Australia gets the Hanovia treatment
The town of Busselton draws its water from deep aquifers which have high micobiological integrity, but to meet new legislation needed to find the optimum method of disinfection. Rockwater Pty Ltd contributes this case study.
Busselton's water supply differs from most municipal sources in Australia because the deep aquifers offer high microbiological integrity and very low organic carbon loadings. Most other municipal water providers draw water from shallow production bores, open water bodies, or a combination of sources that contain particulates such as dispersive clays, plant material and algal debris.
In the past, Busselton Water had not disinfected it water supplies and, historically, the water has been free from microorganisms. However, under the Australian Drinking Water Guidelines (1996), set up by the National Health & Medical Research Council and the Agricultural & Resource Management Council of Australia & New Zealand, it was recommended that all water supplies should be disinfected. The 1996 guidelines state that the ideal disinfectant should:
- Effectively remove pathogens over a range of physical and chemical conditions;
- Produce a disinfectant residual which is stable and easily measured;
- Produce no undesirable by-products;
- Be easily generated, safe to handle, and suitable for widespread use;
- Be cost effective.
To assist them in identifying the most suitable disinfection technology, Busselton Water commissioned groundwater consultants Rockwater Pty Ltd to investigate the available methods and make recommendations for a suitable system. The objectives for the study were to determine a disinfection system that effectively met the 1996 guidelines. Rockwater found that none of the available disinfection methods it investigated met all of the requirements, so a compromise was necessary to obtain the optimum process for Busselton's reticulation system.
The five most commonly used methods of disinfection - chlorination, chloramination, chlorine dioxide, ozonation and ultra-violet (UV) irradiation - were all investigated as they offer good bactericidal activity.
Chlorine-derived methods all provide residual disinfection, whereas UV and ozonation do not. UV produces no disinfection by-products, whereas all the other methods produce very low concentrations of potentially harmful organic compounds. Chlorine substances emit minor tastes and odours that are objectionable to some consumers, while ozone and UV do not affect taste or odour.
Chlorine substances are all corrosive, and plant operators need to be trained
in handling procedures. Ozone has a complex operating system that can present
a danger to workers in the vicinity of the gas generating system. UV presents
minimal risks to operators because all the energy is confined to a sealed container
without generation of gases. There is no danger of overdosing with UV, whereas
accidental overdosing of water supplies using chlorine and ozone is possible
and may present a health hazard.
All disinfection methods with the exception of UV are both temperature and pH dependent - the higher the temperature and pH the less efficient the disinfection process. Busselton's water supply is warm and slightly alkaline and is therefore less suitable for chlorine disinfection than many other water supplies.
Many water supply authorities prefer to use chlorination because their reticulation systems are exposed to the open air after treatment (for example, water is often stored in open-air tanks prior to distribution), and chlorine provides residual disinfection. Ozone and UV do not provide residual disinfection, so these methods must be used in-line after storage, as tanks can become contaminated by airborne bacteria or organic waste matter.
UV was finally selected because it met all of the design criteria except for one - maintaining a persistent residual barrier. However, the water is maintained in a closed system following treatment, and the chance of contamination is minuscule given that in the past the water supply maintained compliance without any disinfection.
Chlorination achieves only one of the conditions in the 1996 guidelines: it maintains a residual barrier, but it breaks down rapidly. To maintain sufficient residual disinfection at the outer extremities of the system, therefore, it may have been necessary to exceed the maximum dose rate at the point of disinfection. Chlorination is also a less effective disinfectant than UV at the recommended dose rates of 0.5 mg/L and 30 mWsec/cm² respectively. UV installation costs at Busselton were estimated to be about 2.2 times those for chlorination and chloramination disinfection systems, but UV was estimated to be about half the cost of chlorine dioxide systems - operational costs were estimated to be only about 25% of those for chlorine.
Move from chlorination
At the Australian Water Association's 18th Federal Convention in 1999, it was reported that municipal water authorities in other countries such as the USA were actually steering away from chlorination because of concerns about disinfection by-products. In the Netherlands, chlorine is no longer used to disinfect municipal water supplies - UV irradiation is now the most widely used method.
Australia's 1996 guidelines state that the implementation of a disinfection system should be a consultative process involving the community. In a previous survey, residents of Busselton had strongly objected to a proposal to chlorinate the town's water supply. In February 1997, therefore, Busselton Water decided to adopt UV as the sole method of disinfection. To achieve a confident level of bacteria destruction, a minimum dose rate of 30 mWsec/cm² was recommended.
In 1997 a study tour of UV equipment suppliers and manufacturers was conducted, and Hanovia won the contract to trial a UV disinfection system.
Two Hanovia single lamp, medium pressure (MP) units were tested for a 12-month period. Biomass destruction tests undertaken using B. subtillis (aged and fresh) and Ecoli across a variety of flow rates proved the MP units to be effective. Subsequently, Hanovia was commissioned to supply eight PMD320 MP systems.
The UV systems in Busselton each contain a single medium pressure UV lamp housed in a protective quartz sleeve. Each unit is equipped with a UV monitor, variable power outputs that increase the UV output as flow rates increase, and an automatic quartz sleeve wiping system.
MP lamps are polychromatic, as they produce UV between 200nm to 300nm. These lamps are specifically manufactured by Hanovia to have optimised spectral outputs between 240nm and 280nm. The most effective germicidal wavelength is at 265nm, and Hanovia MP lamps are designed with this wavelength as their peak output. Recent work undertaken in the USA has shown that UV light centred on 271nm to be most effective at deactivating Cryptosporidium parvum, while UV at 263nm is best for deactivating the MS-2 coliphage virus.
Low Pressure (LP) UV lamps are monochromatic as they only have a single spectral output at 254nm. Typically, between 8-12 LP lamps are required to match the output of a single MP lamp. The output of each Hanovia MP lamp is compared against the absolute standard which is held at the PTB in Berlin.
UV dose is calculated by measuring lamp intensity, water flow rate and water transmissivity. Lamp intensity is measured by a monitor positioned on the inner wall of the UV chamber. If more than one UV lamp is used, a dedicated monitor is fitted for each lamp. Multi-lamp LP systems suffer from the inability of a single (or even several) monitors to detect failure or reduced output from lamps not positioned immediately adjacent to the monitor. But for effective monitoring, UV intensity must be measured, not inferred.
As the geometry of UV chambers is fixed, the flow rate of water can be used to calculate the residence time in the chamber, with a high flow rate leading to a reduced residence time. Hanovia uses detailed CFD modelling techniques to determine the actual distribution of residence time within the chamber. A third method of calculating UV dose uses a transmittance monitor which continuously measures water transmittance and feeds the data to a control unit.
For a high quality, stable groundwater source like Busselton's, however, transmittance was assumed to be constant, so a figure of 90% (when measured against double distilled water) was used. The UV dose is displayed in mWscm-2, or mJcm-2. The UV dose is displayed on a Photon control panel, which is capable of data-logging variables such as water flow, UV dose, UV intensity and wiper frequency, allowing any faults to be date- and time-stamped.
Each Hanovia system operates in duty/standby mode. The power to each lamp is continuously varied, ensuring that the specified UV dose is always delivered. Both systems may be operated at 50%, and in the event of failure a 'hot' standby is available immediately.
The operating protocols are strictly fail-safe and will not under any circumstances
allow untreated water forward. The control panel provides a broad range of signals
to actuate valves, initiate pumping, and generally provide meaningful monitoring
and control of the system.