcium carbonate units and adjust for charge, then this is 2800 * (1000 / 1168) = about 2400 ppm.

So if the TDS test were more accurate (and assuming the salt test is accurate), it would measure at least 2400 ppm and maybe even a bit higher. The calcium wouldn't necessarily be added to this because much of the calcium could have been added from calcium chloride so is already accounted for in the (chloride) salt reading. I don't know why the TDS reading is as low as 2100 ppm, but its accuracy can't be better than +/- 50 ppm and depending on the sample size for your salt test I suspect that it is 200 ppm per drop so +/- 200 ppm which gets these two numbers at least in the ballpark. You could use a 25 ml sample size for the salt test and then multiply the number of drops by 80 ppm instead of 200 ppm. Note that as chlorine is used over time, it increases the sodium chloride salt level and nothing else, so measuring the salt (chloride) level is a darn good proxy for what is mostly in the water in most pools, especially in an SWG pool.

Again, the TDS does not matter. It does not tell you if there is too high a TA, CH or pH combination that can lead to scaling or too low a combination that can lead to dissolving of plaster; it does not tell you if the CYA is too high which makes chlorine ineffective or if it is too low which can have chlorine break down too quickly in sunlight; it does not tell you if the TA is too high which can lead to the pH rising over time or if it is too low which may not provide enough pH buffering if using acidic sources of chlorine; it does not tell you if the pH is too high or too low which can cause eye irritation or scaling or metal corrosion.

 

[polyvue]

Now I feel like I'm beginning to understand this a bit. (Though I guess I'm like the student who wants to be sure he's getting an A on the final.) Thanks for your patience.

Again, the TDS does not matter.

This began because the Aqua Chek strips I was using to check salinity starting showing a dramatic increase in salt --far, far over what the SWG reported; this was accompanied by a surge in measured TDS (also using test strips.) My goal was to resolve the discrepancy between the SWG cell and the tests. As a corollary to this finding I thought it would be good to know if there were other minerals that were tweaking the test strip results.... copper or phophates or nitrates or something else that might be problematic. My first (and so far, last) step was to determine the total dissolved solids, before trying to identify which constituents were causing the increase.

So, that's the history. I'll try to work through the above in the next few weeks and come to some agreement with myself over how precisely I really want to identify the constituents. Though I call myself a testing fanatic, I'm not really. Just thirsty to learn more.

 

[JasonLion]

This is the deep end, but it is still important to keep in mind that there is some complexity here that is difficult to disentangle and really won't be of anything other than academic interest. Both salt and TDS tests can be wrong for a number of different, difficult to explain, reasons. Plowing through all of those possibilities is complicated and never worth the effort (in my experience).

 

[PaulR]

I was interested in the original question, how stuff in the water affects freezing point, partly because my chemical-engineer dad had relayed the story that Fahrenheit chose his zero as the freezing point of a saturated salt solution. (Wikipedia tells a slightly different story.) However, it takes a ridiculous amount of salt to reach the saturation point, and it looks like the effect is on the order of 0.1F per 1000ppm. Not enough to have a practical effect in a pool/spa.
--paulr

 

[chem geek]

This began because the Aqua Chek strips I was using to check salinity starting showing a dramatic increase in salt --far, far over what the SWG reported; this was accompanied by a surge in measured TDS (also using test strips.)

Test strips are often not accurate and this is apparently very true for the TDS test strips.