Implementation of Ultraviolet / H₂O₂ Treatment for inactivation of microorganisms and pesticide control

In 1920, when N.V. PWN Water Supply Company North Holland (PWN) was founded, the demand for drinking water was satisfied by ground water extraction . However by the growing drinking water demand PWN was compelled to utilize surface water as an additional source. Therefore in the 1960's water treatment plant Andijk was constructed for the direct production of drinking water from IJssel Lake (River Rhine) water. Originally the plant consisted of microstraining, breakpoint chlorination, coagulation, sedimentation, rapid filtration and post disinfection. In 1978 the plant was upgraded with a pseudo moving bed GAC filtration. After about 40 years of operation, wtp Andijk still complies with all Dutch drinking water standards. Nevertheless an upgrade is desired in view of by-product (THM) formation and the barriers against pathogenic micro-organisms such as protozoa and organic micropollutants such as pesticides. PWN has pursued the application of advanced oxidation for both primary disinfection and organic contaminant control followed by the already present GAC filtration for removal of residual H₂O₂ and AOC with very promising results. Ultraviolet / H₂O₂ will be installed and become operational mid 2004.

PROCESS SELECTION

PWN's aim was to select an oxidative process able to inactivate pathogenic micro-organisms and to degrade pesticides aselectively with a restricted by-product formation compared to breakpoint chlorination. Initially the focus was on ozone based technologies. (1) Ozone was able to achieve the required disinfection capacity. However ozone proved to be a selective barrier against pesticides. In advanced oxidation technologies hydroxyl radicals react 106 to 109 times faster than ozone, causing a much more aselective pesticide degradation. O₃/H₂O₂ treatment proved to be a sound barrier against pesticides. In pretreated IJssel Lake water 80 % pesticide degradation could be achieved by an O₃ dose of 2.8 g/m³ and a H₂O₂/O₃ ratio of 2 g/g. (2) However bromate was found at high levels especially at low water temperature. (3) Feasible treatment options (increase H₂O₂/O₃ ratio, increase pH) could not control bromate formation completely caused by mass transfer limitations. (see figure 1) (3)

Under optimal conditions for pesticide and bromate control disinfecting properties of the O₃/H₂O₂ process were very poor. Corresponding with high H₂O₂/O₃ ratios and very low c.t. values for ozone, inactivation of spores of sulphite reducing clostridia was insignificant. (4) In view of these findings PWN rejected O₃/H₂O₂ treatment for primary disinfection and pesticide control at wtp Andijk and decided to pursue alternative strategies with the focus on Ultraviolet technologies.

PHOTOLYSIS AND ADVANCED OXIDATION BY ULTRAVIOLET / H₂O₂ TREATMENT

Ultraviolet photolysis is based on the absorption of Ultraviolet photons by contaminants to cause their degradation into smaller products. The absorbed energy must be higher than the energy of the band to be broken to cause photolysis.
In addition to the absorption of Ultraviolet photons the quantum yield of the degradation is an important parameter. Compounds with a high Ultraviolet absorbance and a high quantum yield are very susceptible to photodegradation. (5)
However not all substances absorb Ultraviolet light significantly or have a high quantum yield. In order to promote a more aselective degradation H₂O₂ may be added to produce hydroxyl radicals.
The general expression for the degradation rate of a contaminant is a combination of direct Ultraviolet photolysis and hydroxyl radical reactions:

Not too many and k_C,OH values for pesticide degradation are given in literature. An electric energy of 0.5 kWh/m³ is reported for 60 % atrazine photolysis. Under these conditions pesticide degradation varied from 18 % for dicamba to 92 % for mecoprop. (6) Additional research is needed to determine quantum yield and hydroxyl radical rate constants for representative organic micropollutants and to determine the process conditions for Ultraviolet / H₂O₂ treatment with medium pressure Ultraviolet lamps.

Disinfection by Ultraviolet

There is an abundance of literature on the inactivation of fecal coliform bacteria by Ultraviolet radiation. It has been known for some time that viruses and spore forming bacteria are more resistant to the effect of Ultraviolet radiation than coliform bacteria. Chang et al (7) showed that viruses, bacterial spores and amoebic cysts required 3 – 4 times, 9 times and 15 times more Ultraviolet exposure to achieve the same level of inactivation as Escherichia coli. It had been thought that encysted protozoa were insensitive to Ultraviolet.(8,9) More recent studies have shown that encysted Cryptosporidium parvum and Giardia muris are relatively sensitive to Ultraviolet radiation. (10, 11, 12) Reactivation of pathogenic micro-organisms is a critical parameter with the respect to the treatment of drinking water. It is well known that select pathogenic micro-organisms are capable of reactivating after exposure to Ultraviolet light. Mechsner et al (13) demonstrated that E coli inactivated by low pressure Ultraviolet repaired its DNA after several days incubation in the dark. More work is needed to determine the overall reactivation potential of water-borne pathogenic microorganisms of concern to the water industry. The ultimate goal for Ultraviolet disinfection of drinking water is to damage DNA beyond repair.

Additional research is needed to establish dose-response curves for several pathogenic micro-organisms and to establish the Ultraviolet dosage where DNA is damaged beyond repair.

OBJECTIVES

Disinfection

Three major objectives were pursued in a collaborative study carried out by the University of Alberta and PWN:
- to select representative organisms and characterize Ultraviolet inactivation by generating Ultraviolet doseinactivation curves for MS2 phages, Bacillus subtilis endospores, Giardia muris cysts and Cryptosporidium parvum oocysts using a bench scale collimated beam apparatus;
- to determine the ability of one small medium pressure Ultraviolet reactor and three larger medium pressure Ultraviolet reactors to inactivate MS2 phages, Bacillus subtilis endospores, Giardia muris cysts and Cryptosporidium parvum oocysts;
- to examine the reactivation of Giardia muris and Cryptosporidium parvum ex-vivo and in vivo and to determine more precisely the Ultraviolet dose required for inactivation of these protozoan parasites in drinking water.

Organic contaminant control

Three additional major objectives were pursued by PWN, partly in collaboration with Trojan Technologies Inc.:
- to select representative organic micropollutants (pesticides) and characterize degradation by Ultraviolet photolysis and hydroxyl radical reaction for 12 selected pesticides;
- to predict and determine the ability of a medium pressure Ultraviolet reactor to degrade those 12 priority pollutants;
- to design a full scale Ultraviolet / H₂O₂ system for both disinfection and organic contaminant control.

Posttreatment

Two additional objectives were pursued by PWN to integrate Ultraviolet / H₂O₂ treatment in the total treatment scheme:
- the removal of excess H₂O₂ by GAC filtration;
- the removal of AOC by GAC filtration.

INACTIVATION OF MICRO-ORGANISMS

Bench scale experiments.

The bacteriophage MS2 has been shown to be relatively resistant to Ultraviolet inactivation. A reduction of approximately 1 log-unit was achieved when MS2 phages, suspended in PWN water, was exposed to Ultraviolet doses of 20 – 25 mJ/cm². (Figure 2)

The correlation between log inactivation and effective Ultraviolet dose demonstrated a near linear relationship, particularly within the Ultraviolet dose range of up to 50 mJ/cm². This contrasts the Ultraviolet inactivation curves obtained for Bacillus subtilis (Figure 3), Giardia muris (Figure 4) and Cryptosporidium parvum (Figure 5) where distinct shouldering and tailing effects were observed. Figure 3 summarizes the results to determine the effect of Ultraviolet inactivation on Bacillus subtilis endospores suspended in PWN water. Similar to data reported by Sommer and Cabaj (14) a distinct shouldering effect was observed when endospores were exposed to Ultraviolet doses less than 20 mJ/cm².

However, unlike the data presented by Sommer and Cabaj a pronounced tailing effect was observed. Sommer and Cabaj reported a log reduction of approximately 4 log units with Ultraviolet doses of 60 mJ/cm². In this study 3 log and 4 log reduction were achieved at about 60 and 120 mJ/cm² respectively. Giardia muris cysts suspended in PWN water were shown to be relatively susceptible to Ultraviolet inactivation. An Ultraviolet dose of as low as 20 mJ/cm² induced greater than 2 log-unit inactivation of Giardia muris. However a pronounced tailing effect in the ability of Ultraviolet to inactivate the infective potential of Giardia muris cysts was observed with Ultraviolet doses greater than 20 mJ/cm². (Figure 4).

Exposure of Giardia muris cysts to Ultraviolet doses as high as 120 mJ/cm² resulted in only 1 additional log-unit of inactivation compared to Ultraviolet doses of 20 mJ/cm² only. Cryptosporidium parvum oocysts suspended in PWN water were extremely susceptible to low doses of Ultraviolet. Greater than a 3 log-unit reduction in oocyst infectivity was observed in trials where oocysts were exposed to Ultraviolet doses as low as 20 mJ/cm². For Cryptosporidium parvum oocysts the tailing effect of the Ultraviolet dose-inactivation curve was not as pronounced as that observed for Giardia muris cysts (Figure 5). Ultraviolet doses of approximately 120 mJ/cm² induced greater than 4.5 log-units of inactivation.

However, the presence of the tailing suggests that a Ultraviolet resistant sub-population of oocysts may be present within any given batch of oocysts, which remains to be confirmed using molecular analysis of the resistant sub-population of the (oo)cysts.

Pilot plant experiments

The collimated beam data demonstrate that medium pressure Ultraviolet radiation effectively inactivated MS2 phages, Bacillus subtilis endospores, Giardia muris cysts and Cryptosporidium parvum oocysts suspended in pretreated IJssel Lake water. The results were confirmed by data obtained with a small scale Berson Ultraviolet reactor and with three larger scale reactors from Calgon, Berson and Trojan. Of the three larger scale Ultraviolet reactors the Berson and Trojan reactors showed inactivation levels equivalent to the collimated beam experiments, while the Calgon unit showed a lower inactivation level (Figure 6).

The findings show that Ultraviolet disinfection is a very reliable technology for the inactivation of pathogenic micro-organisms in drinking water. A Ultraviolet dose of 105 mJ/cm² gives a high inactivation of phages and endospores and a complete inactivation of protozoan cysts and oocysts.

REACTIVATION OF MICRO-ORGANISMS

It is important to understand that it is debatable as to whether or not Ultraviolet-irradiation causes permanent or temporary inactivation of micro-organisms. Relatively little is known about the repair of Ultraviolet-damaged DNA in encysted water-borne protozoan pathogens. Our data indicated that neither Giardia muris cysts nor Cryptosporidium parvum oocysts are capable of in vitro reactivation. The results for in vivo reactivation of Giardia muris cysts are summarized in table 1.

From the data it can be concluded that Giardia muris cysts are capable of in vivo reactivation after Ultraviolet treatment when the Ultraviolet dose applied is as low as 25 mJ/cm². This dose is significantly lower than the Ultraviolet dose of 105 mJ/cm² needed for a 3.5 log-unit inactivation of MS2 phages. For Cryptosporidium parvum oocysts the in vivo repair could not be tested due to the "end-point" animal model.

DEGRADATION OF ORGANIC MICROPOLLUTANTS

Bench Scale Experiments in Pretreated IJssel Lake Water
Four test runs were carried out with pretreated IJssel Lake water for a first optimization of electric energy input and H₂O₂-dosage. Atrazine degradation is presented in figure 7.

The EE/O was ranging from 0.65 – 1.30 kWh/m³ depending on the H2O2-dose (see Table 2).
Under no reaction conditions bromate formation was found.
Because the process was considered as cost effective pilot plant research was carried out.

Pilot Plant Experiments Phase 1
Four test runs were carried out with new Ultraviolet lamps under conditions corresponding with the bench scale experiments.
Atrazine conversion in bench and pilot plant scale testing was in the same order of magnitude.
(see Table 2)

The range of EE/O values for pilot testing was only slightly higher than for bench scale testing:
0.70 – 1.50 kWh/m³ and 0.65 – 1.30 kWh/m³ respectively.

Two conditions being 0.6 kWh/m³ + 15 g/m³ H₂O₂ and 0.9 kWh/m³ + 4 g/m³ H₂O₂ were chosen for further research with five selected priority pollutants: atrazine, pyrazone, diuron, bentazone and bromacil. The experiments were carried out with new lamps and after a lamp life of 2000 h.. The results for 0.9 kWh/m³ + 4 g/m³ H₂O₂ are shown in figure 8.

For new lamps pyrazone and diuron were degraded slightly higher than atrazine, while bentazone and bromacil conversion were slightly lower.
After 2000 burning hours pesticide degradation dropped significantly and amounted roughly 50 % of the original degradation. Lamp control showed that the Ultraviolet-output of the lamps was decreased by 23 – 68 % showing that, in addition to electric energy data, measurement of Ultraviolet-dose is urgently needed.
Irrespective of reaction conditions no bromate was found.
Ultraviolet extinction dropped with 20 %, DOC-removal was insignificant. AOC increased from 10 µg/l to 110 – 140 µg/l. Therefore biological stability after Ultraviolet / H₂O₂-treatment is an important issue.
Ultraviolet / H₂O₂-treatment enables 80 % pesticide degradation without a significant formation of primary pesticide metabolites and bromate. Until now the experiments were focused on the combined effect of Ultraviolet-photolysis and hydroxyl radical oxidation. In the second phase of the pilot plant research the effect of Ultraviolet-photolysis only followed by the additional effect of advanced oxidation was studied.

Pilot Plant Experiments Phase 2

Degradation of 10 pesticides atrazine, pyrazon, diuron, bentazone, bromacil, methabenzthiaxuon, dicamba, 2,4-D, TCA, trichlorpyr and the atrazine metabolite desethyldesisopropylatrazine by Ultravioletphotolysis was studied. The electric energy was ranging from 0.25 – 2.0 kWh/m³ (see Figure 9).

Figure 9: Pesticide Degradation by Ultraviolet-Photolysis as a Function of the Ultraviolet-Dose

All priority pollutants showed a significant degradation by Ultraviolet-photolysis. The conversion for an electric energy of 1 kWh/m³ is summarized in table 3.

By Ultraviolet-photolysis with 1 kWh/m³ degradation ranged from 18 % for trichloroacetic acid (TCA) to 70 % for atrazine. The criterion of 80 % conversion had to be achieved by an additional H₂O₂-dosage to initiate a supplementary hydroxyl radical reaction. Examples for a compound with a low and a high Ultravioletphotolysis conversion are shown in figure 10 and 11.

Degradation of atrazine by an electric energy of 1 kWh/m³ amounted 70 %. This degradation was increased to the desired 80 % by adding 13 g/m³ H₂O₂. Atrazine degradation was caused primarily by photolysis with a polishing effect of hydroxyl radical oxidation.
Degradation of bromacil by an electric energy of 1 kWh/m³ amounted 42 %. This degradation was increased to the desired 80 % by adding 15 g/m³ H₂O₂. For bromacil degradation hydroxyl radical oxidation played a much more important part.
For most priority pollutants a degradation of 80 % can be achieved. Depending on Ultraviolet-absorbance and quantum yield on the one side and chemical structure (double bonds, H-atoms) on the other side photolysis or hydroxyl radical reactions will play a predominant part. Only for TCA and desethyldesisopropylatrazine 80 % degradation may not be achieved.

PERSPECTIVE

At water treatment plant Andijk Ultraviolet / H₂O₂-treatment will be implemented. For primary disinfection the breakpoint chlorination will be replaced by Ultraviolet-disinfection. A dose of 105 mJ/cm² gives a high inactivation of phages (viruses) and spores and an complete inactivation of Giardia muris and Cryptosporidium oocysts. Although some reactivation of Giardia muris cysts is observed, this will not take place at Ultraviolet-doses as high as 105 mJ/cm². Ultraviolet / H₂O₂-treatment will be implemented for organic contaminant control prior to the two step GACfiltration already present. A pesticide degradation of 80 % can be achieved for a broad range of process conditions (i.e. from 0.4 kWh/m³ and 15 g/m³ H₂O₂ to 0.9 kWh/m³ and 4 g/m³ H₂O₂). This Ultraviolet dose is much higher than needed for disinfection. Under these conditions bromate formation is completely absent, while primary metabolite formation is insignificant. PWN decided to pursue Ultraviolet / H₂O₂-treatment for full scale application. Important issues were removal of residual H₂O₂ and biological stability. Removal of H₂O₂ by GAC-filtration is presented in Figure 12.

Removal of 15 g/m³ proved to be complete within 5 minutes. After GAC-filtration with an empty bed contact time of 40 minutes the biofilm formation rate remained very low (~ 2 pg/cm².d) during the total filter run, although AOC increased from 10 to 40 µg Ac C eq/l. Based on these results PWN will upgrade water treatment plant Andijk to the following configuration (Figure 13).

Figure 13: Projected Water Treatment Plant Andijk

The combination Ultraviolet / H₂O₂-treatment / GAC-filtration provides a very reliable and flexible process for:
• removal / inactivation of micro-organisms;
• organic contaminant control.

ACKNOWLEDGEMENT

The authors thank P. Hoogzaad, R. Kooijman (PWN), D.W. Smith, N.F. Neumann, S. Craik (University of Alberta), M.I. Stefan (Trojan Technologies Inc.) and J.R. Bolton (Bolton Enterprises) for their stimulating discussions and their assisting in realizing this study.

REFERENCES

1 Kruithof, J.C., Kamp, P.C.: Proceedings IOA/EA3G Conference, Moscow p. 331 – 348 (1998)
2 Hoigne, J.: Science of the Total Environment 47, 169 – 185 (1985)
3 Kruithof, J.C., Kamp, Meijers, R.T.: Proceedings 14th Ozone World Congress, Dearborn (1999)
4 Kruithof, J.C., Kamp, P.C.: Proceedings AWWA Annual Conference, Chicago (1999)
5 Sonntag, C. von, Mark, G. Mertens, R., Schuchmann, M.N., Schuchmann, H.P.: J Water SRT –
Aqua 43, 201 – 211 (1993)
6 Bourgine, T.P., Chapman, J.I.: Proceedings IOA/EA3G Conference, Amsterdam (1996)
7 Chang, J.C.H., Ossoff, S.F., Lobe, D.C., Dorfman, M.H., Dumais, C.M., Qualls, R.G.,
Johnson, J.D.: Applied and Environmental Microbiology 49, 1361 – 1365 (1985)
8 Ransome, M.E., Whitmore, T.N., Carrington, E.G.: Water Supply 11, 75 – 89 (1993)
9 Rice, E.W., Hoff, J.C.: Applied and Environmental Microbiology 42, 546 – 547 (1981)
10 Bukhari, Z., Hargy, T.M., Bolton, J.R., Dussert, B., Clancy, J.L.: Journal of the American Water
Works Association, 91, 86 – 94 (1999)
11 Clancy, J.L., Hargy, T.M., Marshall, M.M., Dyksen, J.E.: Journal of the American Water Works
Association 90, (92 – 102 (1998)
12 Craik, S.A., Finch, G.R., Bolton, J.R., Belosevic, M.: Water Research (2000)
13 Mechsner, K., Fleischmann, T., Mason, C.A., Hamer, G.: Water Science and Technology 24,
339 – 342 (1991)
14 Sommer,R., Cabaj, A.: Water Science and Technology 27, 357 – 362 (1993)