Pinellas County, FL Pool Study 1992 (1994)

[chem geek]

(...CONTINUED FROM PREVIOUS POST)

WHAT COULD POSSIBLY BE GOING ON?
Consider the following pieces of information:

• Many studies have shown that hypochlorous acid (HOCl) is the disinfecting form of chlorine for bacteria, protozoa (not that good for cysts, however), and viruses and that it is the form that inhibits algae though monochloramine is also effective for some forms of algae (yellow/mustard). These studies are mostly done in distilled water though some are done in simulated pool water, but simulated only in terms of a carbonate buffer and calcium, not in terms of organics from bather load.
• Clifford White’s “Handbook of Chlorination” and other sources refer to “base potential” or “poise” of different waters such that different amounts of chlorine are needed to reach a given ORP value. Or put another way, an amount of chlorine added to different waters such that they each measure with the same FC can produce different ORP values and therefore have different HOCl concentrations.
• Several studies have found that ORP is a far better predictor of water quality than FC alone. Other studies found that ion-specific membrane HOCl sensors were also effective and that ORP seemed to track HOCl concentration at a given pH and temperature (ORP varies with pH and temperature in ways beyond corresponding to the HOCl concentration).
• Cyanuric Acid (CYA) combines with hypochlorous acid (HOCl) to form several chlorinated isocyanurate compounds. Several studies have shown that these compounds are not effective sanitizers. Experience of users on pool forums shows that they also do not inhibit algae (notwithstanding one of the two studies I sent earlier that disputes this, but whose methodologies were flawed). The HOCl is released from the Cl-CYA compound at half-life rates on the order of seconds (0.25 seconds for one species; 4 seconds for another) such that the Free Chlorine tests (OTO, DPD, FAS-DPD) register the HOCl, OCl-, AND the Cl-CYA species by taking up (with a dye) the chlorine from all of these compounds.
• The Total Chlorine test which allows one to determine Combined Chlorine (CC), releases the bound chlorine from chlorinated organics and chloramines.
• Data from pool forums with residential pool users strongly validate the chlorine/CYA relationship with respect to green algae and appear to do so for mustard/yellow algae as well though there are too few incidents with that algae to be as certain. Residential pools do not appear to exhibit the sort of behavior of commercial pools as found in the Pinellas study. Instead, residential pools appear to exhibit the theoretical S-curve behavior.

If you haven’t figured out where I am going yet, consider the following link to a study that shows that monochloramine will register in the FC test if given enough time :

http://pubs.acs.org/cgi-bin/abstract.cgi/esthag/1984/18/i05/f-pdf/f_es00123a011.pdf?sessid=6006l3

The paper in the above link states the following:

This study confirmed that NH2Cl oxidizes DPD to the colored intermediate at a rate of 5.6 and 6.0% of the NH2Cl concentration, in the first minute, at 25ºC and 5x10^-5 M (3.5 mg as Cl2/L) and 1x10^-4 M (7.0 mg as Cl2/L) NH2Cl, respectively.

Have you guessed what I am about to hypothesize yet? Think about it. Monochloramine is normally considered to be a “combined chlorine” that should not get measured in the Free Chlorine test. And yet it does slowly give up its chlorine to be measured in the Free Chlorine test. We already know that Cl-CYA compounds will give up their chlorine relatively quickly (in seconds) to be measured in the FC test. So why can’t there be other organic compounds that behave like CYA in that they combine with HOCl to form new compounds, but are not bound as tightly as combined chlorines measured in the TC/CC test, but rather are more loosely bound more like the Cl-CYA compounds? These would be part of the “base potential” or “poise” of different waters referred to above.

In fact, if there were such chemicals, then they would precisely explain the results of the Pinellas study. The bulk of the pools exhibited the expected S-curve behavior requiring only a very low amount of around 0.002 ppm HOCl or perhaps even 0.001 ppm HOCl (measured in units of ppm Cl2, of course). This is, in fact, the expected amount of HOCl based on the hetrotrophic 2-log CT(99%) value off 0.08 since this corresponds to a CT(50%) value of 0.08*0.301/2 = 0.012 and using 15 minutes as the low end of the 15 minute to 1 hour typical generation time for bacteria this implies 0.012/15 = 0.0008 ppm HOCl.

If there were a CYA-like substance, then pools with higher calculated HOCl (using pH, FC, CYA) could have far lower actual HOCl. But because we are only measuring CYA and not these other substances, we don’t know about them nor account for them to find proper correlations! So let’s take a look at the worst-case for this, namely pool #318. What CYA level would it take to make that pool have a ppm HOCl of 0.001 and therefore be much more theoretically likely to have a high HPC? It would take about 2000 ppm of CYA. Now that sounds like an extraordinary amount, but consider that this mystery chemical (or chemicals, since there could very well be more than one) could bind to chlorine tighter than CYA (i.e. the Cl-Mystery has a smaller equilibrium constant of hydrolysis to release HOCl) and this normally implies that the rate of release of HOCl would also be slower. We know that it can’t be too slow or else it wouldn’t measure in the FC test, but it could be, say, 5 times slower and bind 5 times as strongly so that the amount of this mystery substance could be 2000/5 = 400 ppm in that worst case pool. In fact, it could be 10 times slower and still be measured as FC since 10*0.25 = 2.5 seconds for half-life.

As for why residential pools do not seem to experience this mystery chemical in sufficient quantities to disrupt the S-curve behavior, I can only speculate. My best guess is that residential pools tend to be smaller than commercial pools and have lower bather load. Many residential pools have DE or sand filters that are regularly backwashed (or cleaned, in the case of DE) so that pool water gets replaced over time. So between a low bather load and higher water replacement, there is less buildup of organics, including the “mystery” chemical. In fact, some residential pools with smaller volumes and that have regular backwashing can use Trichlor as their sole source of chlorine and not build up CYA to unacceptable levels — they can keep the CYA below 50 ppm. I looked at the TDS of pools in the Pinellas study to see if there was any sort of trend, but I couldn’t find any — most pools had rather high TDS and a few of the pools with high calculated HOCl but high bacterial counts had low TDS, so regular changing of the water isn’t necessarily the only answer. However, if TDS was measured using a conductivity test, then the “mystery” chemical could very well be neutral, though polar to dissolve in water, and not show up in a TDS test — CYA itself does not show up as TDS, though the CYA- ion does.

HIGH CYA HAS OTHER PROBLEMS
One of the recommendations from the study was the following:

Review with state regulatory agencies and the EPA the feasibility of increasing the cyanuric acid limit in public pools.

Raising the CYA limit above 100 ppm is not a wise thing to do because there are other problems with high CYA other than just needing higher FC (which I believe, but you may still not). The following links show how high CYA levels cause pitting in plaster surfaces:

http://www.tricitypoolservice.com/tc-plaster_study.html
http://www.findarticles.com/p/articl...4/ai_n15932555
http://www.aquamagazine.com/data/archive/AQ-406-61.pdf

Note that the above study was done by Arch Chemicals who manufacture CYA so this is an example of good corporate responsibility so that products will be used responsibly. [EDIT] This study on CYA and plaster has since come into question and may not be valid. [END-EDIT]

(CONTINUED ON NEXT POST...)