• About this site|
• Site index|
• Contact us|
• Jobs

An Introduction to Drinking Water Contaminants, Treatment and Management

1 Introduction

2 Roles and Responsibilities in the Production of Drinking Water

3 Water Supply Management – Information Sources

4 Origins of Contaminants in Drinking Water Sources

5 Water Treatment

Appendix 1: Activities and the Contaminants That May Contribute to Source Waters

Appendix 2: Treatment Processes and Their Efficacies

5 Water Treatment

5.1 Introduction

The NES aims to ensure that the effects of new catchment activities on the ability of water supplies to produce safe water for their communities are adequately considered. The ability of a water supplier to provide safe water when an activity is introduced into a catchment depends on three factors: the changes in the quality of the water at the water supply’s abstraction point¹⁴ as the result of the new activity; the types of processes being used to treat the water; and how well these processes are operated.

To assess the effect of an activity, an understanding of water treatment is necessary. This section provides an introduction to water treatment. In particular, it:

• describes the types of contaminants the treatment processes remove

• identifies the main treatment processes used in New Zealand

• provides semi-quantitative guidance on the ability of these processes to remove the contaminants listed in the DWSNZ.

5.2 Treatment processes

5.2.1 Introduction

Source waters, whether they are surface waters or groundwaters, can contain a range of contaminants that may make the water unsafe to drink or aesthetically unacceptable (eg, bad taste, odour or appearance). Such contaminants include: particles, microbiological contaminants, naturally occurring chemical substances and chemical substances derived from human activities. Of these, the two for which treatment is most important are particles and microbiological contaminants. Treatment for these contaminants is particularly important for surface waters and shallow groundwaters that are affected by events above ground. Deep groundwaters, or groundwaters from confined aquifers¹⁵ are expected to be of much better quality than surface or shallow groundwater, and in some instances are untreated, eg, Christchurch’s groundwater sources.

The generic elements of a full treatment train (series of treatment processes) are depicted in Figure 2 in the order in which they will occur in the treatment plant. The ‘pre-treatment’, ‘particle removal’ and ‘disinfection’ processes will always occur in the order shown, although some may be omitted if the quality of the source water does not require them. The ‘additional processes’ cover processes that are specific to a particular supply, and their location will depend on the design of the particular treatment system.

One or more pre-treatment steps may be used. These are often unsophisticated processes designed to reduce the load of contaminants reaching the main particle removal processes, or to condition contaminants in the water to make their removal by later processes easier. They include settling (eg, in a reservoir or sedimentation basin to allow particles in the water to sink to the bottom), aeration and chemical pre-oxidation.

Particle removal is the first of the main treatment steps, and usually consists of a series of processes. The last of these is filtration which is preceded by steps designed to improve filter performance. Particle removal is important because these processes remove the larger microbiological contaminants (protozoa, such as Giardia and Cryptosporidium), some of which are resistant to chlorine, as well as the non-living material that contributes to the cloudiness (turbidity) of the water. Particle removal is also important for the efficacy of the disinfection step.

The disinfection step must take place after the particles have been removed, when the water is as ‘clean’ as possible. Micro-organisms adsorb to particles in the water. Once adsorbed, they are shielded to a degree from the effects of the disinfectants. As much as possible of the particle load in the water must be removed before disinfection to ensure adequateinactivation¹⁶ of the organisms and to remove micro-organisms already adsorbed to particles.

The main treatment processes are not primarily intended to remove any of the large number of chemical contaminants listed in the DWSNZ.¹⁷ As will be seen in a later section, some contaminants are coincidentally removed by particle removal and disinfection processes. Where contaminants in the source water cannot be removed satisfactorily by the main treatment processes, additional treatment processes may need to be incorporated into the treatment train. Additional treatment processes that do not have a direct role in removing contaminants may also be required to adjust the water chemistry to improve the performance of a process that is removing contaminants, eg, adjustment of the pH. The location of the additional treatment processes will depend on their function and the needs of the other processes.

This section briefly describes the treatment processes most likely to be encountered in New Zealand and explains their function. The principles of operation of a particular process are independent of the treatment plant, but the physical design and implementation of the process can vary with the treatment plant. At the end of the section there is information about the number of treatment plants employing different treatment processes in New Zealand to indicate how common a particular process is.

5.2.2 Pre-treatments

Water treatment plants are likely to reliably produce safe drinking water, if the conditions under which they operate remain constant (see section 5.4.4.2). A source water of changing quality is difficult to treat. Treatment plants drawing water from underground or from a lake or reservoir will usually have a source water that changes little, or if it does, it changes gradually. Rivers and streams, however, are subject to rain events, and treatment plants abstracting water from these types of source can be exposed to rapidly changing source water quality. One of the functions of pre-treatment processes is to provide a ‘buffer’ against changes in source water quality, so that quality changes and the rate of change are reduced.

Pre-treatment processes may also be used to modify the water chemistry and possibly the contaminants themselves, to improve their removal by later treatment processes.

Where treatment plants experience biological growths in parts of their system, such as the clarifier tanks, pre-treatment may also be used to control these growths.

• Sedimentation basins: Sedimentation basins reduce the load of sediment in the water reaching the main treatment processes, and they reduce the magnitude of water quality changes. This is done by providing a large impounded area in which the water flow is reduced, which gives time for particles to settle out under gravity. During rain events they provide a buffer against rapid changes in the quality of water entering the treatment plant. Insoluble chemical contaminants may also be partially removed by the settling process.

• Infiltration galleries: Levels of turbidity and natural organic matter¹⁸ (NOM), and to some extent microbiological contamination, in river or stream water can be reduced by abstracting the water indirectly from the source through an infiltration gallery. By burying open-jointed or slotted pipes in the bed of a river, stream or lake, water percolates through the gravels and sands of the bed and into the pipes where it is diverted to a collection well on the bank and pumped out for the water supply. This crudely filters the water as it passes through the media of the riverbed so that a fraction of the particles, and contaminants that may adsorb to the riverbed media, are removed. This form of pre-treatment achieves little removal of Cryptosporidium.

• Pre-oxidation: Pre-oxidation may be carried out using oxidising chemicals such as chlorine, ozone or potassium permanganate. It is typically used to modify NOM (the substances that give some waters a yellow-brown colour) to improve its removal during the coagulation/flocculation step. It may also be used to oxidise soluble iron or manganese (usually in groundwaters) and sometimes arsenic, to precipitate them for removal by particle removal processes. This process may also control unwanted biological growths in other parts of the treatment plant. A drawback of pre-oxidation is that it tends to increase disinfection by-product (DBP) formation. To minimise DBP formation, it is usually preferable to remove as much NOM as possible before chemical disinfectants are added to the water. This may require avoiding the use of chlorine or ozone when the NOM concentration in the water is causing unacceptable levels of DBPs.

Pre-oxidation can destroy some cyanotoxins (toxins produced by cyanobacteria: blue-green algae). See Table A2.5 for guidance on the efficacy of oxidants in destroying toxins.

• Aeration: Aeration of a source water can introduce oxygen into the water to oxidise contaminants, such as iron or manganese, to an insoluble form so that they can be removed as particles (as with pre-oxidation above). Passing air through the water will also assist in removing gases (eg, hydrogen sulphide, carbon dioxide) or volatile contaminants (eg, vinyl chloride, trichloroethene). The removal of contaminants by aeration is also known as air-stripping. Aerators can be designed to entrain air in the water by ‘breaking up’ the water and passing it through the air, eg, sprays or trickling towers, or by bubbling air through the water.

• Copper sulphate treatment: Copper sulphate is an algaecide, and is sometimes used to control algal blooms in static source waters, such as reservoirs. This approach to controlling algae can result in enhanced taste and odour problems and elevated toxin levels in the water – on death, the algal cells break up and release toxins and taste and odour compounds into the water. Bloom development is better controlled by minimising factors, such as nutrient concentration, that encourage algal growth.

5.2.3 Particle removal

By weight, clay, silt and sand particles are the main contaminants removed by this group of processes, but particle removal processes also improve the microbiological quality of the water by physically removing the micro-organisms. The most important task of particle removal, from a public health view point, is the removal of protozoa – some of which are not easily inactivated by chlorine. Particle removal processes can also contribute to the removal of bacteria. Adsorption of bacteria onto larger particles in the water ensures the bacteria are removed with the material to which they are adsorbed. Free bacteria (those not adsorbed) are not as easily removed because of their small size.

The processes within this group may be used individually, but it is more common for two or more processes to be used in series to remove particles more effectively from the water.

5.2.3.1 Coagulation/flocculation

This is the first step in the main treatment train of a full conventional treatment system, and prepares the water for particle removal by subsequent processes. A coagulant, usually an aluminium (eg, alum) or iron salt, is added to the water. This encourages small particles in the water to stick together to form larger particles, which are more readily removed from the water by the processes that follow. The addition of the coagulant also results in the formation of ‘flocs’ (particles) of insoluble metal hydroxides. The flocs further assist in contaminant removal by providing surfaces for adsorbing contaminants, and trapping contaminants as floc formation occurs.

Particle removal is often the chief purpose of the coagulation/flocculation process, but by adjusting the coagulation conditions, NOM can also be removed. This helps to control the formation of DBPs (see section 5.2.4 and 5.4.4.1) following the disinfection process.

The coagulation/flocculation process, combined with the other processes discussed below, can remove metals, bacteria and protozoa, although to varying degrees that depend on the metal and the type of micro-organism. Its role in removing Cryptosporidium is important because chlorine cannot inactivate this protozoan under water treatment conditions. Some non-metals and pesticides are also partially removed by this process.

5.2.3.2 Clarification

Clarification follows the coagulation/flocculation step and provides more time for the particles to stick together and settle out of the water, thereby reducing the sediment load that has to be removed by the filters.

• A process called direct filtration is used in some supplies if the turbidity (particle content) of the source water is low. In this process, the clarification step is omitted and coagulant is dosed directly before the filters. This reduces the turbidity by increasing the efficiency with which particles stick to the sand grains within the filter.

• DAF (dissolved air flotation) is used in one New Zealand treatment plant, instead of clarification. DAF works by floating the larger particles out of the top of the clarifier rather than letting them settle to the bottom.

5.2.3.3 Filtration

Filters of one type or another are the final process by which particles are removed from the water. When the source water is highly turbid, coagulation/flocculation and clarification steps usually precede the filters, but for some low-turbidity source waters or in small water supplies with limited resources, the filter may be the only particle removal process.

No matter which type of filter is used, particles and other contaminants become trapped by the filtering medium. The amount of trapped ‘dirt’ will eventually reach a point at which the filter cannot satisfactorily operate and the filter must be cleaned. How this is done depends on the type of filtration, but in all cases treatment plant operators aim to maximise the period the filter is operating before cleaning is required. This is because the cleaning process reduces the efficiency of the operation. The cleaning process is discussed below for the commonly used rapid sand filtration.

The following filter types are found in New Zealand.

• Rapid sand filters: Sand filters are widely used for particle removal, but usually in combination with other processes. Although called ‘sand’ filters, they often contain two types of sand overlain by a layer of small coal particles. They strain out particles that are too big to pass through the spaces between the sand grains, and allow smaller particles to travel down into the sand where they are removed by adsorbing to the sand grains.

Rapid sand filters and other granular filtration systems (eg, granular activated carbon) are cleaned by a process termed ‘backwashing’. This process forces water backwards through the filter (possibly in combination with the injection of compressed air to help in dislodging the ‘dirt’) and discharging this water to waste. Backwashing must be done with previously treated water, therefore the more frequently backwashing is required, the greater the volume of treated water that has to be sent to waste during backwashing, and the less efficient the operation.

• Diatomaceous earth filtration: Diatomaceous earth (DE) is a porous sedimentary material made of the silica skeletons of diatoms (a type of microscopic algae). The DE is continuously fed into these filters with the water and builds up as a cake on the support membrane. DE can be used as a ‘polishing’ step following other filtration treatments, because it can remove fine particles more efficiently, or for filtering fairly clean source waters.

• Bag filtration: Bag filters are typically constructed from porous woven or felted fabrics. The fabric from which they are made is non-rigid. Water under pressure is forced through the fabric from the inside of the bag to the outside; as the water passes through the bag, particles are removed on the fabric surface or within the fabric. The shape of the bag is maintained during use by a rigid support or housing. A range of pore sizes can be removed by bag filters, but to provide protection against protozoa, they must be able to remove particles larger than 1 µm. As particles accumulate in and on the bag, the pressure drop across the bag fabric (ie, the pressure required to push water through it) increases and eventually reaches a point at which the bag must be replaced.

Bag filters tend to be used in small water supplies rather than large systems. A single filter unit may be used or several may be used in series.

• Cartridge filtration: These filters are similar to bag filters, but differ in the following respects. Cartridge filters are typically constructed of rigid or semi-rigid material, and they are housed in pressure vessels so that water is forced through them from the outside to the inside. Like bag filters they must be capable of removing particles greater than 1 µm in size to remove protozoa. As with bag filters, ‘dirty’ cartridges must be replaced.

• Slow sand filtration: Although these filters use sand as a medium, as the name suggests, the rate at which the water passes through them is about 10 times slower than rapid sand filters. Other than filtration rate, the most important difference in the two types of sand filter is that slow sand filters make use of biological activity to treat the water, ie, there is a microbial community living within the sand and on the surface of the sand bed. This biological activity plays an important part in removing micro-organisms and NOM. Unlike a rapid sand filter operating in conjunction with coagulation/flocculation and clarification, slow sand filters are not designed to remove large amounts of particulate matter from the water.

• Membrane filtration: Membrane filters are ‘high-tech’ systems. The types of membrane filtration normally used in drinking water treatment are microfiltration (MF) and ultrafiltration (UF). Their primary task is the removal of particles (including protozoa) and bacteria. They remove contaminants by size-exclusion, ie, the contaminant can be removed from the water because it will not fit through the pores in the membrane. All membranes have a distribution of pore sizes. The pore size of a particular membrane may be specified as a ‘nominal pore size’ (the average pore size) or the ‘absolute pore size’ (maximum pore size). MF membranes generally have a nominal pore size of 0.1 µm and UF membranes a nominal pore size of 0.01 µm.

Nanofiltration (NF) and reverse osmosis (RO) membranes are generally employed to remove dissolved contaminants, eg, in water softening, as they are not designed to remove particles, although they can do this. They operate on a different principle from MF and UF, which allows the removal of particles as small as 0.001 µm approximately in the case of NF and 0.0001 µm in the case of RO. They operate at a higher pressure than MF and UF membranes, and are more expensive to purchase and operate.

Membranes are housed in modules. Banks of modules are set up within membrane filtration plants: the greater the number of membranes the greater the ability of the treatment plant to treat larger volumes of water.

In 2005, 13 New Zealand treatment plants reported the use of membrane filtration systems. The type of membrane was not recorded, but they are most likely to have been microfiltration or ultrafiltration technologies.

5.2.4 Disinfection

There are three methods of disinfecting presently in use in community water supplies in New Zealand.

• Chlorination: Chlorination is the most widely used disinfecting method world-wide. It inactivates bacteria, viruses and the protozoan, Giardia. It will not, however, inactivate Cryptosporidium rapidly enough for use in water treatment.

An important advantage that chlorine has over the other two main disinfectants used in New Zealand is that it remains present long enough in the water to provide a disinfectant residual after treatment.¹⁹ This is important for the maintenance of a safe water supply. In the event of low levels of contamination entering the distribution zone, the chlorine provides a degree of protection against microbiological contaminants.

The efficacy of chlorine as a disinfectant is determined by the pH; higher acidity (that is, lower pH) enhances disinfection. As well as being a good disinfectant, chlorine is a moderately strong oxidising chemical and is therefore also used for oxidising contaminants²⁰ during treatment (see section 5.2.2).

• Ozonation: Ozone is a more powerful oxidising agent than chlorine and a stronger disinfectant, and it can be used in both roles during water treatment. It can rapidly inactivate Cryptosporidium and therefore provides a satisfactory barrier to this organism, as well as to viruses and bacteria. It is a very reactive gas and even in ‘clean’ water it decomposes rapidly which means it cannot provide a disinfectant residual after treatment.

• Ultraviolet irradiation: Ultraviolet (UV) light at a wavelength of 254 nm can inactivate micro-organisms by damaging their DNA. Like ozone, UV light can inactivate protozoa and bacteria. Some viruses are resistant to inactivation by UV light, although it is effective against the majority of viruses. The disinfection efficacy of UV light is compromised by particles in the water, as also happens with the chemical disinfectants. The intensity of the light determines the ability of UV light to inactivate micro-organisms, and particles decrease the intensity of light passing through the water.

• Disinfectant combinations: Some treatment plants may use more than one disinfectant. Common combinations are ozone and chlorine, or UV disinfection and chlorine. Chlorine is used in combination with these two disinfection systems because neither ozone nor UV irradiation provides a disinfecting residual. Ozonation or UV disinfection, therefore, is used to inactivate Cryptosporidium, and the chlorine is then added to maintain the good microbiological quality of the water during distribution to consumers.

All chemical disinfectants have the drawback of reacting with naturally occurring organic contaminants in the water to produce DBPs. These substances can have undesirable health effects (eg, cancer). However, it is generally agreed that adequate disinfection should not be compromised in trying to minimise the extent of DBP formation during treatment (WHO, 2004), because the consequences of a microbiologically unsafe water are felt within a matter of days, not decades as is the case if DBPs exceed their MAV.

Each of the oxidising disinfectants, ie, chlorine, ozone and chlorine dioxide (presently not used in New Zealand), can destroy some, but not all, cyanotoxins. Table A2.5 provides guidance about the toxin-destruction capabilities of each disinfectant.

5.2.5 Additional treatments

• Activated carbon adsorption: Activated carbon contains a very high surface area per unit weight that can adsorb contaminants. Activated carbon adsorption can remove a wide range of contaminants from water, particularly trace organic contaminants including industry solvents and pesticides. In New Zealand, activated carbon is primarily used to remove taste and odour compounds formed in minute quantities by micro-organisms in the water. Some supplies may also introduce activated carbon treatment to deal with the cyanotoxins produced by blooms of blue-green algae.

Algae and some bacteria are the usual sources of tastes and odours in drinking water. The growth of these organisms, and therefore the concentrations of the taste and odour compounds or toxins they produce, depend on several factors – some of which are influenced by season. These factors include nutrient (nitrogen and phosphorus) concentrations which may be influenced by catchment activities, temperature, light intensity and oxygen concentration in the water. Algal blooms are more likely to develop in the summer when the water is warm and greater sunlight assists with photosynthesis.

Activated carbon can be used in two forms: powdered or granular. Whichever form is used, once the carbon’s adsorption capacity is exhausted it is disposed of to waste.²¹ Granular activated carbon is used like another filter medium, either as the medium in a standard rapid sand filter, or in a separate contactor that the water stream passes through. The contactor is located near the end of the treatment train so that high-quality water passes through it. Activated carbon used in this way can become biologically active (micro-organisms colonise it) which can improve its ability to remove contaminants. Powdered activated carbon is usually added to water before the final filtration step. This allows contact with the water and adsorption of the problem contaminants, before the carbon is removed by the filters and discharged to waste during the filter backwash. Use of carbon in this way avoids the major capital costs of installing special contactors, and is favoured when use of the carbon is intermittent.

• Ion-exchange adsorption: Synthetic organic resins that can attract and adsorb positively or negatively charged ions (depending on the design of the resin) in the water are used in the ion-exchange adsorption treatment process. As with any adsorbing material, there is a limit to the amount of contaminant they can adsorb, but they can be regenerated, usually by pumping a brine solution through them. The most widely used ion-exchange systems are those used for removing positively charged ions. These are used to soften water, by removing calcium and magnesium, and for removing soluble iron and manganese (these metals in their insoluble form will foul the resins and inhibit their operation). Other contaminant metals in their soluble forms can also be removed by ion-exchange systems with varying degrees of efficacy.

• Greensand filtration: Greensand is a naturally occurring material that is treated to form a layer of manganese oxide on the surface of the grains. This coating oxidises soluble iron and manganese to their insoluble form when they come in contact with the surface so that they can by removed by filtration. The coating is regenerated with potassium permanganate or chlorine, or a combination. Permanganate dosing needs to be carefully controlled to avoid excess manganese contaminating the treated water.

• Precipitation softening:²² This form of hardness control is little used in New Zealand. It creates conditions in the water that make calcium and magnesium compounds precipitate so they can be removed as solids. Adjustment of the pH is required after the process to return it to a satisfactory level. Some heavy metals can also be removed from the water during this process.

• pH adjustment: Optimum operation of some treatment processes requires adjustment of the water’s pH, eg, oxidation of iron and manganese and the coagulation process. There may also be a need to adjust the pH of a water to reduce the tendency of the treated water to dissolve materials in the distribution zone or consumers’ plumbing.

5.2.6 Combinations of treatment processes

As shown by Figure 2, there will usually be more than one treatment process in operation in a treatment plant. For some contaminants this will result in more than one process contributing to the removal or inactivation of a contaminant. For example, bacteria are removed to some degree by particle removal processes and also by disinfection. For other contaminants only one process may reduce the contaminant’s concentration, eg, Cryptosporidium is only removed by particle removal processes.

There are also some instances where a combination of two or more processes is required to achieve removal of a contaminant and removal of the contaminant fails if one of the processes is omitted. An example is the removal of soluble iron or soluble manganese by precipitation. The soluble metal is oxidised by aeration, chlorine or ozone to form an insoluble form of the metal. This is then removed by particle removal processes. Omission of either the oxidation step, or the particle removal process, results in the iron or manganese being present in the finished water.

The combination of coagulation/flocculation/clarification/filtration is commonly used in the treatment of surface waters. Treatment plants using direct filtration, in which the clarification step is dropped from this combination, will achieve a lower removal of particulates and protozoa than can be achieved using the full combination. The maximum turbidity level in the raw water that can be satisfactorily treated is also lower when direct filtration is used.

5.2.7 Treatment processes used in New Zealand

Table 2 lists treatment processes used by New Zealand water treatment plants, and the numbers of each in use.²³

The treatment processes have been grouped into the generic categories used in Figure 2. The order in which the generic processes would operate in the treatment plant are shown in descending order, except for the additional processes which are likely to used before disinfection. As indicated in the earlier discussion, more than one process will usually be in use at a treatment plant.

5.3 Efficacy of contaminant removal by treatment processes

Estimates of the abilities of treatment processes to remove specific contaminants listed in the DWSNZ are provided in tables in the Appendix 2 (Tables A2.1–A2.6). The information contained in the tables is not exhaustive. The absence of an entry indicates that either the treatment process has no effect on the contaminant concentration, or no information has been found about its efficacy in removing that contaminant. Whichever is the case, it should be assumed that the process has a negligible effect on the contaminant concentration until information is available to show otherwise.

Tables A2.1–A2.6 should be used as a starting point to assess whether there are any obvious concerns about the ability of a treatment plant to remove the contaminants expected from a particular activity. Further information that might help in determining the ability of the treatment plant to deal with a particular contaminant should then be sought from the water supply engineer or treatment plant operator.

The tables, particularly Tables A2.2–A2.6 covering chemical contaminants, are guides only and should not be used to attempt quantitative calculations. The estimates of removal in Tables A2.2–A2.6 are based predominantly on data from laboratory studies, although some pilot and full-scale studies also contribute to the estimates. The results from laboratory and pilot-scale studies do not always translate directly to what is found at full scale, and the tables should therefore be regarded as providing ‘best-case’ removals. Further, it is not possible in summary tables such as these to take account of the different variables that may affect the performance of a treatment process. These include different types of membrane, different activated carbon types, the nature of coagulants used, and the chemistry of the raw water.

Other factors that will influence the efficacy of a treatment process, but which cannot be included in the tables are discussed in section 5.4.

Table 2: Treatment processes used in New Zealand, the approximate numbers of treatment plants recorded as using the process, and primary contaminants the process is designed to remove

View treatment processes used in New Zealand, the approximate numbers of treatment plants recorded as using the process, and primary contaminants the process is designed to remove (large table).

5.4 Factors affecting the efficacy of treatment processes

5.4.1 Introduction

The guidance to the efficacy of treatment processes provided in Appendix A2 assumes they are operating as well as they can. This section discusses factors that may influence the ability of a particular treatment plant to achieve the estimated levels of removal given in Appendix A2.

5.4.2 Process optimisation

One or more parameters determine treatment process performance. Some of these parameters can be controlled by the treatment plant operator, others cannot. For example, the pH and coagulant dose are important controllable parameters that determine coagulation performance, and chlorine dose and pH influence the effectiveness of disinfection by chlorination. The turbidity and NOM content of the raw water, which are outside the operator’s control, also influence the quality of the water produced by the treatment plant.

To achieve the best contaminant removal, the treatment plant operator needs to optimise the controllable parameters for the quality of the water being abstracted. Continuously changing water quality makes this difficult, especially if the processes are under manual control.

With experience, the operator develops an idea of the approximate parameter settings required to treat the normal range of raw water quality. These will act as guides for establishing optimum plant performance, but monitoring of other parameters that show how well the treatment is working will be needed to fine-tune the process. For example, the turbidity of the water leaving the clarifiers indicates how well the coagulation/flocculation and clarification processes have been optimised, and the chlorine residual in the water and water’s pH value can be used to assess the effectiveness of the chlorination.

The removals noted in the tables in Appendix A2 are unlikely to be achieved without treatment processes being optimised.

5.4.3 Process control

Once the optimum treatment conditions are established, the process must be controlled to maintain optimum treatment. In New Zealand, the levels of treatment plant control range from fully automated systems with alarms to alert operators to conditions that require their attendance, to small treatment plants with manual controls where checks on operation of the plant may only be undertaken every few days.

As well as hardware influencing treatment control, the training, skill and experience of the operator affects how well optimised treatment is maintained.

Without adequate control hardware and operator training, treatment plants may not function reliably. Failings in treatment control will create the greatest threat to the production of safe drinking water when source water quality is changing, such as during a rain event (see section 5.4.4.2).

5.4.4 Source water quality

5.4.4.1 Contaminants

The contaminants or constituents of water that are most likely to influence the performance of treatment processes are not the trace contaminants that have direct health consequences, but the major constituents of the water, which have no direct health consequences. These include: turbidity, NOM, hardness and some major ions, such as sodium ions.

Turbidity

Turbidity (particles) needs to be removed to avoid deterioration in the effectiveness of filtration, disinfection and adsorption processes. The ability of a treatment plant to adequately treat the water to remove particles must be evaluated if a new activity is likely to result in a major increase in the turbidity of the source water.

Filtration processes are part of the combination of treatment processes that remove particles from water, but filters will rapidly clog, or there will be breakthrough of particles through them (ie, there will be turbidity in the filtered water) if the turbidity of the water entering them is too high. Pre-treatment sedimentation or the combination of coagulation/flocculation and clarification must reduce the turbidity to a level that the filters can handle.

The efficacies of all disinfection processes, whether they are chemical or physical (UV), are adversely affected by particles in the water. Microbes that are adsorbed onto the surface of particles are given some protection from chemical disinfectants, and the intensity of UV radiation passing through water is reduced by scattering caused by particles.

Processes that remove contaminants by adsorption depend on the contaminants of concern reaching the adsorbing surface so that adsorption can take place. Water with unacceptably high levels of particles will rapidly reduce the surface area available for adsorption, and with it reduce the removal efficacy of the process.

Natural organic matter

Disinfection, oxidation and adsorption efficacies are reduced by NOM. New catchment activities that could contribute to the NOM concentration in the source water can adversely affect the capabilities of a treatment plant²⁴ unless the processes in use can adequately reduce the NOM concentration.

The concentration of a chemical disinfectant in water is one of the factors determining how quickly it can inactivate microbes. Chemical disinfectants react with NOM, which reduces their concentration, and so reduces their disinfection ability. The disinfectant dose added to the water can be increased to compensate for this, but this increases the concentrations of DBPs that will form, which is undesirable because of their possible health effects.

Disinfection by UV radiation is made less effective by NOM, because NOM absorbs light at the same wavelength generated by UV lamps.

The reduction in the concentration of chlorine or ozone that results from their reaction with NOM reduces their capacity to oxidise contaminants, affecting their disinfection ability.

Fouling of ion-exchange resins and activated carbon surfaces affects the ability of both adsorption processes to remove contaminants from water.

Hardness and other major ions

Water hardness arises from calcium and magnesium ions in the water, and is often the result of the water having been in contact with limestone or marble (rock types consisting of calcium carbonate). High hardness creates problems of scale formation on water heating elements, and inhibits the lathering of soap.

The predominant treatment problems resulting from waters that are hard or contain high concentrations of other major ions concern are ion-exchange and UV irradiation.

Ion-exchange can be used to soften water, but if the primary use in a particular situation is to remove iron and manganese, an increase in source water hardness may result in the exchange resin removing calcium and magnesium and not iron and manganese. Waters high in sodium may also compromise the removal of iron and manganese.

The intensity of the output from UV lamps can become reduced because of the formation of calcium scale on the quartz sleeves in hard waters. This affects their disinfection capability.

5.4.4.2 Source water quality variability

The importance of process optimisation has been noted in section 5.4.2. Treatment processes function best when water quality conditions are constant. An optimised set of control parameters is only of value for a given source water quality. A change in source water quality therefore requires re-optimisation of the control parameters. The larger and more rapid this change, the more difficult it is for the treatment plant operator to continue to produce good quality water. A complete loss of control over water quality may lead to unsafe water being pumped into the distribution zone. Failure of the particle removal processes, for instance, may result in any Cryptosporidium in the raw water being present in the water supplied to consumers.

Rain events are the most common cause of changes in source water quality, and the more extreme the event the greater the likelihood of a breakdown in treatment barriers. Rain events present treatment plant operators with two problems simultaneously. First, changes in water quality require the process control parameters to be re-optimised. The speed with which this can be done will determine how well the barrier to contaminants is maintained. Second, at the time when maintaining optimum treatment is at its most difficult, the concentrations of contaminants in the source water are often at their greatest. The most apparent quality changes that occur with rain events are increases in turbidity and increases in colour (NOM content) of the water. The changes in water quality that are not so apparent are the increases in microbial contaminants.

Pathogen (disease-causing microbes) concentrations can increase substantially in source waters during rainfall. From non-point sources of contamination²⁵ this is partly due to rainfall washing more microbes from faecal material into the source water, and partly because of increased river flow re-suspending microbes contained in the sediment on the riverbed. Point contamination sources may also add to faecal contamination of the water in rain events. For example, a waste water treatment plant that receives both stormwater and sewage may not be designed with sufficient buffering storage. A rain event will increase the amount of stormwater entering the treatment plant. This increase in flow may exceed the plant’s ability to treat the water, and treatment stages may have to be by-passed resulting in untreated or partially treated water entering the water source.

With regard to implementation of the NES, new catchment activities that may lead to marked increases in turbidity during rain will increase the threat of the treatment plant producing unsafe drinking water during these periods. Examples of such activities might be major earthworks, or developments that increase the volume or flow of water through a catchment, so entraining more particulate matter.

For more information on water treatment, refer to, for example Williams and Culp (1986) or Letterman (1999).

5.5 References

Letterman RD. 1999. Water Quality and Treatment (5th edition). American Water Works Association, McGraw-Hill Inc, New York.

WHO. 2004. Guidelines for Drinking-water Quality, 3rd ed., Vol 1, World Health Organization, Geneva.

Williams RB, Culp GL. 1986. Handbook of Public Water Systems. Van Nostrand Reinhold Co, New York.

[ Previous|Next