This thread is a response to this post some of which is quoted below.
The German DIN 19643 standard apparently uses a reference bacteria of Pseudomonas aeruginosa with a required 3-log (99.9%) kill time (one source says 4-log, but I think it's wrong; all the others say 3-log) in 30 seconds. It's hard to find a consistent CT (chlorine concentration in ppm times time in minutes) value, but I've seen it range in studies from 0.05 to 0.16 for a 99.9% kill time. If we use the high number, then this implies a chlorine concentration of 0.16*/0.5 = 0.32 ppm. Other sources refer to a 750 mV ORP level, but to achieve that at a pH of 7.5 requires about 0.6 ppm using a Chemtrol sensor or 1.1 ppm using an Oakton sensor. So something is inconsistent here and I suspect that the ORP level for the required kill rate is closer to 700 mV. The thing is that a 99% kill rate in 1-2 minutes for most bacteria is probably more reasonable to prevent person-to-person transmission and that corresponds more closely to 650 mV (with an Oakton sensor) and to 0.1 ppm FC at a pH of 7.5. This is roughly the target FC of around 10% of the CYA level.
I'm well aware of Dr. Bernard's studies as well as many others all showing respiratory and ocular problems associated with indoor pools and most of these due to nitrogen trichloride (the most irritating and volatile of chloramines). There is also concern for trihalomethanes (THMs) such as chloroform, but that's more of a concern for long-term health effects such as cancer. I've written to Dr. Bernard (and others) about the chlorine/CYA relationship but I usually don't get a response (sometimes I do).
The production of nitrogen trichloride from the breakpoint of ammonia is modeled by the Jafvert & Valentine (1992) model. Earlier models from Wei & Morris and Selleck & Saunier come to similar conclusions, but have different absolute numbers for concentrations and the Jafvert & Valentine model is the newest and matches experimental data the best. I have a spreadsheet with these models and they show the following when there is much more chlorine than ammonia to oxidize (which is the normal situation one wants to have in a pool; otherwise a higher FC level should be used):
1) The rate of formation of monochloramine is proportional to the active chlorine concentration, but it's very fast regardless, in under a minute even at relatively low active chlorine concentrations.
2) The peak concentration of dichloramine and of nitrogen trichloride is roughly proportional to the active chlorine concentration and increases greatly at lower pH.
3) The final concentration of nitrogen trichloride is roughly proportional to the active chlorine concentration and increases greatly at lower pH. Monochloramine and dichloramine oxidize towards zero, but nitrogen trichloride stays around (doesn't further oxidize quickly) and is very volatile so tends to outgas.
On the other hand, high pH favors chloroform production if precursor molecules are present (chloroform does not come from ammonia; it comes from carbon-based organics).
The open question is what happens with urea as no clear model has been developed for chlorine oxidation of urea. Wojtowicz has speculated about some reactions and these produce both nitrogen trichloride directly as well as producing dichloramine that would then follow the breakpoint reaction with the characteristics noted above. Since urea is the primary component of sweat and urine (ammonia is the second largest component), knowing what happens with urea is very important so it's amazing that very little research has been done in this area. As I've noted in other posts, this article shows how Chip Blatchley of Purdue has a grant from NSPF to look at disinfection by-products in pools and part of this work is to develop a model for the chlorine oxidation of urea. I am very much looking forward to the results of that work.
A researcher named Samples in 1959 proposed the following mechanism to explain the products that he observed.
Cl2NCONCl2 + HOCl ---> NCl3 + HCl + CO2 + NCl
Quadchloro Urea + Hypochlorous Acid ---> Nitrogen Trichloride + Hydrochloric Acid + Carbon Dioxide + intermediate
NCl3 + HOCl + 2H2O ---> HNO3 + 4HCl
Nitrogen Trichloride + Hypochlorous Acid + Water ---> Nitric Acid + Hydrochloric Acid
NCl + OH- ---> NOH + Cl-
intermediate + Hydroxyl Ion ---> intermediate + Chloride Ion
2NOH ---> H2N2O2 ---> N2O + H2O
intermediate ---> intermediate ---> Nitrous Oxide + Water
Wojtowicz proposed another mechanism as follows:
Cl2NCONCl2 + HOCl ---> NCl3 + NHCl2 + CO2
Quadchloro Urea + Hypochlorous Acid ---> Nitrogen Trichloride + Dichloramine + Carbon Dioxide
He then describes further reactions that are consistent with the Jafvert & Valentine model. He notes that chlorine and urea do not form (or measure as) Combined Chlorine (CC) and he also shows that the oxidation of urea by chlorine is fairly slow taking many hours to days depending on concentrations. This implies that the reaction rate of urea combining with chlorine to form quadchloro urea may be slow but that the above reaction may be fast or that the quadchloro urea (and other intermediate chlorourea) don't measure in the CC test. In any event, you can see that dichloramine is produced so would then follow the breakpoint chlorination model, but nitrogen trichloride is also directly produced so that part is an unknown as to reaction rates and concentrations. We don't yet know enough about which reaction steps are rate-limiting so we don't know if urea will show the same sort of chlorine dependence on nitrogen trichloride production. So half of the nitrogen trichloride will show such dependence (due to the dichloramine) but the other half is unknown. However, with a linear dependence on active chlorine concentration, this would be a linear result of amount of nitrogen trichloride per time period though the total cumulative concentration would be dependent on the amount of urea (again, for half of the nitrogen trichloride concentration). This is why I think that even with urea, lower active chlorine levels would be better and hence CYA in indoor pools would lower the rate of production and total amounts of nitrogen trichloride.
Richard
My spreadsheet (here) does not have the HOCl turn red except when it is low. It stays green at high levels as there is no single number indicating safety, unlike the rough limits for certain bacteria kill rate goals or algae inhibition. So you can't go by any color coding and generally just want to use the minimum amount of active chlorine necessary for your purposes, which usually means the lowest amount needed to prevent algae growth since that requires higher active chlorine than killing most pathogens.smallpooldad said:ChemGeek,
The HOCl (as ppm Cl2) is still in the green zone so presumably safe and if maintained little or no algae, assuming a SWG system or gas injection system, or am I wrong?
:
:
The other cultural consideration is that Europeans do not bathe as often as persons living in the US so having a high ORP of 750 might be beneficial to other swimmers in the pool in that it helps to avoid cross infection. Although strangely, generally speaking, Germans bathe much more often than many other Europeans.
:
:
One thing got me concerned and that was your mention of nitrogen trichloride (trichloramine), based on this I would be vary wary of sending young children to an indoor pool, especially one that does not have good modern ventilation, or is public. See here:
http://en.wikipedia.org/wiki/Pool_chlorine_hypothesis
The German DIN 19643 standard apparently uses a reference bacteria of Pseudomonas aeruginosa with a required 3-log (99.9%) kill time (one source says 4-log, but I think it's wrong; all the others say 3-log) in 30 seconds. It's hard to find a consistent CT (chlorine concentration in ppm times time in minutes) value, but I've seen it range in studies from 0.05 to 0.16 for a 99.9% kill time. If we use the high number, then this implies a chlorine concentration of 0.16*/0.5 = 0.32 ppm. Other sources refer to a 750 mV ORP level, but to achieve that at a pH of 7.5 requires about 0.6 ppm using a Chemtrol sensor or 1.1 ppm using an Oakton sensor. So something is inconsistent here and I suspect that the ORP level for the required kill rate is closer to 700 mV. The thing is that a 99% kill rate in 1-2 minutes for most bacteria is probably more reasonable to prevent person-to-person transmission and that corresponds more closely to 650 mV (with an Oakton sensor) and to 0.1 ppm FC at a pH of 7.5. This is roughly the target FC of around 10% of the CYA level.
I'm well aware of Dr. Bernard's studies as well as many others all showing respiratory and ocular problems associated with indoor pools and most of these due to nitrogen trichloride (the most irritating and volatile of chloramines). There is also concern for trihalomethanes (THMs) such as chloroform, but that's more of a concern for long-term health effects such as cancer. I've written to Dr. Bernard (and others) about the chlorine/CYA relationship but I usually don't get a response (sometimes I do).
The production of nitrogen trichloride from the breakpoint of ammonia is modeled by the Jafvert & Valentine (1992) model. Earlier models from Wei & Morris and Selleck & Saunier come to similar conclusions, but have different absolute numbers for concentrations and the Jafvert & Valentine model is the newest and matches experimental data the best. I have a spreadsheet with these models and they show the following when there is much more chlorine than ammonia to oxidize (which is the normal situation one wants to have in a pool; otherwise a higher FC level should be used):
1) The rate of formation of monochloramine is proportional to the active chlorine concentration, but it's very fast regardless, in under a minute even at relatively low active chlorine concentrations.
2) The peak concentration of dichloramine and of nitrogen trichloride is roughly proportional to the active chlorine concentration and increases greatly at lower pH.
3) The final concentration of nitrogen trichloride is roughly proportional to the active chlorine concentration and increases greatly at lower pH. Monochloramine and dichloramine oxidize towards zero, but nitrogen trichloride stays around (doesn't further oxidize quickly) and is very volatile so tends to outgas.
On the other hand, high pH favors chloroform production if precursor molecules are present (chloroform does not come from ammonia; it comes from carbon-based organics).
The open question is what happens with urea as no clear model has been developed for chlorine oxidation of urea. Wojtowicz has speculated about some reactions and these produce both nitrogen trichloride directly as well as producing dichloramine that would then follow the breakpoint reaction with the characteristics noted above. Since urea is the primary component of sweat and urine (ammonia is the second largest component), knowing what happens with urea is very important so it's amazing that very little research has been done in this area. As I've noted in other posts, this article shows how Chip Blatchley of Purdue has a grant from NSPF to look at disinfection by-products in pools and part of this work is to develop a model for the chlorine oxidation of urea. I am very much looking forward to the results of that work.
A researcher named Samples in 1959 proposed the following mechanism to explain the products that he observed.
Cl2NCONCl2 + HOCl ---> NCl3 + HCl + CO2 + NCl
Quadchloro Urea + Hypochlorous Acid ---> Nitrogen Trichloride + Hydrochloric Acid + Carbon Dioxide + intermediate
NCl3 + HOCl + 2H2O ---> HNO3 + 4HCl
Nitrogen Trichloride + Hypochlorous Acid + Water ---> Nitric Acid + Hydrochloric Acid
NCl + OH- ---> NOH + Cl-
intermediate + Hydroxyl Ion ---> intermediate + Chloride Ion
2NOH ---> H2N2O2 ---> N2O + H2O
intermediate ---> intermediate ---> Nitrous Oxide + Water
Wojtowicz proposed another mechanism as follows:
Cl2NCONCl2 + HOCl ---> NCl3 + NHCl2 + CO2
Quadchloro Urea + Hypochlorous Acid ---> Nitrogen Trichloride + Dichloramine + Carbon Dioxide
He then describes further reactions that are consistent with the Jafvert & Valentine model. He notes that chlorine and urea do not form (or measure as) Combined Chlorine (CC) and he also shows that the oxidation of urea by chlorine is fairly slow taking many hours to days depending on concentrations. This implies that the reaction rate of urea combining with chlorine to form quadchloro urea may be slow but that the above reaction may be fast or that the quadchloro urea (and other intermediate chlorourea) don't measure in the CC test. In any event, you can see that dichloramine is produced so would then follow the breakpoint chlorination model, but nitrogen trichloride is also directly produced so that part is an unknown as to reaction rates and concentrations. We don't yet know enough about which reaction steps are rate-limiting so we don't know if urea will show the same sort of chlorine dependence on nitrogen trichloride production. So half of the nitrogen trichloride will show such dependence (due to the dichloramine) but the other half is unknown. However, with a linear dependence on active chlorine concentration, this would be a linear result of amount of nitrogen trichloride per time period though the total cumulative concentration would be dependent on the amount of urea (again, for half of the nitrogen trichloride concentration). This is why I think that even with urea, lower active chlorine levels would be better and hence CYA in indoor pools would lower the rate of production and total amounts of nitrogen trichloride.
Richard