Chlorine Causes Cancer . . . or Not!

[chem geek]

There were a series of scientific papers released about the same time and a flurry of media reports on them. I've written a post about it here, but as Ben writes above, the pools in the study are very different than typical residential pools because they were indoor pools (so no UV from sunlight and relatively poor air circulation), were public pools with high bather load, and likely did not use Cyanuric Acid (CYA) in the pools.

Virtually every scientific study done investigating Disinfection By-Products (DBPs) has found that they are roughly proportional to bather load (see Fig. 1 in this paper as one of many examples). Unfortunately, the investigators did not identify the specific bather load of the pools used, but they were likely to be high since they were public pools and had high DBP levels including Combined Chlorine (CC).

The chlorinated pool in the mutagenecity study was 33 x 25 x 2 meters in size so around 436,000 gallons with a Free Chlorine (FC) level of 1.28 mg/L, monochloramine of 0.29 mg/L and dichloramine (that is likely to really be chlorourea) of 0.38 mg/L (so combined chlorine of 0.67 mg/L). The THMs in breath study used an indoor pool that was 25 meters long with the pool located in a sports center with an average Free Chlorine (FC) level of 1.17 mg/L, dichloramine of 0.43 mg/L (monochloramine wasn't measured but was implied in the study to be roughly the same level). The genotoxicity study used an indoor 25 meter pool with an average Free Chlorine (FC) level of 1.17 mg/L. This implies that this study used the same pool as the THMs in breath study.

I have a Breakpoint Chlorination spreadsheet where a steady-state of 0.3 mg/L monochloramine with 1.3 mg/L FC (with no CYA) can be achieved with around 0.7 mg/L Nitrogen per hour from ammonia alone in sweat and urine (if I run the model with urea assumptions, the implied bather load is even higher). Typical assumptions for the amount of sweat and urine in swimming result in around 800 mg Nitrogen per bather per hour. So this implies around 1140 liters per bather or around 300 gallons per bather. This is unlikely (seems too high) so there are probably other sources of nitrogenous compounds (or people were urinating in the pool), but it still indicates that the bather load was probably very high in these pools -- at least 5-10 times higher than in residential pools.

In the mutagenicity paper, it was noted: "We found a greater number of DBPs in the chlorinated and brominated indoor pools studied here than were found in chlorinated outdoor pools (Zwiener et al. 2007), which was not surprising, considering that DBPs can be volatilized or photolyzed (Lekkas and Nikolaou 2004) in outdoor settings." This is an important point. The greater air circulation and the exposure to ultraviolet (UV) rays of sunlight in outdoor pools significantly reduce the amount of DBPs, probably by at least a factor of 2-5.

Finally, the studied pools very likely had no Cyanuric Acid (CYA) in the water so the active chlorine level was far higher than in pools with CYA. The pools in the study had an FC of around 1.2 ppm while pools at roughly the "target" level in Ben's Best Guess CYA chart have the same active chlorine (hypochlorous acid) level as a pool with roughly 0.1 ppm FC and no CYA. So the pools in the study had over 10 times the active chlorine level of residential pools that use CYA which would have every (elementary) reaction that chlorine participates in be over 10 times faster. The breakpoint chlorination models all demonstrate that this higher active chlorine level leads to a greater rate of production AND total amount of very irritating and volatile nitrogen trichloride (this is also described in this article).

The bottom line is that outdoor residential pools are much, much safer in terms of having much lower amounts of disinfection by-products (DBPs). For commercial/public high bather load pools, I recommend that they use a small amount of CYA to lower the active chlorine concentration, so around 4 ppm FC with 20 ppm CYA to achieve roughly the equivalent of 0.2 ppm FC with no CYA, and that supplemental oxidation (UV, ozone, non-chlorine shock, enzymes) be used. These two work together since the lower active chlorine level gives precursor molecules more time to be exposed to the supplemental oxidizers and also gives them more time to volatize before reacting with chlorine.

Richard