Disinfection Rates and Minimum Chlorine Levels

[chem geek]

The purpose of this thread is to document the disinfectant Concentration times contact Time (aka CT) constants that determine how much of a disinfectant is needed to kill a specific microorganism. This lets us determine minimum chlorine levels required for sanitation under different circumstances (e.g. pools vs. spas).

First, some background. It is found that generally speaking the product of the concentration of a disinfectant with the time it takes to kill a certain percentage of microorganism is roughly constant. That is, if you double the concentration, then you kill the same percentage in half the time. Technically, this rule works when the rate of kill is what is known as "first-order", meaning that it is directly proportional to the concentration of the disinfectant (as opposed to being proportional to the square of the concentration or some other power).

Typical CT values are quoted in disinfectant concentration units of mg/liter, which for our purposes is the same thing as ppm, and in contact time units of minutes. The CT value is a function of temperature and generally gets lower at higher temperature which means that it is easier to kill pathogens at higher temperature (which is interesting because cells reproduce faster at higher temperature, but apparently with increased temperature the rate of disinfection rises faster than the reproduction rate). It is also generally true that CT values increase somewhat with higher "C" (concentration) which is probably an indication that the reaction rate is a little less than first order. The CT value is also a function of pH, but this pretty much tracks the proportion of hypochlorous acid (HOCl) out of Free Chlorine (which includes hypochlorite ion, OCl-). There is never any CYA present when computing any of the CT constants I have seen.

Some terminology that is used to describe the percentage that is killed is called "log reduction" or "n-log" as in "2-log", "3-log" or "4-log". This refers to powers of 10 of inactivation of microorganisms. 1-log is 90%, 2-log 99%, 3-log 99.9%, 4-log 99.99%, 5-log 99.999%, etc. The CT values for each "log" inactivation are related through simple math of the "log" numbers, so that doubling the 1-log CT value gets the 2-log CT value; tripling the 1-log CT gets the 3-log CT value; doubling the 2-log CT value gets the 4-log CT value; and so forth. In practice, CT values are almost always looked at in the 1-log or higher category which means at least 90% inactivation, but in reality there are two competing factors: 1) the disinfectant is inactivating the microorganisms at some rather constant rate and 2) the microorganism population reproduces at a fairly constant rate (i.e. doubling every so many minutes or hours, known as the "generation time"). The higher "log" inactivation constants focus on this first factor and not the second. However, to determine the absolute minimum and "critical" amount of chlorine, one needs to know how much is needed to kill half of the microorganisms in roughly the same time as these microorganisms would double through reproduction (typically, cell division). Obviously, we want to have higher levels than this critical amount in our pools, but it is useful to know this amount nevertheless.

So here's how to calculate the CT value at 50% (critical time) from a CT value at another "n-log":
CT(x%) = CT(y%) * log(1-(x%/100%)) / log(1-(y%/100%))
log(1-(50%/100%)) = -0.30103
log(1-(y%/100%)) = -n (where "n" is "n-log")
CT(50%) = CT(n-log) * 0.30103 / n
The CT(50%) is in some sense "0.3-log", but nobody calls it that.

So what is the "generation time" for varioius microorganisms. This link gives some useful numbers for bacteria and though it ranges widely, most bacteria have a generation time from 15 minutes to 1 hour. Viruses do not reproduce in water as they need a host in order to reproduce. Protozoa and cysts that are resistant to disinfection have a generation time of at least several hours (different sources varied considerably). Most algae seem to have a generation time of about 6 hours though some are closer to 3 hours.

Now let's look at some CT numbers. Appendix B in this EPA document gives CT values for hard-to-kill microorganisms. The highest temperature listed for Giardia cysts is 25C (77F) which is still lower than typical pool and certainly spa temperatures, but at a pH of 7.5 and low disinfecting chlorine levels (as typically found in the presence of CYA), the 3-log (e.g. 99.9% inactivation) CT value is 42. This hearty fellow simply isn't going to be killed quickly in water in the presence of CYA without some serious shocking. At 30 ppm CYA it would take about 7 ppm FC to inactivate 99.9% in 3 hours. However, the CT(50%) value is only 42*0.30103/3=4.21 so at 3 ppm FC (equivalent to 0.087 ppm FC with no CYA) it would take about 48 minutes to inactivate half of the cysts and that is probably quite a bit faster than their generation rate. So though this would be unacceptable in a commercial pool, in a residential pool taking longer to kill a microorganism isn't as important so long as it does get killed.

The protozoa Cryptosporidium, according to this source has a CT value of 7200 (though values vary widely, but are all very high). Even pools with no CYA are going to have a tough time with that one!

In a spa, the bacteria that causes "hot tub itch" is Pseudomonas aeruginosa and this document and this document imply CT(99.99% or 4-log) values on the order of 30-50. That means that hot tubs should be run with either no CYA or a little CYA (less than 20 ppm). The CT(50%) value is only 50*0.30103/4=3.76 so at 15 minutes generation time this means an FC level of 0.25 without CYA or 8 ppm FC at 20 ppm CYA (4.1 ppm FC at 10 ppm CYA). Since the CT values were at nominal water treatment temperatures while spa temperatures are much higher (at least when in use), having an absolute minimum of 4 ppm FC at 20 ppm CYA is probably a reasonable value. [EDIT] However, this earlier source implies a CT value of less than 0.2 and also seems to imply that some chlorinated cyanurate species may have bactericidal properties but only at higher concentrations (i.e. when the FC is higher). [END-EDIT]

Fortunately, the previous pathogens are not the common ones found in most pools. Easy-to-kill bacteria include Escherichia coli with a CT(2-log or 99%) of 0.04 for hypochlorous acid (so equivalent to a CT of 0.08 at a pH of 7.5) and heterotrophic bacteria with a CT(99%) of 0.08. And these CT values are at near freezing temperatures. Most viruses have CT(99.99% or 4-log) values around 2 though according to this WHO document, Polio may have CT(99% or 2-log) of 6 while Hepatitis may have CT(99%) of 10. Remember that viruses won't reproduce without a host so the issue is more of one for public pools with high bather loads than for residential pools. [EDIT] This CDC page gives higher CT numbers for some bacteria. Part of the difference for E. coli is that the CT of 5 is for 4-log or 99.99% kill, but theoretically doubling the 0.04 to 0.08 should give the same number, but clearly does not. [END-EDIT]

So let's calculate what the effective CT(50%) time is for chlorine levels found in SWG pools with 80 ppm CYA and 3 ppm FC, which is probably the lowest disinfection amount recommended on this forum. At a pH of 7.5, this is 0.015 ppm HOCl and is equivalent to 0.03 ppm FC. Taking the worst case of a bacterial generation rate of 15 minutes, this corresponds to a CT(50%) of 0.45 so generally speaking, such pools will inactivate microorganisms with CT(90% or 1-log) lower than 1.5, CT(99% or 2-log) lower than 3.0, CT(99.9% or 3-log) lower than 4.5 or CT(99.99% or 4-log) lower than 6.0. That means that easy-to-kill bacteria and most viruses will be inactivated, but the protozoa Giardia and Cyrptosporidium cysts will not nor will this be sufficient for spas. The high chlorine concentrations near the chlorine generation plate in salt cells are quite high, though contact time is very brief, so the issue is mostly about the general pool water, biofilms on pool surfaces, and the length of time it takes to circulate water through the salt cell.

I found a very interesting piece of information at this link which describes not only how chlorinated isocyanurates (that is, chlorine combined with CYA) are not effective in inactivating cysts, but that the presence of chlorinated isocyanurates may actually make the cysts less vulnerable to hypochlorous acid (HOCl). The effect was to increase the CT constant by about 50% which is not huge (by CT standards), but is quite interesting and unexpected. At a minimum, this article just confirms that the chlorinated isocyanurates are NOT disinfectants and that the focus should be on the hypochlorous acid (HOCl) concentration.

Richard

[fcfrey]

WOW

This is deeper than the diving end of my pool but this is good stuff. I have read this several times and am still digesting the info. With the other information and charts you have provided in the past we should find comfort in what we try to do. That being provide a safe sanitary pool.

Thanks Richard for your continued diligence and expertise.

By the way, still no ice on the pool here in Central PA. I have been throwing the robot in once a week just to keep it clean. My test results confirm my thoughts ---- A cold, clean pool doesn't need much treatment. Free CL is still in range 3.8 with CYA 35ppm and the last Bleach addition was 11-12-06.

[CarlD]

Not sure what y% is--tho x% is obviously 50% in this example.

I think the reason that the disinfection at higher temps (when the organisms are also reproducing the most) works better can be explained simply. The organisms' metabolism is faster at warmer temps--that's why the reproduction rate increases. That means they are also metabolizing the POISON faster too--after all, that's what our FC is, poisoning the organics--which is why the it's more effective.

But the idea that there's giardia and crypto in the pool, along with e-coli, makes my skin crawl--and pour in another gallon of bleach!

[chem geek]

Thanks for the explanation of the faster kill at higher temps. The higher temps make all chemical reactions go faster (due to higher energy so more reactants can get "over the hump" of an activated complex to form products), but even with faster uptake of the "poison" if the bugs were killed twice as fast AND reproduced twice as fast, then there would have been no net change. So it must be that the increased temps increase the uptake faster than the reproducion rate increases. That makes sense since the reproduction rate isn't based on a single simple chemical reaction, but on a complex series of them PLUS physical processes of moving organelles around as part of cell division.

In the formula, the "y%" is the CT value that you are converting FROM, say CT(99%) which is 2-log or CT(99.99%) which is 4-log. As you point out, I was developing a formula for how to get TO CT(50%) (which is "0.3-log" if anyone called it that). The purpose of the formula was just to convert from the "we most definitely have killed this bug since 99% of it is gone" CT value to the "we are on the cusp of just keeping this bug at bay, killing it about as fast as it reproduces" so that we could get an absolute bare minimum "critical" value. The CT values were developed mostly for water treatment where what was important was ensuring that some large percentage of bugs were killed since they were generally only exposing the water to high disinfectant levels for a limited period of time. That is different than our concern, especially in a residential pool, where we have continuous and long-term exposure from disinfectant and our goal is also to kill the bugs, but over a relatively much longer period of time.

At least with E.coli, it is killed very, very easily so we have nothing to worry about. But that Crypto, hey, now that's some hearty protozoa! Fortunately, the really hard-to-kill bugs that are of concern are ones that come from people, which is why they are an issue in public pools. In a residential pool, it's very unlikely to see such bugs unless you've already got the bug and are introducing it (I won't say how) into your own pool! Then, again, a paranoid pool owner can test their pool party invitees for Crypto (not...)!

Richard