CYA vs UV Chlorine Loss Test – Observations

[Tom ONeill]

During the summer of 2021 we switched from using 3-inch Trichlor tablets in Pentair feeders to peristaltic pumps using 10% - 12.5% Sodium Hypochlorite (Bleach) hoping to reduce operating costs and to save water (periodically having to dump a lot of water due to constantly rising CYA). At peak summer we had typically used about 3.88 PPM chlorine per day in each of our 2 approximately 32,500 semi-public pools here in Tucson, AZ, averaging 2.33 8-oz Trichlor tablets/day/pool. After the conversion to Bleach we noted we were having to use substantially more Bleach than we had estimated, based on the 12.8 fl oz/PPM/10,000 gallons (10% Cl2 equivalent) vs 1.48 oz Trichlor/PPM/10,000 gallons guidance.

As a result, I started trying to figure out what was causing the higher-than-expected Bleach use. It wasn’t “weak” Bleach solution. I started testing the chlorine content and found it was as advertised. During my “research” I found a posting on the TFP forum by member mas985, dated 7/1/21, titled “CYA vs UV Chlorine Loss Test – Updated”.

A Recap of Test Conditions presented by mas985 of the single sample source pool water for the initial condition and subsequent dilutions, and the results:

Initial Sample, FC 7.2, CYA 80, FC/CYA Ratio .09, pH 7.2.

Resulting 2-hour FC PPM Loss .4 PPM

1st Dilution 1:1, FC 3.6, CYA 40, FC/CYA Ratio .09, pH 7.02 (adjusted from dilution with pH 6.9 distilled water)

Resulting 2-hour FC PPM Loss 1 PPM

2nd Dilution 1:3, FC 1.8, CYA 20, FC/CYA Ratio .09, pH 6.98 (adjusted from dilution with pH 6.9 distilled water)

Resulting 2-hour FC PPM Loss 1.4 PPM

Initial, 1st and 2nd dilutions were in open-top clear containers placed in mid-day summer sun for 2 hours.

The first thing I noticed was the critical role of CYA in shielding the chlorine from solar UV burn-off.

At one-half the CYA concentration (CYA 40), the FC loss was about 2 times that of the Initial of 80 PPM CYA sample. At one-quarter CYA concentration (CYA 20) the FC loss was about 4 times that of the Initial condition of 80 PPM CYA.
But, note that the FC/CYA ratio is exactly the same in each prepared sample.

Of considerable help to me in understanding the CYA – Chlorine relationship was a web-based application I had found: “Free Chlorine and Cyanuric Acid System Simulator”, created by David G Wahman, U.S. EPA. The tool is based on the equilibrium model presented by O'Brien (1972) and O'Brien et al (1974). Accessible at: https://usepaord.shinyapps.io/cyanuric/ . The application generates graphic representation of resulting values and can generate and download to your computer a CSV file of computed results. Note: All chemical relationships are at 25 °C.

The Application's results are given as negative log molar concentrations. So, you need to figure out the math and conversions.

EPA Application Definitions and Terminology Equivalents:

Total Chlorine: CYA-Bound + Unbound Chlorine = DPD tested Free Chlorine (FC)

Total Free Chlorine: Chlorine Unbound from CYA, HOCl + OCl-. Can’t be directly tested, has to be computed, and is the concentration of the chlorine exposed to Solar UV. Equivalent to Free Available Chlorine (FAC).

Total Cyanuric = Stated CYA concentration

Resultant data points for the 3 test conditions:

Initial 1st Dilution 2nd Dilution
FAC Molar Concentration 0.009295707 0.019057203 0.037339955

FAC PPM 0.066929089 0.068605932 0.067211919

Reported 2-hr FC PPM Loss .4 1.0 1.4

Computed FC PPM Loss .4 .820 1.607

I attribute the bulk of the variance between Reported and Computed FC Loss to the +/- .2 PPM accuracy range of the FAS-DPD test, along with some small variance due to the inability to represent lesser than a .1 pH precision in the computational simulator.

Observations:

At the constant FC/CYA Ratio of .09, the starting FAC PPM remains essentially flat across the Initial, 1st and 2nd dilutions tested.

The FC PPM Loss rate is directly proportional to the FAC Molar Concentration, not the FAC PPM.

Therefore, I believe a Rate Constant loss is present: R * FAC Molar Concentration.

I calculated the 2-hr FC PPM loss Rate in this test as 43.0306 * FAC Molar Concentration across all three test solutions. The Rate Constant expression is dependent on the amount of clear-sky solar insolation, which varies throughout the year and by geographic latitude.

The FC/CYA Ratio is highly correlated to maintaining adequate HOCl PPM for disinfection:

Initial 1st Dilution 2nd Dilution
HOCl PPM (computed) 0.045933451 0.053248808 0.052166838
Note the increased HOCl PPM in 1st and 2nd Dilutions due to the reduced pH of those solutions having shifted the HOCl/OCl- ratio a bit more toward HOCl.

The mas985 test was a “static” test over a two-hour period – declining balance at half-life intervals such as the approximate 12-min interval of OCl- over a 2-hr period, I assume that a “dynamic” constant-feed of chlorine will result in a higher chlorine loss at any given CYA level, particularly when adding unstabilized chlorine as with 10% – 12% pool bleach or with a SWCG, versus using Trichlor, which will continuously and progressively change the FC/CYA ratio.

Caveat:

My calculated FC PPM Loss Rate Constant of 43.0306 * FAC Molar Concentration for the mas985 test samples group is Not Applicable to a real-life swimming pool situation, in which the ratio of exposed pool surface area in square feet / the total cubic feet of water will have significant effect on the calculated loss rate of the full FAC content in the pool, due to UV shielding at depth from the limited UV penetration of water.

However, for me, this was an eye-opener into the mechanism of solar UV chlorine loss and the validity of the previously established FC/CYA ratios.

 

[Jimrahbe]

and the validity of the previously established FC/CYA ratios.

Tom,

What? Did you just think we made this stuff up???

Thanks,

Jim R.

 

[JoyfulNoise]

Thanks for posting this data and taking the time to run down the calculations.

Where in Tucson do you operate your pools?

 

[mgtfp]

Nice work. Just one comment on the Wahman model:

The Wahman model seems to show good agreement with the O'Brien calculation (at 25°C), as long as the FC/CYA ratio is above about 1. But at ratios relevant for swimming pools (FC/CYA around 10%), the model deviates about 30%-35% from the O'Brian calculation.

There seem to be some differences between the concentration ranges in pool and drinking water.

For pool related calculations, I'd recommend to use Chem Geek's pool equations spreadsheet:

https://www.troublefreepool.com/~richardfalk/pool/PoolEquations.xls

 

[Tom ONeill]

No. I never believed that the FC/CYA ratio concept was "made up". I know that Lowry and Richard Falk both worked on developing the concept and that it appeared to be sound.
However, following mas985's posting of his test there was a follow-on posting by Richard, which I can't locate again, in which Richard commented that the FC/CYA ratio may need to be revisited in light of the test results.

The FC/CYA ratio is totally sound. And, when followed, can be relied on to ensure predictable and adequate Free Available Chlorine, both for disinfection and algae control.

TFP has been an invaluable resource for me over the last 6 years. I came into swimming pool operations late - I'm now 75. Background: mechanical, industrial and computer systems engineering, programming, business management, blah blah. No chemistry background. Our location in Tucson is a condominium complex on Pantano Rd, between Broadway Blvd and Speedway Blvd.

I thank every TFP contributor for all their wonderful and knowledgeable information. I've learned a lot.

 

[mgtfp]

I just realised that your link goes to version 0.5 of the Wahman calculator, which is perfect, because it uses the O'Brien constants for 25°C.

I first thought that you were using version 1.0 (https://shiny.epa.gov/fcedts/) of their calculator that uses constants that they measured themselves for much higher FC/CYA ratios than relevant for pools.

 

[mgtfp]

The FC PPM Loss rate is directly proportional to the FAC Molar Concentration, not the FAC PPM.

This one confuses me a bit: Shouldn't the FAC (which is just HOCl + OCl^- in this notation) in ppm be proportional to the FAC molar concentration, and the loss rate therefore be proportional to both, FAC ppm and molar?

I would expect 1 mol/L of HOCl or OCl^- to each be equivalent to 70.906 g Cl₂ / L = 70906 mg Cl₂ / L = 70906 ppm.

Or the other way round, 1 ppm equivalent to 1/70906 mol/L = 1.41 10^-5 mol/L.

Or am I misunderstanding something?

 

[Tom ONeill]

When I first started estimating this I also thought that FAC ppm should be proportional to the FAC molar concentration. It is, but not in an intuitive way. The FAC molar fraction times FC ppm = FAC ppm. But after running mas985's test parameters in the EPA app and doing the math on the results, then analyzing those, it became clear that the FAC PPM had stayed nearly constant across the 3 test conditions, reflecting the constant FC/CYA ratio of .09 in all of the mas985 test samples, but that the FAC molar concentration (or fraction), the portion "exposed" to UV, doubled then quadrupled through the dilutions from FC/CYA 7.2/80 to 3.6/40 and 1.8/20, with virtually no change to the FAC PPM. It was then that the lightbulb lit up.

BTW, PoolEquations.xls calculates the FC sunlight loss as about 2 times higher at FC/CYA 7.2/80 than at 3.6/40. The exact reverse of what occurred in mas985's empirical test. If it hadn't been for that test series, I might never have found the proper relationship.

Fundamentally, the higher the CYA, the lower the FAC (HOCl/OCl- unbound from CYA) and the lower the loss of chlorine to UV. But also lower FAC to kill pathogens. But we already knew about the CT values for pathogens.

I'm neither a chemist nor a math genius. When trying to find base relationships I generally just run ranges of numbers through known or tested values and look for recurring series, patterns or near-correlating end values. Then I go back and try to figure out the details of how the heck I got those results, compared to formal research, until I finally understand it. Works - most of the time for me. I spent the whole of this past winter reading all sorts of papers from very narrowly-focused lab tests and more broadly-afield into atmospheric, drinking water, wastewater treatment, and Advanced Oxidation Process chemistry. The degradation of HOCl/OCl- by sunlight is simply driven by basic Energy inputs and molecular energy absorption coefficients. The higher its starting temperature (molecular excitation level), the less additional energy it takes to break the molecular bonds. For example, in lab experiments they have generated Hydroxyl Radicals (OH*) from HOCl/OCl- directly by UV irradiation through to near-visual wavelengths, as well as with ultrasound. Fascinating.

Still having "fun".

 

[Tom ONeill]

Oh - And the doubling and quadrupling of the FAC molar concentration matched the doubling and quadrupling of the FC losses in the as985 test series.

 

[PoolStored]

TFP has been an invaluable resource for me over the last 6 years. I thank every TFP contributor for all their wonderful and knowledgeable information. I've learned a lot.

If you found value, you might consider a contribution to keeping the lights on....

Become a TFP Supporter

Help Support TFP Trouble Free Pool is run by a dedicated group of volunteers that … Read more…

www.troublefreepool.com

 

[Tom ONeill]

Also, am I breaking this forum's protocols by using my real name? At 75 years old, I don't really care. But . . . I'd rather not be a thorn in everybody else's backside or make anybody uncomfortable.

 

[PoolStored]

Also, am I breaking this forum's protocols by using my real name? At 75 years old, I don't really care. But . . . I'd rather not be a thorn in everybody else's backside or make anybody uncomfortable.

Not at all. It is personal preference. Nobody will feel uncomfortable. If you want to change it then start a conversation with a Mod and they can change it to any name you want.

 

[Jimrahbe]

Tom,

Not sure I've seen that done before...

I hate screen names and always use the OP's 'real' name when known.

It just feels more personal to me.

Jim R.

 

[mgtfp]

Just to make that clear: You should not be able to reproduce Mark's test results with either Richard's spreadsheet or the Wahman calculator. The calculations assume that chlorine bound to CYA is protected, and HOCl and OCl^- decay with their respective half life times.

Richard had done calculations on decay times based on these assumptions and noticed that there seem to additional protection effects that are not explainable by the fractions of FC bound to CYA alone. Like shielding of FC in lower water layers by those above.

That's where Mark came in and verified Richard's ideas on that on an experimental level.

I regards to that proportionality: Could you explain again what you mean exactly by FAC molar versus ppm? I still think these should be proportional if it's just the sum of HOCl and OCl^-. When adding chlorinated CYA species into the picture, it might get not exactly proportional anymore when considering CYA with two and three chlorines attached.

 

[Tom ONeill]

I guess I didn't do a good enough job of presenting my information clearly enough in my initial posting of my "Observations".

I did in fact validate Mark's test results and did in fact use the "Wahman calculator" - The EPA application I referred to and provided the link for.

When I was investigating Mark's test information I was unaware of a prior discussion and/or tests performed between Richard and Mark. In my investigation, one of the things I was trying to figure out was why the three test instances had progressively larger chlorine losses even though the FC/CYA ratio was absolutely identical in each case. As I worked my way through the data it appeared there was a rate constant loss - not against the FC PPM, but rather against the Molar Fraction of the Free Available Chlorine (FAC Molar). And, of course, CYA has a huge effect on the exposure of chlorine to UV. After all, at the initial FC/CYA ratio of 7.2/80, 99.07% of the chlorine is bound to CYA, and therefore nearly totally shielded from UV. But the UV chlorine loss is coming from the un-shielded fraction of chlorine - the Free Available Chlorine.

Your posting clarifies that they were interested in confirming whether chlorine loss by UV would be mitigated by the depth of water. And yes, water does shield the HOCl/OCl- that is unbound from CYA from the UV significantly. UVC doesn't make it through the atmosphere to Earth's surface. Only about 10% of UVB makes it to Earth's surface and can barely break the surface of water. About 95% of UVA reaches the surface and the majority is absorbed in water within the first 12 inches.

If you placed 10,000 gallons of 7.2/80 FC/CYA ratio water into an area of 4,010 sq ft, that might equal close to the depth of water Mark used in the test (guessing 4 inches), and left it for 2 hours, you would probably see close to the same loss of .4 PPM as in Marks test. In a 10,000 gallon pool with an average depth of 4.5 feet the same FC/CYA 7.2/80 ratio would probably lose about .09 PPM or less.

As to FAC Molar Fraction versus FAC PPM -
At 77 degrees F water temperature (25 C) -
The Free Available Chlorine, FAC Molar Fraction multiplied by the Total Available Chlorine PPM (which is the DPD-tested "FC") = FAC PPM.
Total Available Chlorine (TAC) Molar Fraction is always = 1. This is the CYA-bound chlorine + the unbound Free Chlorine Fraction.
If your DPD-tested FC is 2.25 PPM and you have 30 PPM CYA (FC/CYA ratio .075) and have 7.5 pH, your FAC Molar Fraction is .026421.
Multiply .026421 x 2.25 PPM chlorine = .059446 FAC PPM (unbound HOCl + OCl), which is really very good.

Mark's initial conditions of 7.2 PPM FC, 80 PPM CYA (FC/CYA ratio.09) and 7.2 pH had a FAC Molar Fraction of .009296 x 7.2 PPM chlorine = .066929 FAC PPM.

I believe Richard's spreadsheet, although astoundingly detailed and accurate in most ways, is inaccurate in the chlorine Half-Life calculations. I think he may have taken the half-life rate constants as being based on PPM, where in fact they are Molar half-life rate constants. I also think that would explain why his spreadsheet calculates a higher chlorine loss rate at 7.2/80 of 0.638 PPM/Hr vs the 3.6/40 ratio calculated loss rate of 0.334 PPM/Hr. I think because it's likely using the tested FC PPM as its basis and not the Molar Fractions with the loss rate constants.
In fact, the chlorine loss rate at 3.6/40 is nearly 2 times the chlorine loss rate of 7.2/80 (the reverse of what the spreadsheet calculates), and at 1.8/20 the chlorine loss is about 4 times the 7.2/80 ratio, as is clearly shown in Mark's sample tests. And each of the sample conditions were at FC/CYA ration .09.

p.s., Richard is extraordinarily good at his math. Way, way better than I am.

0.026421​

 

[Tom ONeill]

This one confuses me a bit: Shouldn't the FAC (which is just HOCl + OCl^- in this notation) in ppm be proportional to the FAC molar concentration, and the loss rate therefore be proportional to both, FAC ppm and molar?

I would expect 1 mol/L of HOCl or OCl^- to each be equivalent to 70.906 g Cl₂ / L = 70906 mg Cl₂ / L = 70906 ppm.

Or the other way round, 1 ppm equivalent to 1/70906 mol/L = 1.41 10^-5 mol/L.

Or am I misunderstanding something?

It is proportional. FAC Molar concentration (Fraction) * Total Available Chlorine (TAC) PPM = FAC PPM.
I'm not sure where you're trying to go with molar mass of Cl2 and FC PPM. But your math looks fine, to a point.
To begin with, I think you'll find that 1 mole Cl2 in 10,000 U.S. gallons of water (37,584.12 liters) will give you about 0.9433 PPM FC.
One half of the Cl2 dissolved in water becomes hydrochloric acid (HCl), the other half becomes HOCl and/or OCl- in a ratio dependent on the pH of the solution. Also why Cl2 is acidic.
So, for every 1 mole of Cl2, if fully dissolved in water, would only result in one-half of that in a DPD test as measurable FC PPM.
That in itself was quite confusing for me, as everything appears to be expressed as the "equivalent" of Cl2, even though only one-half of Cl2 becomes HOCl/OCl- and therefore measurable as FC (excluding the further complications of the presence of CYA).

That kind of calculation is at the root of, for instance, 1.48 oz TriChlor as 90% available Chlorine being equal to 1.34 oz Cl2 per 1 PPM measurable FC in 10,000 gallons water. The actual ounce calculations are a bit different than those "standard" values.

I don't know if that helps you or confuses you as much or more than I was when I started trying to figure this out.

 

[mgtfp]

So, for every 1 mole of Cl2, if fully dissolved in water, would only result in one-half of that in a DPD test as measurable FC PPM.

Look at it like that: The chlorine in the water is measured in g Cl₂ gas that had been added per L of water. 1 atom of that molecule turns into HOCl, the other into chloride.

 

[Tom ONeill]

My apologies to any all who have followed my postings with resulting confusion and/or frustration with my ignorance, sometimes amplified by some of my muddled explanations. I'm sure there are many members with vastly more knowledge and ability than I in chemistry and mathematics for whom my ignorance could lead to frustration.

I should have clearly stated where I was coming from and what I was hoping to learn, and the approach I had taken to find some answers for myself.

What I wanted to find out:
How can I figure out, under whatever combinations of CYA, FC and pH, what percentage (fraction) of each Mole of chlorine in the pool is in the forms of HOCl and OCl-? How do those results change as you change those conditions? Allowing me then to do some comparative analysis. For me, this is not a purely scholarly pursuit. The answers to my questions will help my decision making in operating our 2 semi-public pools, hopefully better, more efficiently and economically. And within County Health Code operating boundaries.

So - Research the problem. I quickly found some guideline estimates like "98% of the chlorine is bound to CYA", approximately 50% unbound chlorine at pH 7.5 is in the form of HOCl. I read, downloaded a lot of information I located that ChemGeek had posted on TFP, as well as from other sites or referenced documents online, that might be applicable and analyzed that info. Read and downloaded a lot of basic research papers, including the O'Brien papers, that presented elements that I felt were relevant to swimming pools, and try as best I could to understand those. Most of the time having to Google for terminology definitions, chemical reactions, etc.

Anyway, after many untutored attempts at calculating (in other words, getting myself wrapped around the axle), I happened upon the EPA (Wahman) calculator app. Aha! Thought I.
However, the app generates all its values as the negative Log of the molar concentrations (mol/l) at the given conditions. OK. I'll presume they're accurate, as they're based on the prior work of O'Brien, et al. Now, how to separate out just the fractional molar relationships from the concentrations?

Ultimately, I figured out a way to derive the Molar Fraction from the negative Log of the Molar Concentration, at least for total free chlorine, HOCl and OCl-. For all the rest, I'll have to learn how to use the negative log molar concentration values directly. That is, if I ever really need to know something like what fraction of each initial mole of TriChlor (Cl3CY) in solution is in the form of H2ClCy under various conditions.

Being able to calculate the Molar Fractions at different FC, CYA and pH conditions, allowed me to solve, at least for myself, how and why Mark's test results came out as they did, and the implications of that for our pool operations.

Thanks for your patience.

 

[mgtfp]

I'd suggest to use Chem Geek's spreadsheet for your calculations. You get all the ppm and molar concentrations directly (when you scroll far enough down, you'll find all the molar concentrations).

 

[Tom ONeill]

In the ChemGeek's spreadsheet: Mark's test conditions are in the "Initial" column = Initial pool sample at 7.2 pH, 7.2 FC, 80 CYA. "Goal" column = Mark's 1st dilution sample at 7 pH, 3.6 FC, 40 CYA. Go to line 397 of the spreadsheet. The calculated "FC usage by sunlight (ppm/hr)" row. Mark's DPD-tested FC loss in his Initial sample after 2 hrs (not 1 hr) was .04 ppm. Mark's DPD-tested FC loss in his 1st dilution sample after 2 hrs (not 1 hr) was 1.0 ppm (for which I postulated .820043 ppm loss. His FC loss at the 40 ppm CYA was a bit over 2 times the FC loss than at 80 ppm CYA.

Please explain how that could be when the spreadsheet shows nearly the reverse results. I'm a firm believer in Actual, reasonably controlled empirical test results. Then I try to figure out why there is a discrepancy between the theoretical result and the empirical test result.

And please don't introduce depth of water. All three of mark's test samples were in rather shallow, clear-walled, open vessels. Other than FC and CYA concentrations, the test conditions of temperature and sunlight exposure were identical.

I wish ChemGeek had published a detailed user manual for the spreadsheet. Most of the lower end of the spreadsheet and the math is well over my head. I'm more a "pattern" and relationships analyst type, and I don't have the years left to go for a Chemistry Masters or PhD.

Thanks for any help you can provide, because at this point, I'm either correct or totally lost.