ymclain said:
(No, I am not a dealer, just an owner and chemist.)
As a chemist, you will appreciate that you do not have any fast-acting sanitizer in the bulk pool water for quickly killing bacteria, inactivating viruses, etc. You also have to carefully manage the copper ion and pH levels to prevent staining.
You can read more about copper/silver ion systems and why they are not used in public/commercial pools on their own on the [EDIT]
Australian APVMA website (via Internet Archive) (current page
here uses a single standard instead of singling out the problems with copper ions) and on the
Health Canada website [END-EDIT]. There are no standalone copper or silver ion systems that pass the U.S.
EPA DIS/TSS-12 due to their slow kill times and all copper/ion systems certified by
NSF Standard 50 require a minimum of 0.4 ppm chlorine or 0.8 ppm bromine.
Metal ions do not kill pathogens quickly enough to be used (by themselves) in commercial/public pools since they may not be able to prevent person-to-person transmission of disease and do not always control bacterial growth (
E.coli, for example). When the FC is around 10% of the CYA level (roughly equivalent to 0.1 ppm FC with no CYA), most common heterotrophic bacteria are killed by chlorine in under one minute (for a 99% kill). It takes silver ion and copper ion far longer for equivalent kill. The following table shows kill times normalized to a 3-log reduction (99.9% kill) for various bacteria, viruses, and protozoan oocysts.
3-log reduction (99.9% kill) times in minutes for chlorine at 0.1 ppm FC with no CYA vs. copper at 0.4 ppm (400 ppb) vs. silver at 20 ppb
Pathogen ..................................... Chlorine ...... Copper ..... Silver
Bacteria (planktonic, not biofilms)
Escherichia coli ................................ < 1.2 ........ no effect* ..... 50 ..... *some studies show injury, but not death, over days
Pseudomonas aeruginosa .................... 1.5 ............ 58.5* ...... 225 ..... *no effect for some phenotypes
Stenotrophomonas maltophilia .............. ? .............. 52.5 ......... 42
Acinetobacter baumannii ................... 100 ............. 129 ........ 1770* ... *effectively no effect
Legionella pneumophila ....................... 60 .............. 12 ........ 1050* ... *effectively no effect
Enterococcus faecalis ......................... < 1 .......... no effect ...... ? ........ (formerly called Streptococcus faecalis)
Staphylococcus aureus ....................... < 1 .......... no effect ..... 225
Virus
Herpes Simplex Virus (HSV) ............. < 1800 ....... 22,500 ...... 5825
Vacciniavirus ................................... < 2500 ..... no effect ... no effect
Adenovirus ....................................... < 8.7 ............ ? ........ no effect
Vesicular stomatitis virus (VSV) ........ < 1500 ............ ? ........ no effect
Poliovirus .......................................... < 95 ....... > 5000 ..... no effect
Hemagglutinating Virus of Japan (HVJ) .... ? ............... ? ........ no effect
Coliphage MS-2 ................................... 1.8 ................ 130 (combo)
Influenza ............................................ 6.0 ............ 617 ......... ?
Protozoan oocyst
Naegleria gruberi ............................... 208.5 ........... ? ............. ?
Naegleria fowleri ................................. 425 ........ 17,000 .... 23,000 ... (data for chlorine is for cysts; for copper/silver it's amoeba in water)
Giardia intestinalis .............................. 232 ............. ? ............. ?
Cryptosporidium parvum .................. 153,000 .... > 4320 .......... ?
Sources (almost all are peer-reviewed scientific papers published in respected journals):
•
This table from the CDC was used for some of the chlorine CT values above (adjusted to a common 3-log 99.9% kill at 0.1 ppm FC with no CYA). Escherichia coli (0.647 is pH factor from 7.0 to 7.5) <0.25*(3/4)/(0.1*0.647) = 2.9 minutes, but we will use the WHO data from the next bullet point. Adenovirus 0.75*(3/4)/(0.1*0.647) = 8.7 minutes but at 5ºC temp so will be less at higher temp. Poliovirus 6.36*(3/4)/(0.1*0.5) = 95.4 but at 5ºC so will be less at higher temp. Giardia intestinalis 15*(3/3)/(0.1*0.647) = 232. Cryptosporidium parvum 15,300 (see source after the next one) so 15300*(3/3)/0.1 = 153,000.
• Table 3.1 in
this WHO document shows a 2-log (99%) reduction in Escherichia coli at pH 6.0 and 5ºC temperature with a CT value of 0.04 so 0.04*(3/2)/(0.1*0.5) = 1.2 so less than 1.2 minutes at higher temp.
•
This paper notes a CT value for Cryptosporiium parvum of up to 15,300 depending on source water and is the CT value the CDC now recommends using for a 3-log (99.9%) inactivation.
•
This paper stated "no inactivation of E. coli was observed after exposure to 0.4 or 0.8 mg/l cupric chloride after 60 min".
This paper showed a 3-log reduction in E-coli (wild-type strain) in 1 minute only above 500 mM (31,773 mg/L or ppm copper), but one cannot extrapolate here because E.coli cells have a mechanism for handling copper ions at a certain rate and therefore concentration (i.e. at low concentrations copper ions don't have any measurable kill effect) so I indicate "no effect".
This paper gives a MIC (minimal inhibitory concentration) of copper ions of 1.3 mM to 3.5 mM depending on strain, but even 1.3 mM is 82.6 ppm copper.
This study shows
E. coli injury after 2 days reaching 90-99% after 5-7 days, but this injury is not death of coliforms and rather causes an underestimation of coliform counts, but that these cells can recover and grow (basically, coliform counts using growth medium with sodium deoxycholate would kill more injured cells, but nonselective TLY agar would not and resulted in high coliform counts; that is, bacterial viability remained high at over 90% even after 7 days of exposure to copper).
This paper elucidates the mechanism for injury to decreased oxygen utilization, but again does not result in death of the cells and that they can recover and reproduce.
•
This paper for
Pseudomonas aeruginosa PAO1 showing a CT of 0.05 mg-min/liter for 90% kill with chlorine. This is a 3-log (99.9% kill) with 0.1 ppm chlorine of 0.05*(3/1)/0.1 = 1.5 minutes. For silver the CT was around 30 mg-min/liter for a 3-log (99.9% kill) so with 20 ppb (0.02 ppm) of 30/0.020 = 1500 minutes. This is higher than the 225 minutes in the paper below so the lower number was used for silver. For
E. coli ATCC 8739, they list a 2-log CT of 0.1 for chlorine and 28 for silver. This implies 3-log (99.9%) kill at 0.1 ppm chlorine and 20 ppb (0.020 ppm) silver of 1.5 and 2100 minutes, respectively, but is higher than the 225 minutes from another paper below so I used the lower number for silver. The E. coli number is higher than in all other studies so I used "< 1" in the table above, but did use the silver number.
This paper shows 3-log reduction of E. coli NBRC-3972 at 900 ppb silver ion in around 7 hours so at 20 ppb this is around (7*60)*900/20 = 18,900 minutes. I used the lower number of 2100 from the first paper.
•
This paper shows copper generally more effective than silver with copper ion 99.9% kill times at 0.4 ppm for
P. aeruginosa of 60*(0.39/0.4) = 58.5 minutes,
S. maltophilia of 60*(0.35/0.4) = 52.5 minutes,
A. baumannii of 60*(0.86/0.4) = 129 minutes,
L. pneumophila of 60*(0.08/0.4) = 12 minutes and silver ion 99.9% kill times at 20 ppb for
P. aeruginosa of 60*(0.075/0.02) = 225 minutes,
S. maltophilia of 60*(0.014/0.02) = 42 minutes,
A. baumannii of 60*(0.59/0.02) = 1770 minutes,
L. pneumophila of 60*(0.35/0.02) = 1050 minutes.
•
This paper for
Pseudomonas aeruginosa gives a MIC (minimal inhibitory concentration) for copper ions of 2 mM (127 ppm) copper though even lower 0.06 mM (3.8 ppm) copper showed only a significant lag of around 20 hours before growth occurred to normal levels by 30 hours. So the 58.5 minutes in the copper study above is really "no effect" in the long-term possibly due to selection for copper-resistant phenotypes, but I will keep the 58.5 minutes in the table. The MBC (minimum biocidal concentration) for an apparent complete kill in 5 hours in one growth medium (MSVP) was 0.01 mM (0.64 ppm) copper while in another (MOPSO) it was 0.125 mM (7.9 ppm) copper, but this is during the lag time where growth may subsequently resume after 20 hours from the copper-resistant phenotypes.
•
This paper shows a 3-log CT for
A. baumannii (and
M. oxydans) of 10 mg-min/L so with 0.1 ppm chlorine (and no CYA) this is 10/0.1 = 100 minutes.
•
This paper gives a 99% (2-log) kill of Legionella pneumophila with 0.1 mg/L FC at 21ºC and pH 7.6 in 40 minutes. So for a 3-log reduction this is 40*(3/2)*(0.1/0.1) = 60 minutes.
•
This paper shows 99.999% (5-log) kill of
Enterococcus faecalis (formerly called
Streptococcus faecalis) in 2 minutes with 0.51 ppm FC and 50 ppm CYA or 0.11 ppm FC with no CYA and of
Staphylococcus aureus in 5 minutes with 1.64 ppm FC and 50 ppm CYA or 0.64 ppm FC with no CYA. Conditions were 25ºC, pH 7.2, TA 50 ppm. The 0.5 ppm FC with 50 ppm CYA is equivalent to 0.01 ppm FC with no CYA at pH 7.5 which with 2 minutes is an implied CT of 0.02. The 1.64 ppm FC with 50 ppm is equivalent to 0.03 ppm FC with no CYA at pH 7.5 which with 5 minutes is an implied CT of 0.15. For 3-log kill this implies less than 1 minute at 0.1 ppm FC with no CYA for both bacteria. The directly calculated CT with no CYA is 0.11*2 = 0.22 for 5-log for
Enterococcus faecalis and 0.64*5 = 3.2 for 5-log for
Staphylococcus aureus so 3-log would be (3/5)*0.22 = 0.13 and (3/5)*3.2 = 1.9, respectively.
•
This paper showed the MIC (minimal inhibitory concentration) for copper-sensitive strains of
Enterococcus faecium of 4 mM (254 ppm copper) so 0.4 ppm copper would have no effect on this bacteria and it is presumed that this is also true for
Enterococcus faecalis.
•
This paper in Figure 2(a) shows 1.5 hours for a 3-log reduction of
Staphylococcus aureus with silver ion at 0.05 ppm (50 ppb) so this implies 1.5*60*(50/20) = 225 minutes while Figure 2(b) shows around 20 minutes for a 3-log reduction of
Escherichia coli at 0.05 ppm (50 ppb) so this implies 20*(50/20) = 50 minutes which is much lower than the 2100 minutes from another paper above so I use this lower amount.
•
This paper showed the MIC (minimal inhibitory concentration) for
Staphylococcus aureus of 200 µM (12.7 ppm) copper so 0.4 ppm copper would have no effect on this bacteria.
•
This paper shows that copper ions do a 90% inactivation of Herpes Simplex Virus in 30 minutes at 100-200 ppm (copper is normally < 0.5 in pools and I use 0.4 ppm in the table above). It is a big stretch to extrapolate, but this would give a 3-log (99.9%) inactivation at 0.4 ppm of 30*(100/0.4)*(3/1) = 22,500 minutes.
•
This paper shows that silver ions have virtually no effect on vacciniavirus, adenovirus, VSV, poliovirus, HVJ, but that with herpes simplex virus there is a 5-log kill in 60 minutes (roughly a 90% kill in about 5 minutes), but at over 3200 ppb (30 µM * 107.8682 g/mole) compared to the usual limit of 20 ppb to prevent silver staining. Extrapolated kill time at the lower level would be (30*107.8682/20)*60*(3/5) = 5825 minutes.
•
This paper shows no inactivation of vaccinia virus with copper alone at 5 µg/ml which is 5 mg/L (ppm).
•
This paper shows that Herpes Simplex Virus (HSV) has greater than a 2-log reduction in under 30 minutes with 4 ppm FC. Being conservative, I calculate 30*(4/0.1)*(3/2) = 1800 minutes though it is likely to be far lower.
•
This paper shows that a 0.0525% hypochlorite solution has a 1.8-log reduction of Vacciniavirus after 3 minutes. These strong chlorine solutions (around 535 ppm FC) are high in pH so the active chlorine level is only 1.7 to 27.9 ppm depending on the TA. If I am conservative and use a 50 ppm FC equivalent at pH 7.5, then a 3-log reduction with 0.1 ppm (with no CYA) is 3*(3/1.8)*(50/0.1) = 2500 minutes.
•
This paper indicates that VSV was inactivated in 10 minutes by 0.645% NaOCl. Graphs of inactivation of other chemicals in the study were at least 6-log reductions and a 0.645% NaOCl solution (6600 ppm FC) will be high in pH so the active chlorine level is only 1.6 to 17.1 depending on the TA. If I am conservative and use a 30 ppm FC equivalent at pH 7.5, then a 3-log reduction with 0.1 ppm (with no CYA) is 10*(3/6)*(30/0.1) = 1500 minutes.
•
This paper gives values for log
10 reductions per minute with 0.3 ppm chlorine vs. 400 µg/L (0.4 ppm) copper and 40 µg/L (40 ppb) silver for coliphage MS-2 and poliovirus type 1. For a 3-log (99.9%) reduction in coliphage, we have (3/4.88)*(0.3/0.1) = 1.8 minutes using 0.1 ppm chlorine, (3/0.023) = 130 minutes for the copper/silver combination. For a 3-log reduction in poliovirus, we have (3/0.036)*(0.3/0.1) = 250 minutes using 0.1 ppm chlorine, (3/0.0006) = 5000 minutes for copper/silver combination (so I designate this as "> 5000" for copper alone). I use these numbers for the copper values alone since silver alone was already tested in another study with no effect. For chlorine, I use the CDC data for 47.7 minutes instead of the 250 minutes in this study.
•
This paper gives CT values for a 3-log reduction of Influenza virus (H5N1) of 0.41 at pH 7 and 0.79 for pH 8 so I'll use 0.60 for a pH of 7.5. So for 0.1 ppm FC (with no CYA) this gives 0.60/0.1 = 6 minutes.
•
This paper showed that 25 µM copper sulfate (1.6 ppm copper ion) had a 3.5-log reduction in Influenza H9N2 virus in 3 hours (copper chloride took 6 hours for the same level of reduction). So at 0.4 ppm copper and a 3-log reduction this is (3/3.5)*(1.6/0.4)*(3*60) = 617 minutes.
•
This paper shows a mean CT product of 13.9 for Naegleria gruberi at pH 7 for 99% kill so for a 3-log (99.9%) kill with 0.1 ppm chlorine (with no CYA) this would be 13.9*(3/2)/0.1 = 208.5 minutes.
•
This paper showed that copper and silver alone, at ratio of 400:40 to 800:80 µg/l (ppb) caused no significant inactivation of N. fowleri even after 72 hours of exposure (k = log10 reduction/min = 0.00017 and 0.00013, respectively). These levels are similar to the 400:20 (0.4 ppm copper, 20 ppb silver) levels found in pools so the time for 3-log reduction is 3/0.00017 = 17647 minutes (so I use 17,000 in the table) for copper and 3/ 0.00013 = 23,000 minutes for silver.
•
This paper shows 3-log inactivation CT for Naegleria fowleri cysts of 42.5 so with 0.1 ppm chlorine (with no CYA) this would be 42.5/0.1 = 425 minutes.
•
This paper showed that with 0.25 to 3 mg/L (ppm) copper ions there was perhaps a 0.5-log reduction in Cryptosporidium parvum oocysts after 12 hours. So a 3-log reduction is > 12*(3/0.5)*60 = 4320 minutes.
Realistically, it's a spectrum of risk and you are much better off using a metal ion system than using nothing at all, but the risk is much higher than using one of the only three disinfectants registered by the EPA for use in pools: chlorine, bromine or Baquacil/biguanide/PHMB. Uncontrolled bacterial growth would be prevented when the 0.301-log (50%) reduction time was shorter than the generation time which is 15-60 minutes. This implies a 3-log reduction time shorter than approximately 150-600 minutes. This is why copper ions can prevent some runaway bacterial growth, but not for most fecal coliform bacteria. As shown in
this link, the three bacteria in the table for which copper ions have no effect (
Escherichia coli,
Enterococcus faecalis,
Staphylococcus aureus) are found in the lower G.I. tract (
S. aureus is also found in the nose and on skin) and are potential pathogens. It is also unlikely that any bacteria that can survive for an extended period of time in blood, such as the pathogenic
Leptospira that causes Leptospirosis (aka Weil's disease), will be able to be killed by copper ions in pools since blood serum contains 0.7 to 1.5 ppm copper ions (see
this link or
this link for normal copper levels of 70 - 150 µg/dL = 0.7 - 1.5 mg/L).
Ecosmarte Planet Friendly filed for
bankruptcy last year and emerged from chapter 11 earlier this year. We have written about some of the Ecosmarte claims in
this thread though they appear to have removed their largely bogus
science summary information since that time. [EDIT] The science summary is back and still wrong in many of its claims. [END-EDIT]
Richard