|
|
|
Ozone and Water Treatment Rolf A. Deininger School of Public Health- The University of Michigan- Ann Arbor, Michigan, USA Janice Skadsen and Larry Sanford Ann Arbor Water Treatment Plant 919 Sunset Road - Ann Arbor, Michigan, USA Anthony G. Myers CH2MHILL 310 W. Wisconsin Avenue - Milwaukee, Wisconsin, USA
ABSTRACT Although ozone has been known for over 100 years, and although it has been used in water treatment for over 100 years, it has not found widespread use as a disinfectant in the water industry. Statistics are not quite reliable, but probably less than 1 percent of the water treatment plants worldwide use it. The major reason for its low use are the costs. Compared to chlorine it is much more costly. Since the water industry has always been trying to provide safe drinking water at the lowest cost it is understandable that it never has been in wide use. But all of this is changing, and actually very rapidly. The regulations for disinfection byproducts in Europe and the US, and guidelines by international organizations such as the World Health Organization, force the utilities to lower the trihalomethanes in drinking water which makes the continued use of chlorine in many cases impossible. This fact combined with the insistence of consumers to receive taste and odorless drinking water points to the use of ozone since it does not only disinfect but also reduces color and odor by oxidizing the organic substances found in water. The paper will describe the ozonation process in detail, and by way of example describe the retrofitting of the treatment process in the City of Ann Arbor, Michigan which installed an ozone facility two years ago. While the water always met water quality standards, the taste and odor were not always good and acceptable. After the installation of the ozone process, it was rated the best tasting water in Michigan. 1. Introduction Ozone might be called an “orphan” disinfectant since it has worldwide a very low use in water treatment plants. Yet it has been known for over a hundred years, and the time line below shows how it evolved (Rice ,1986): 1785 von Marum describes characteristic odor at an electrostatic machine 1801 Cruikshank smells the same odor at an anode 1840 Schoenbein names the substance ozone, after the Greek word “ozein”which means “to reek, smell” 1857 Werner von Siemens designs an ozone generator, cylindrical dielectric type 1893 Oudshoorn, Holland first plant 1906 Nice ,France, Bon Voyage plant, “birthplace of ozonation in a water plant” In the US prior to 1980 there were less than 10 plants, by 1995 there were about 100 plants, and there were more than 50 under design or construction (deMers,1996). The number of plants in Europe are in the thousands. So why are there not more plants using ozone? The simple answer to that is that the water industry has always been concerned with costs, and has tried to provide water at the lowest possible cost. Table 1 shows relative, representative costs of the chemicals used in water treatment plants, and it clearly shows that chlorine is a much less expensive disinfectant (Akness, 1996). But when it is viewed in context with other chemicals, it is not exorbitant. Table 1 Unit Chemical Costs
But things are changing, and actually very fast. The new regulations on drinking water quality, particularly the concerns about disinfection byproducts like trihalomethanes make the use of chlorine no longer an option. In addition the inactivation of viruses and other microorganisms like Cryptosporidium would require high chlorine dosages which would cause higher byproduct concentrations. Therefore, a disinfectant which is strong and does produce only low levels of byproducts is the ideal choice. When compared to other disinfectants like chlorine, chloramine and chlorine dioxide, ozone is the strongest disinfectant and also the fastest acting. Table 2 shows the CT values for the inactivation of viruses and Giardia taken from the US EPA guidance document (USEPA, 1991). Table 2. CT Values for Inactivation by Ozone
*(From USEPA, 1991) These CT values are the product of the ozone concentration in mg/L, and the time measured in minutes. Higher values are necessary at cold temperatures than at warm temperatures, and the values for 3 log reduction (99.9%) range from .3 to 1.0. By contrast, the values for the inactivation of bacteria are around 0.1 (Reiff, 1992). So it is clear that at the moment everything is driven by the CT values for virus and microorganism inactivation. 2. The Ozonation Process There are three components to an ozonation system: the ozone generator, the ozone contactor, and an ozone destruction device. 2.1 Ozone Generation Ozone is produced in an ozone generator. The feed gas can be either air or pure oxygen. A high voltage (6,000-20,000V) is applied to two electrodes and the high voltage produces an arc. In the arc part of the O2 is converted into O3. Ozone is very unstable and reverts back into O2 in minutes. That is why ozone must be generated on-site and cannot be shipped to the water treatment plant. About 1-10 percent of the oxygen flowing past the electrodes is converted into ozone. When air is the feed gas ozone concentrations between 1 and 4 percent are generated. When the feed gas is pure oxygen, the ozone concentrations will be between 4 and 12 percent by weight. About 80 to 95 percent of the energy will be converted to heat, and must be removed at the ground electrode, usually through cooling water. The operational variables are the applied power, the efficiency and design of the generator, the flow of feed gas, and the temperature. A schematic of this process is shown below in Figure 1 following. Figure 1. Dielectric Ozone Generator
Air feed systems have to remove dust and moisture from the air. This can be done using filters, driers, and compressors. A lot of machinery and supervision is necessary. A pure oxygen system uses liquid oxygen (LOX) and is much simpler. Only a vaporizer is needed. 2.2 Ozone Contactors For the ozone to do its work of disinfection and oxidation it must be brought into the water and dispersed as finely as possible. This is accomplished generally through fine bubble diffusers located in baffle chambers, or in a turbine type contactor. Baffled chamber diffusers seem to most prevalent, and the number of chambers, their geometry, the diffuser systems, and their operation differ from plant to plant and are subject to the experience of the design engineers. Figure 2 following shows a typical arrangement for a baffled chamber contactor. Figure 2. Baffled Chamber
A typical ozone contactor usually has several compartments in series with bubble diffusers at the bottom. In the first compartment the water flows downward against the rising bubbles, and in the second compartment the water flows upward. The chambers are covered to prevent the escape of ozone and to increase the partial pressure of the ozone in the contactor. Additional chambers follow to guarantee a contact time between the ozone and the water. Each of the chambers has sampling ports so that the ozone concentration in each chamber can be determined. This is needed to calculate the product of concentration and detention time to get the required CT value. The last chamber should still have a ozone concentration of 0.1 ppm. Figure 3 shows a turbine diffuser contactor, which mixes the ozone with the water. Contact chambers to establish contact time must follow. Figure 3. Turbine Diffuser 2.3 Ozone Exhaust Destruction The off-gas from the ozone contactors generally exceed the Occupational Safety and Health Administration (OSHA) limit of .1 ppm by volume and therefore the remaining ozone has to be recycled or destroyed. The off-gas is first passed through a demister, which traps small water droplets on stainless steel mesh. Then the gas is heated and passed through a destruct unit which contains a catalyst to speed up the process. The power requirement is between 1 to 3 kW per 100 scfm (standard cubic feet/min)( 3 m3/min) of gas flow (DeMers,1996) 2.4 Safety Considerations Ozone occurs naturally in the environment. Probably the largest short-term concentrations occur during thunderstorms when lightening causes a production of ozone. In the office environment ozone is detected near copy machines. Welders are exposed to ozone produced by the arc during the welding process. And urban dwellers living in large urban centers like Denver, Los Angeles, Mexico City, Bogota, Caracas, Sao Paulo, etc., are exposed to ozone concentrations in the .5-1.0 ppm range when the exhaust from cars and industries react with sunlight. Since ozone is a strong oxidant it will react with human tissues, most notably the lungs and will lead to breathing difficulties. The eyes and nose are also affected. The OSHA has established limits for the work place and are shown in Table 3 below. Table 3. Ozone ExposureExposure Limits
In a water treatment plant, ozone monitors continuously monitor the ozone concentrations in the water of the contactor cells, in the air of the off-gas stream, and in the ambient air in and around the ozone building. General alarms sound at a concentration of .1 ppm, and at a concentration of .3 ppm the ozone generators will shut down instantly. This is a great improvement over the “sniff” test used in the older European plants (Reiff,1992). Shutdowns of ozone systems due to leaks have been rare, but they do occur due to leaks around fittings and sample taps. But operators at water treatment plants are probably better protected than urban dwellers. 3. The Ann Arbor Water Treatment Plant The Ann Arbor Water Treatment Plant is a lime softening plant, which uses water from the local Huron River (85%) and well water (15%). The river water and the well water are both very hard, and the hardness is removed through the addition of lime (CaO). After softening the water is filtered through GAC (granular activated carbon) with a layer thickness of 40-60 cm which is on top of a 15 cm thick sand layer with a mesh size of .45-.55 mm. The filtration rates vary from 2-7 m/hr. Prior to the installation of the ozone facility the water was chloraminated twice, once at the intake at the river, and once, just before storage of the finished water at the treatment plant reservoir. Chlorination of the water was not an option since the organic content of the river water is high and would lead to high levels of THMs. Contact time for virus inactivation could not be guaranteed using chloramines. Hence the conversion of the plant to using ozone as the primary disinfectant, and chloramine as a secondary disinfectant to maintain a residual in the distribution system. A sketch of the plant processes is shown in Figure 4 below. Figure 4. Treatment Plant Schematic
4. Pilot Plant Studies at the Ann Arbor Water Treatment Plant No two waters are the same, and a pilot plant study is essential to determine the proper design and operation parameters of the ozonation system. In 1990, the Ann Arbor Water Treatment Plant conducted a pilot plant study with ozone. The study was conducted during the warm and cold water seasons. The pilot testing facility was housed in a climate-controlled trailer and contained the ozone generation and contacting equipment, pilot filters, and a laboratory for routine analyses. The primary objectives for the ozone pilot plant study were to determine:
The pilot plant results were extremely useful for the final design of the facilities. The results ultimately saved Ann Arbor millions of dollars in capital and operating costs for a modest investment in the pilot plant study. The major conclusions from the pilot study were:
5. Retrofitting the Water Treatment Plant In 1992, the final design of the ozone facilities commenced. One major issue was physically locating the ozone facilities on the plant site. The plant site is extremely congested, with residential housing on all sides. A solution was to build the ozone contactors between two existing clarifiers with only several feet clearance on each side. The ozone generating building was placed on top of the contactors because there was literally no other available space. Special provisions to seal the top of the contactors from leaks into the building were made. Plant hydraulics were another major issue for design. The chosen location had yard piping nearby so that contactor inlet and outlet piping was relatively short. However, there was not sufficient head between the clarifiers and filters for the ozone process. Therefore, a low lift pump station was constructed just before the ozone contactors using submersible pumps. This provided sufficient head to pump to the ozone contactors and gravity flow through the rest of the plant. Liquid oxygen (LOX) was chosen as the feed gas. The reasons for choosing liquid oxygen include:
The ozone contactors consist of 4 parallel trains with 8 minutes contact time at a maximum design flow of 200,000 m3/day (50 mgd). Each train has 7 cells. Ozone can be added through fine bubble diffusers in 2 cells, or through an in-line eductor system in the first cell. The contactors also have capability to recycle ozone off gas into the eductor system. Hydraulic tracer tests indicate that the contactors approach 80% of plug flow. A major factor in this is the baffle plate placed in the first cell for even flow distribution. There are 4 ozone generators that can produce 250 kg/day (550 ppd) of ozone. The cooling water system is a closed loop system with a chiller for warm water conditions. The off-gas ozone destructors are thermal/catalytic. The ozone facility was completed in 1996. The ozone system has been in operation for 2 years. The facility meets design criteria, and the water quality has significantly improved. 6. Study of Operation and Byproducts After completion of the plant additions and several months of use, two periods of operation were closely monitored. One period was during February-March 1997 with a water temperature around 10 degrees Co, and the other period was from August -September 1997 with an average temperature of 20 degrees Co. The periods will be called cold and warm period, respectively. The required CT values for virus inactivation during the cold period is .5 mg/L*min, and .25 mg/L*min for the warm period. The actually achieved CT values were 2 mg/L*min during the cold period, and .5 mg/L*min for the warm period. MICROORGANISMS: The raw water was monitored for Giardia, Cryptosporidium, and enteric viruses. While some were found in the raw water at very low concentrations, none were ever found in the finished water. BROMIDES: The bromide levels were monitored both during the cold and the warm water seasons. The average concentrations were around 80 mg/L and there was no significant difference between the seasons. BROMATES: Bromates were detected during both seasons and were usually around 5 mg/L, well below the currently proposed limit of 10 mg/L ALDEHYDES: Aldehydes were formed during the ozonation process, and the major compound was formaldehyde. The concentrations ranged from 20 to 40 mg/L. They were partly removed by the GAC filtration, and during the cold season the concentrations in the finished water were around 20 mg/L, and during the warm season around 10 mg/L. This indicates that during the warmer temperatures the GAC filters are more biologically active. HAA6,HAN,TTHM: Haloacetic acids, haloacetic nitriles and the trihalomethanes were monitored and found at extremely low concentrations. Figure 5 below shows that the pH value of the water seems to have an effect on both the TTHM and HAA6. Figure 5. THM and HAA6 in Finished Water
The average value of TTHM of 5 mg/L, and the average HAA6 value of 2 mg/L show that the processes are well under control. The HAN (Haloacetonitriles) were even lower and mostly under 1 mg/L. TOC: The total organic carbon was removed mostly at the softening stage. The Huron river water is rather hard (300 mg/L CaCO3), and through the addition of lime is softened to about 130 mg/L. The typical TOC in the river water is between 6 and 7 mg/L, and after softening is reduced to about 3 mg/L. The ozonation process reduces a small amount of TOC, usually around 4 percent. There is no significant difference between the cold and warm water season. The GAC filtration averages about a 15 percent TOC removal, with a 20 percent removal during the warm period, and a 10 percent reduction during the cold period. Thus, the overall plant removes about 60 percent of the TOC. The range and average of the TOC removals is shown in Figure 6 below. Figure 6. TOC Removal by Plant Processes
Figure 7. AOC Increase due to Ozonation
AOC: The assimable organic carbon is of concern since it can stimulate the regrowth of bacteria in the distribution system. Figure 7 below shows the increase in AOC during the ozonation process. While the average AOC of the water after softening is about 50 mg/L, the water after ozonation contains more than triple that amount. Contrary to predictions that the bacterial count in the distribution system should go up this has not happened, even during the warm season. COSTS: The two graphs below, Figure 8 and Figure 9, show the costs during the cold season and the warm season, and the effects of the pH. Roughly speaking, the costs during the cold season are about 3 times higher than during the warm season. This can be explained by the fact that during the cold season a 3 times higher ozone concentration was used to satisfy the CT requirements. Figure 8. Chemical
Costs
Figure 9. Chemical Costs
7. Summary and Conclusions The consumers of the Ann Arbor drinking water are very pleased with the quality of the water. The water bills have gone up, but not to the point that the consumers watch the water meters. And indeed, if one looks at the cost of water in comparison to other utilities like gas, electricity, or telephone, it is still under priced. By using ozone as a primary disinfectant, the drinking water quality has greatly improved, and will meet any present and anticipated future regulations despite the fact that the major source of water is rather high in organic compounds. Byproducts occur but are at such low concentrations that they raise no public health concern. So while ozone is today what might be called an “orphan” disinfectant because of low worldwide use, a strong future and increased use can be safely predicted. 8. References DeMers, L.D., et al., Ozone System Energy Optimization Handbook, AWWA Research Foundation,1996, ISBN 0-9648877-1-1. George,D.B. et al., Case Studies of Modified Disinfection Practices for Trihalomethane Control, AWWA Research Foundation,1990,ISBN 0-89867-515-4. Masschelein, W. J.,Ozonation Manual for Water and Wastewater Treatment, John Wiley & Sons,1982, ISBN 0-471-10198-2. Rakness,K.L. et al.,Ozone System Fundamentals for Drinking Water, Opflow, Amer. Water Works Assoc., Vol 22,No 7,July 1996. Reiff,F. and V.M.Witt,Guidelines for the Selection and Application of Disinfection Technologies for Small Towns and Rural Communities in Latin America and the Caribbean, PAHO Technical Series No 30,Washington,D.C.,1992. Rice, R.G.,Analytical Aspects of Ozone Treatment of Water and Wastewater, Lewis Publishers, 1986, ISBN 0-87371-064-9. USEPA, Guidance Manual for Compliance with the Filtration and Disinfection requirements for Public Water Systems Using Surface Water Sources, 1989. Additional Reference: International Ozone Association: www.int-ozone-assoc.org | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
The largest supplier of ozone
equipment on the web. See our New Site www.ozonesupplies.com |
