Treatment and Chemical Application

Chapter 5: Treatment

This chapter describes design considerations for various water treatment processes. The design of treatment processes and devices should depend on evaluation of the nature and quality of the particular water to be treated, seasonal variations, the desired quality of the finished water and the mode of operation planned.

5.1 General

For the design of a new plant, the designer should use pilot plant data that has been accumulated over a period of time that at a minimum would cover winter and summer operations, and over a wide range of historically demonstrated raw water conditions (at a minimum winter and summer). Where no such data are available, treatability studies should be conducted covering the full range of water quality conditions. Care should be exercised in interpreting treatability study data since they may not provide information directly appropriate for scale-up of attributes such as reaction and retention times and filtration rates. Samples should be taken at the proposed intake location and depth to establish a data base for design. The design of the process units should become progressively more conservative as the quantity and reliability of experimental or operating data decreases. Consideration of variable surface water quality due to seasonal changes can be crucial in the selection of processes and equipment.

In some cases, design parameters may be determined by other means such as experience with similar water in other locations or literature searches, if properly documented.

When designing an expansion of an existing water treatment plant, the designer should determine the treatment needs for the specific raw water source, as described in Section 4.1 General, based on operating data accumulated over a substantial period of time (i.e., at least three years). Where a particular water quality parameter is the principal factor in treatment design, sufficient data should be gathered for proper statistical analysis.

The designer should consider water quality and regulatory requirements together with the following site specific treatment considerations:

• Pathogen removal/inactivation;
• Control of disinfection by-products (DBP);
• Control of health-related parameters;
• Control of aesthetic parameters;
• Flexibility to respond to 'emerging' or newly identified and detectably present contaminants;
• Potential for accelerated corrosion or scale formation; and
• Readily available biological nutrient levels in the treated water.

Other considerations include operability, reliability, flexibility for expansion, site limitations, social considerations (impact on neighbours and surrounding area), minimizing the impact to the natural environment and economics (both capital and operation/maintenance costs).

5.1.1 Blending of Dissimilar Waters/ Treatment Changes

If a new water supply source is brought into service with the existing supply, it is possible to create water quality problems when the dissimilar waters are blended within the distribution system. Likewise, changes to treatment and/or disinfection that alter the chemical characteristics of the treated water may also have an adverse impact on the distribution system and household plumbing. For example, if surface water with higher dissolved oxygen levels is blended with ground water, iron or manganese can be precipitated. Another potential impact is increased corrosivity in the zone served by the new/ changed source of water supply. The designer should consider the potential water quality impacts of adding a new source or changing the existing one and provide additional/ modified treatment if necessary.

5.2 Pre-engineered water treatment components

Pre-engineered water treatment components are normally modular process units which are pre-designed for specific process applications and flows. Multiple units may be installed in parallel to accommodate larger flows.

Pre-engineered treatment components are especially applicable for small systems where individually engineered treatment plants may not be cost effective. Factors to be considered when selecting a pre-engineered water treatment component include:

• Demonstration of treatment train/unit process effectiveness under all raw water conditions and system flow demands, especially for winter conditions and northern waters;
• Means to optimize treatment and the flexibility to handle the process residuals generated;
• Sophistication of equipment and the reliability and experience record of the proposed treatment equipment and controls;
• Operational oversight that is necessary (i.e., full time operators or automation plan);
• Formal commissioning, start-up and follow-up training, operations and maintenance manuals and troubleshooting available from the manufacturer or contractor;
• Manufacturer warranty, replacement guarantee and confirmation of meeting performance objectives;
• Timely availability of parts and service; and
• Estimated annual operating and maintenance costs.

Pre-engineered treatment components require significant engineering and integration with other components, such as chemical feed and storage systems, building, electrical and plumbing systems, as well as instrumentation and controls.

5.3 Screening

Raw water screens should be provided for removal of large solids, with the screen mesh size and materials of construction consistent with the raw water quality. A screen mesh size of 10 mm (0.4 in) is common (a smaller size is typically required for membrane applications). The screen should be sized for a maximum velocity of water through the screen of 0.6 m/s (2 ft/s) regardless of water level in the screen well. Where parallel screens are used, the maximum velocity with one screen out of service should not exceed 0.9 m/s (3 ft/s).

A minimum of two screens should be provided. Depending on the size of the plant, the design may consider provision of either two fixed screens in series, two rotating mechanical screens in parallel, or one rotating screen plus one fixed screen in series. In each case, the designer should ensure that sufficient space exists above the screen to permit removal.

Fixed screens should have suitable lifting lugs on the screen, and a lifting hook or beam positioned above the screen to assist in removing it. When it is intended that screens be cleaned manually, the designer should pay particular attention to the size of screen sections and materials of construction to ensure that screen removal and handling can be readily accomplished.

Rotating screens should have either automatic or manual advance mechanisms arranged so that complete washing of the screen is accomplished. Head-loss metering equipment should be installed to monitor and react to excessive head loss, and for automatic systems, trigger cleaning. The screen rotation should be such that even wear is obtained from all sections.

For either type of screen, washing facilities should be provided using high pressure water through an appropriately sized line. Screen washings should be diverted to a holding tank, preferably with a basket type screen, so that screen wastes can be simply dewatered prior to adequate disposal. Screen wastes should not be returned to the raw water well.

Where micro-screens are provided for removal of algae, larger pore size pre-screens should be provided to protect the mesh of the micro-screens.

Screen wells should be of watertight design with provision made either through valving or stop logs for isolation of the well for cleaning or inspection. Where there are multiple screen wells, means for isolation of these wells for maintenance purposes should be provided to ensure continued operation of the plant. The screen well should be covered with a curb around any floor openings to prevent water running from the floor into the well. Wells should be provided with differential level indicating devices to permit the headloss across the screen to be determined to assess the need for cleaning.

The designer should be aware of the potential of frazil ice formation and plugging of screens with a shallow depth of intake. Protection from this may best be achieved by installation of a boom enclosing the area over the intake to accelerate ice cover formation. The provision for back flow and/or local heating may also be considered. Refer to Section 4.2 Surface Water for information regarding the design of intake facilities.

5.4 Coagulation & Flocculation

5.4.1 General

The effective removal of colloidal/micron-sized particles using granular media depth filters requires chemical destabilization of suspended particles so they will agglomerate into settling floc or adhere to the media. Design considerations for coagulation and flocculation processes include: the type, concentration and dosage range of coagulants, coagulant aids or flocculant aids; rapid mixing of the chemicals into the water; flocculation methods, intensities and time; and downstream processes.

5.4.2 Rapid Mixing/Coagulation

The design objective of the rapid mix/coagulation step is to provide high intensity mixing to thoroughly distribute the coagulant chemical in advance of the hydrolysis reaction that takes only seconds to complete. Coagulation may be achieved either in a separate process tank or by the use of an in-line mixer. The detention period in the mixing zone should be minimized and limited to no more than thirty seconds.

Static mixing devices can be used to provide effective mixing only where the flow is constant and close to the design maximum of the mixer. When selecting a static mixer, the designer should take into consideration that minor reductions in flow rate through these mixers can result in significant reductions in the mixing power delivered. This may reduce the flexibility to operate treatment effectively at reduced throughput. Power mixers are preferred where flow is expected to vary. The designer should ensure easy access for maintenance and replacement of the rotating seal of power mixers.

Chemical diffusers are a component of all rapid mixing systems. The designer should be aware that solids-forming reactions that may plug the diffuser can occur and should make appropriate provisions for removal and cleaning.

Typically, a rapid mixer with a mixing intensity velocity gradient (G value) in the order of 1000 s-1 is effective. Adding coagulants to raw water wells and allowing pumping units to perform the mixing process is not recommended. Where alkalinity or pH adjusting chemicals, powdered activated carbon (PAC) or potassium permanganate use is anticipated, addition of these should take place 3 to 5 minutes upstream of coagulant addition.

5.4.3 Flocculation

Following coagulation, flocculation is used to enhance the collisions of destabilized colloidal particles and their enmeshment into settleable and filterable floc sizes. Polymer flocculation aids and activated silica should not be subjected to high shear mixing. Provisions should be made for separate addition downstream of coagulant mixing. A delay period of 3 to 5 minutes is recommended. The mixing should be thorough enough to provide opportunities for the particles to collide but also gentle enough to prevent the flocculated particles from breaking apart.

Required detention time for adequate flocculation is variable depending on water temperature and the type of downstream processes. When sedimentation is included, detention times of 25 to 30 minutes are usually sufficient in summer. When water temperatures are <5°C (<41°F), floc formation is slower and longer detention times of 30 to 40 minutes or longer may be needed. For direct filtration, the detention time required is usually less, typically 15 minutes. Even shorter times may be adequate for coagulation/flocculation for membrane filtration. However, if the flocculation time prior to membrane filtration is too short, or control of pH, alkalinity and buffering capacity of the water is inadequate when using aluminum based coagulants, dissolved aluminum concentrations in the permeate may exceed the Ontario operational guideline and/or may prematurely foul the membrane.

Typically, G values of 10 to 70 s-1 are needed for successful flocculation. Tapered flocculation (reducing G in each stage) is desirable, typically designed as three or four sequential process tanks. Lower velocity gradients are required for the more fragile colour floc than for flocculated suspended material (turbidity). Higher velocity gradients are needed for direct filtration to produce denser pin-point floc.

Optimum G and Gt (incorporating the time that the mixing intensity is applied) values are best determined by pilot studies. Jar testing does not always involve back-mixing which is typical for flocculation processes and is therefore a limited guide for full scale design.

To permit flexibility of operation and for maintenance purposes, two separate flocculation tanks should be provided as a minimum. To prevent short-circuiting, each tank should be divided into at least two stages. The design of the basin inlet and outlet should consider short-circuiting and shearing of floc. A drain and/or pumps should be provided to handle dewatering and cleaning. A superstructure (i.e. cover or enclosure) over the flocculation basins is needed.

Mixing may be provided either mechanically or hydraulically, provided that sufficient flexibility of operation is possible, and that G values can be varied to allow for optimization of the process. Where mechanical agitation is provided, submerged bearings are not recommended, and all submerged parts should have sufficient corrosion resistance to withstand long term use with coagulated water.

Flocculation and sedimentation basins should be as close together as possible. The velocity of flocculated water through pipes or conduits to settling basins should not be less than 0.15 m/s (0.5 ft/s) or greater than 0.6 m/s (2 ft/s). Flocculated water should never be pumped between the flocculation and sedimentation units as this will break floc. The designer of pipes and conduits should minimize changes of direction in order to avoid turbulence.

5.5 Clarification

5.5.1 General

The design objective of the clarification process is to reduce the settleable solids loading on subsequent filtration processes. A minimum of two clarifier/sedimentation tanks is recommended. Where only one sedimentation tank is provided, sufficient finished water storage should be provided to allow continuous water supply while the clarifier/settling tank is out of service. Alternatively, provision should be made to produce water without clarification during tank cleaning periods. It is recommended that a by-pass pipe or conduit be provided.

Clarification processes can be categorized into general types: horizontal flow sedimentation basins, upflow reactor and sludge blanket clarifiers, adsorption clarifiers, dissolved air floatation and ballasted sand or high-rate microsand process. High rate settlers are modified sedimentation tanks or clarifiers with plate or tube modules placed into the basin to increase the settling area and reduce the distance flocs have to fall.

The primary design parameter is the surface overflow rate. The type of clarification process, the performance required in terms of clarified water suspended solids, generally measured as turbidity, the type of flocculated material generated prior to clarification and the temperature of the water all dictate the optimum overflow rate.

A superstructure (i.e. cover or enclosure) over the tanks is needed. If there is no mechanical equipment in the tanks and provisions are included for adequate monitoring under all expected weather conditions, a cover may be adequate. Roof drainage should be provided and should not discharge into the tank. The designer should allow for the possibility of ice formation within settling tanks which could fall and cause damage to submerged tubes or other components within the tank if the water level is dropped. Corrosion-resistant materials should be used for tanks, piping and appurtenances.

Inlets and outlets should be designed to ensure that water is distributed evenly across the clarifier/settling tank at uniform velocities to minimize short-circuiting. In selecting the sludge collection system, the designer should consider site-specific weather conditions (potential for ice formation), the nature and quantity of suspended solids in the raw water, the type of coagulant(s) used, the shape of the tanks and provision for installation of high-rate settlers.

Drainage systems should allow the tank to empty within a reasonable time (e.g., 8 hours).

Safety of the employees must be considered in the design of tanks. As a minimum, the design should conform to Ontario Ministry of Labour requirements, other applicable laws and regulations of the Province and local municipal building department requirements. Confined space entry requirements should be considered. Ladders, ladder guards, railings, handholds and entrance hatches should be provided where applicable. The design should incorporate easily accessible fall arrest systems for use by employees or emergency response workers. Openings into tanks should be curbed and covers should have a locking device. Additional, appropriately sealable, small openings into the tanks may be appropriate for venting, testing purposes such as dye tests for short circuit detection, or observation of settling characteristics.

5.5.2 Horizontal Flow Basins (Sedimentation/Settling Tanks)

Typically, surface overflow rates (SORs) for sedimentation tanks are from <1.0 to 2.4 m/h (<0.4 to 1.0 USgpm/ft²). Low rates are normally needed for colour removal and high rates are suitable for turbidity removal. Where water temperatures are consistently lower than 10°C (50°F), SORs should be toward the lower end of the range. For plant capacities less than 10,000 m³/d, where sedimentation efficiencies are frequently lower than in larger plants, SORs may need to be reduced by 15 to 25% to achieve the desired results on a regular basis. The designer should also consider the following parameters:

• Water depth of 3 to 4.5 m (10 to 15 ft);
• Mean flow velocity of 0.3 to 1.1 m/min (1 to 3.6 ft/min);
• Length to width ratio minimum 4:1;
• Water depth to length ratio minimum 1:15; and
• Weir loading 9 to 13 m³/(m·h) (12 to 18 USgpm/ft).

Maximum entrance velocities should not exceed 0.6 m/s (2 ft/s). Fixed or adjustable baffles should be provided as necessary to achieve the maximum potential for settling. In evaluating different inlet baffling methods or hydraulic scale model studies, the designer should ensure that the maximum number of ports is provided, that the ports uniformly distribute flow across the baffle wall and that the headloss through the ports allows for equalization of flow distribution across the entire cross section of the tank inlet with minimum floc breakage. The water exiting the settling tank should be uniformly collected across an area that is perpendicular to the flow direction either by a submerged pipe or across a weir.

The velocity of the flow into the submerged pipe or across a weir will depend on the individual design. The use of submerged orifices is recommended in order to provide a volume above the orifices for storage when there are fluctuations in flow. Submerged orifices should not be located more than m (3 ft) below the water level. Where submerged outlets are used, each tank should be provided with a suitably sized overflow or other means to prevent flooding. Any overflow should be located so as to be readily visible.

Sludge collection systems should ensure full tank coverage. Where it is proposed to remove sludge manually, the tank bottom should be sloped, typically 1:100, toward the inlet end. Flushing lines or hydrants should be provided and should be equipped with appropriate backflow prevention devices. For sludge removal by travelling siphon or scraping mechanisms, the tank bottom should be flat. Where sludge is to be removed by pumping from sludge hoppers, the hopper design should be consistent with the flow characteristics of the sludge produced.

Sludge draw-off methods should take into account that sludge loadings near the tank inlet may be substantially higher than at other locations. Sludge withdrawal piping should be designed so that material withdrawn can be observed to ensure that sludge rather than settled water is being removed. Sludge pipes should be not less than 75 mm (3 in) in diameter and arranged to facilitate cleaning. The entrance to sludge withdrawal piping should be designed to prevent clogging. Valves should be located outside the tank for accessibility. All valve operators which are not within buildings should be tamper proof with provision for locking.

5.5.3 Solids Contact, Upflow Sludge Blanket & Reactor Clarifiers

Solids contact, upflow and reactor clarifiers are proprietary settling units that have their basic sizes and associated equipment pre-established by the manufacturers based on flow. Solids contact, upflow sludge blanket or reactor clarifiers are most efficient where water characteristics, especially temperature, do not fluctuate rapidly, and where flow rates are effectively constant and operation is continuous. In the evaluation of proprietary settling units, the designer may consider the following factors.

Clarifiers should be designed for the maximum uniform flow rate and should be adjustable to react to gradual changes in flow and water characteristics.

Effective back-mixing devices should provide good mixing of the influent water (raw water plus coagulant) with previously formed sludge particles and prevent deposition of solids in the mixing zone. Depending on the design of the clarifier, a separate rapid mixing process to distribute the coagulant uniformly throughout the process stream upstream of the clarifier may be required.

The flocculation and mixing detention time should not be less than 30 minutes at the expected design maximum flow. If applicable, flocculation equipment should be adjustable (speed and/or pitch) and the clarifier design should provide for coagulation in a separate chamber or baffled zone within the unit.

The units should be equipped with either overflow weirs or orifices constructed so that water at the surface of the unit does not travel more than 3 m (10 ft) horizontally to the collection trough. Weirs should be adjustable, and the summed total lengths should be at least equivalent to the perimeter of the tank. Weir or orifice loading rates will depend on the individual design of the clarifiers. Either should produce uniform rise rates for the entire area of the tank.

Surface overflow rates range from 1.2 to 6.0 m/h (0.5 to 2.5 USgpm/ft²), depending on the design of the clarifier and the water being treated. The high rates are for clarifiers which include plate or tube settlers. Low rates are normally needed for colour removal and higher rates are suitable for turbidity removal.

Recirculation impellers should have an adjustable speed ratio of 1 to 4. Rake speed should be variable from 0.3 to 4.0 m/min (1 to 13 ft/min). Where the proposed operation is "stop-start" mode, the design should allow sludge recirculation to continue when raw water flow stops to prevent process upsets of sludge recirculation type clarifiers. The raw water inlet valve should be of the slow opening type operating over not less than one minute to prevent disruption of the floc blanket.

Sludge removal design should include:

• Sludge pipes greater than or equal to 75 mm (3 in) in diameter and arranged to facilitate cleaning;
• An entrance to sludge withdrawal piping that prevents clogging;
• Valves located outside the tank for accessibility; and
• Observation, sludge density monitoring, sampling and control of sludge being withdrawn from the unit.

Either internal or external concentrators should be provided in order to obtain a concentrated sludge with a minimum of wastewater. Typically, total water losses should not exceed 5%. Solids concentration of sludge bled to waste should be ≥3% by weight.

Discharge from blow-off outlets and drains should be treated as wastewater. Cross-connection control should be included for the drinking water lines used to backflush sludge lines.

Clarifiers should be covered either by locating them within the plant or by the use of a separate cover with personnel access to permit visual inspection of the treatment. Where open top clarifiers are proposed, the equipment should be properly weatherproofed and the rake mechanisms should be equipped with torque switches to prevent overloading. Ice blockage of effluent launder orifices may occur unless these orifices are sufficiently covered to remain ice-free, thus increasing the operating depth of the clarifier.

5.5.4 Tube or Plate Settlers

Tube or plate settlers should be inclined at an angle of 55° to 60°. Settling tanks with tube or inclined plate settler units should be designed to ensure uniform flow distribution into an entire unit, to minimize short-circuiting and to maintain velocities suitable for settling within the unit [an average velocity of 0.15 to 0.2 m/min (0.5 to 0.7 ft/min) is normally used for settling alum floc]. An approaching flow velocity of approximately 0.6 m/min (2 ft/min) should be used in the settling tank upstream of the tube or plate settler unit. SORs for tanks with tube/inclined plate settlers range from 2.5 to 5.0 m/h (1.0 to 2.0 USgpm/ft²) where the effective settling area is the footprint area (i.e., before the plates or tubes are installed).

The designer of high rate settlers units should consider the following: settling velocity and characteristics of the suspended solids; flow velocity within the settler unit; surface loading; selection of the appropriate sludge removal equipment to be installed underneath the settler unit; spacing of launders to be installed above the settler unit with weir loadings of 3.7 to 7.5 m³/m·h (5 to 10 USgpm/ft).

Provisions should be made for cleaning and/or removal of plates or tubes and sludge removal. Flushing lines should be provided to facilitate maintenance and should be properly protected against backflow or backsiphonage. Drain piping from the settler units should be sized to facilitate a quick flush of the settler units and to prevent flooding.

5.5.5 Dissolved Air Flotation

Dissolved air flotation (DAF) offers special advantages with high colour, low turbidity raw water, or water with high algae content which causes floc to settle very slowly or to float upwards. The retention time and loading rates for dissolved air flotation units depends largely on the water being treated, the nature of the contaminant being removed, the chemicals used and the design of the DAF process. Traditional loading rates have been in the 10 to 12 m/h (4 to 5 USgpm/ft²) range; however, higher loading rates, up to 29 m/h (12 USgpm/ft²) may be used if confirmed through appropriate pilot testing.

The tank length should be no greater than 12 m (39 ft) to control the density of the bubble blanket. Tank depth should be 1.5 to 3.0 m (5 to 10 ft); greater depths are recommended for high algae loads. Flow velocities should be designed to limit scouring of the float from below. An inlet baffle should create a contact zone volume large enough to provide good floc/bubble contact time, but not so large as to encourage short circuiting. The angle of the baffle should be 60 to 90°, depending on the hydraulic loading rate.

The air saturated recycle flow should be adjustable and introduced at a location which ensures even distribution of the released air at the tank influent. The recycle ratio should be between 5 and 12% of inlet flow. The air flow should also be adjustable and the air injection designed to ensure an even distribution of air across the inlet baffle. Bubble diameter should be between 10 and 100µm. Saturation pressure should be 415 to 725 kPa (60 to 105 psi).

The DAF effluent (subnatant) should be removed from a submerged location near the basin floor, usually by way of an underflow baffle or perforated pipe laterals. The float- sludge may be removed hydraulically or mechanically. In selecting the float-sludge removal system, the residuals handling system should be considered.

The designer may consider the use of DAF on a seasonal basis (e.g., for algal blooms).

5.5.6 Ballasted Flocculation & Clarification

This is a proprietary treatment system which operates in up-flow mode and uses microsand-enhanced flocculation and lamellar settling for clarification. The process may allow very high loading rates, and therefore a smaller footprint, in comparison to conventional flocculation and sedimentation. Typical hydraulic loading rates are 35 to 73 m/h (14 to 30 USgpm/ft²). This process may need a specific combination of chemicals for effective treatment depending on a raw water quality. A means to recycle and clean the sand ballast for reuse in the process should be provided. The designer may consider this process, particularly for situations where the site area is limited, for the clarification process in the main treatment train or where backwash water is to be clarified before recycling (Section 11.2.3 Membrane Filtration).

5.5.7 Roughing Filters

Roughing filters are an alternative clarification process which may be operated in either up-flow or down-flow mode. They create a zone of laminar flow and suspended particulate deposits on the filter media. To clean the filter, the media is agitated to loosen the solids, or in the case of upstream roughing filters, is such that the flow can be controlled to allow water to flow backwards (down) through the filters. Roughing filters are frequently used upstream of slow sand filters when the source water requires pre-treatment to remove coarse particles that could lead to unacceptable rates of headloss development.

5.5.8 Adsorption Clarifiers

Adsorption clarifiers operate on the principle of granular media flocculation. The flocculation and clarification process takes place as the coagulated water travels upwards through a buoyant plastic media. Flocculated solids are trapped in the media. These systems are proprietary and use a combination of hydraulic flocculation, roughing filtration and rapid rate filtration. SORs are in the range of 19.5 to 25.5 m/h (8 to 10 USgpm/ft²). Air scour should be provided for cleaning the filters. Pilot testing is recommended for site specific applications.

5.6 Granular media depth filtration

5.6.1 General

The designer should evaluate the following interrelated factors when designing a granular media filtration process: the requirements of the filtered water quality specified in the Procedure for Disinfection of Drinking Water in Ontario (Disinfection Procedure); site specific conditions including raw water quality; pre-treatment process(es) and associated chemical application points; plant size/capacity; materials of construction; type of filtration technology; filtration rate; control of filtration rate; type of filter bed including the media size, thickness of media layers, and number of independent filters; available headloss for filtration; type of media support and underdrain; type of filter wash system; and filter-to-waste. Treatment and/or disposal of waste residuals should also be considered in the selection and design of the filtration process.

The number of filters/filter trains provided should depend on the process selected; i.e., with or without prior clarification, the method of plant operation, the method of filter control and the quantity of available storage.

At least two filters should be provided, each capable of independent operation and backwashing. Where only two units are provided, the filters should be capable of meeting the plant gross design capacity at the design filtration rate and, for security of supply, consideration should be given to having additional filter area so that each filter is capable of meeting the majority of plant gross design capacity at the design filtration rate.

Where more than two filter units are provided, the filters should be capable of meeting the plant gross design capacity at the design filtration rate and each filter should be capable of independent operation and backwashing.

To avoid the potential for turbidity breakthrough, the designer should select an adequate number of filters and area of each filter bed so the filtration rate can remain the same or not increase substantially (less than 10% gradual change in hydraulic loading) during the backwashing of filters.

The filtration rate and terminal head loss for a particular type of filter and filtration medium design should be selected considering total required area of the filter bed, the available hydraulic loss during filtration, the anticipated terminal head loss prior to turbidity breakthrough in the filter bed and the anticipated filter run.

All filters should be equipped as follows:

• Means for obtaining influent and effluent samples;
• Indicating flow meter and flow control to each filter/filter train;
• Continuous effluent turbidity measuring and recording device; particle monitoring equipment may be useful in enhancing overall treatment operations;
• Indicating loss of head gauge;
• Provisions for filtering-to-waste with appropriate measures for backflow prevention or operational procedures to achieve the same water quality results;
• Wall sleeves providing access to the filter interior at several locations for sampling or pressure sensing or equivalent other devices; and
• Pressure hose [25 to 37 mm (1 to 1.5 in)] and storage rack at the operating floor for washing filter walls (gravity filters).

The filter structure should be designed to provide for:

• Vertical walls within the filter;
• No protrusion of the filter walls into the filter media;
• Cover or enclosure by superstructure;
• Head room to permit normal inspection and operation;
• Minimum depth of filter box of 2.5 m (8.2 ft);
• Minimum water depth over the surface of the filter media of 1 m (3 ft) [1.m (5 ft) for high rate filtration] to prevent air binding due to dissolved air coming out of solution in the filter bed;
• Effluent piping designed to prevent backflow of water and air to the bottom of the filters and to provide minimum operating conditions for flow meters;
• Prevention of floor drainage to filters with a minimum 100 mm (4 in) curb around the filters;
• Prevention of flooding by providing overflow;
• Maximum velocity of treated water in pipe and conduits to filters of 0.6 m/s (2 ft/s);
• Cleanouts and straight alignment for influent pipes or conduits where solids loading is heavy or following lime-soda softening;
• Washwater drain capacity to carry maximum backwash flow;
• Walkways around filters, to be not less than 600 mm (24 in) wide;
• Safety handrails or walls around all filter walkways; and
• Construction to prevent cross-connections and common walls between potable and non-potable water.

Washwater troughs should be constructed to have:

• The bottom elevation above the maximum level of expanded media during washing;
• A 50 mm (2 in) freeboard at the maximum rate of wash;
• The top edge level all at the same elevation (adjustable weirs are recommended);
• Spacing so that each trough serves the same filter area;
• A horizontal travel distance for suspended particles to reach the trough of 1 m (3 ft) for sand media; and
• Trough spacing of 1.8 to m (6 to 10 ft) for dual media with anthracite or granular activated carbon (GAC).

The designer should refer to the Disinfection Procedure for more information on the design, operation criteria and performance requirements of specific types of granular media filtration processes and applicable pathogen removal credits.

The design should include provision for modifications of the filters or the addition of more filters so that future construction will have minimal impact on water treatment plant operations.

5.6.1.1 Flow Control

Flow control can be designed using different strategies; however, the system should control the flow to each individual filter, apportion the total flow among the individual filters and accommodate rising head loss through each individual filter run. The designer should consider cost, complexity and reliability when selecting a control strategy.

There are two basic modes of filtration control:

• Constant rate filtration; and
• Declining rate filtration.

In constant rate filtration, the flow to each filter should be maintained at as constant a rate as possible with clearwell storage absorbing fluctuations in demand and in filter output (i.e., when a filter is taken out of service for backwashing). This is typically accomplished by a flow meter and a flow modulation valve on each filter effluent pipe, or constant level filtration with equal flow splitting inlet weirs, a water level sensor and a flow modulator valve.

5.6.1.2 Filter Media

Filter media should conform to NSF/ANSI Standard 61: Drinking Water System Components - Health Effects and the applicable AWWA Standard B100: Filtering Material or AWWA Standard B604: Granular Activated Carbon. When GAC is used, provisions should be made for periodic replacement or regeneration.

The selection of media type, size (effective particle size and uniformity coefficient), distribution, depth and L/d ratio (L=depth, d=filter media effective particle size) should be such that, in operation, the filter reaches its design terminal headloss at approximately the same time as either turbidity or colour breakthrough occurs, based on whichever is the controlling process parameter. The media selection also depends on the concentration and type of suspended solids to be removed by the filter.

For traditional dual media filter designs, the media should consist of a lower level of silica sand, not less than 200 mm (8 in) deep, and an upper layer of anthracite coal or GAC not less than 450 mm (18 in) deep. The designer should be aware that direct filtration is more sensitive to variations in media selection. The media selection should ensure fluidization of each layer of media during backwashing and subsequent re-stratification of the media.

In addition, some types of filter bottom/underdrain systems require supporting media to prevent the passage of filter media through the filter bottom. Typically, gravel support layers in three or four overlapping gradations should be provided. The smallest size is usually in the 2 to 5 mm (0.08 to 0.2 in) range; the largest size and the depth of the layers will depend on the type of filter bottom used. Other media support methods (gravel-less systems) are also available.

Alternate configurations, including multi-media, coarse deep beds and proprietary media designs, should be pilot tested to ensure their suitability, or have appropriate documentation of past performance.

5.6.1.3 Underdrains

The most important functions of the filter bottom or underdrain are to provide an even rate of filtration over the entire area of the filter and uniform distribution of backwash water and/or scouring air. The filter bottom should be designed so that all head losses on backwashing occur at the final openings to ensure an even distribution of washwater.

Porous plate underdrains should not be used where iron or manganese may clog them, with waters softened by lime or with water susceptible to algae growth or biofouling.

5.6.2 Rapid Rate Gravity Filters

The use of rapid rate gravity filters requires pre-treatment (chemically assisted filtration). Filters should be designed to achieve an individual filter effluent turbidity of <0.1 NTU, other than during the ripening period when the effluent should be controlled. The rate of filtration should be determined through consideration of such factors as raw water quality, degree of pre-treatment provided, filter media type(s), specifications and depths, and the competency of operating personnel. For traditional dual media filter designs, a maximum filtration rate of 11.7 m/h (4.8 USgpm/ft²) is recommended, although filter rates of up to 20 m/h (8.1 USgpm/ft²) have successfully been achieved. Filtration rates of 11.7 m/h (4.8 USgpm/ft²) may not be achievable with floc formed from high colour water. For all filter designs, filtration rates greater than 11.7 m/h (4.8 USgpm/ft²) should be confirmed through pilot testing. Pilot testing in cold water conditions is also advisable to establish acceptable rates.

5.6.2.1 Backwash Systems

A sufficient volume of water should be available for backwashing all filters every 24 hours. The backwash rate should be variable, with the maximum rate designed to provide 50% expansion of the filter media bed at the highest water temperature. Generally, the rate needed for this expansion is 37 to 50 m/h (15 to 20 USgpm/ft²). Lower backwash rates are required in GAC filters or contactors, since GAC is less dense than anthracite.

Lower backwash rates are also needed to fluidize the bed at the beginning of the wash and to allow the media to re-stratify at the end of the wash. The design should allow for a backwash duration of at least 15 minutes. An effective backwash of a filter is required before return to service and may require more than 15 minutes depending on conditions. For filters with air scour, a lower maximum wash rate and a shorter duration may be sufficient.

Filtered water should be used for backwashing and provided at the required rate by a minimum of two washwater pumps (one duty and one standby). The use of high pressure sources with pressure reducing valves is not recommended as failure of pressure reducing valves may disrupt filter media which would then need to be re-stratified.

A flow regulator, flow meter and flow indicator should be provided on the main backwash header. An air release valve should be placed at the high point of the header. The system should be designed so that rapid changes in backwash water flow do not occur.

Backwashes should be operator initiated; or alternatively, automated systems should be operator adjustable.

5.6.2.2 Supplementary Wash & Air Scour

A supplementary surface/subsurface wash system or air scour should be provided. A supplementary wash could be either a system of fixed nozzles or a revolving-type apparatus. All such systems should be designed with:

• Provision for water pressures of at least 310 kPa (45 psi) or as specified by the manufacturer;
• Backflow prevention to prevent backsiphonage if connected to the treated water system; and
• Flow of 4.9 m/h (2.0 USgpm/ft²) with fixed nozzles or 1.2 m/h (0.5 USgpm/ft²) with revolving arms or as specified by the manufacturer.

Air flow for air scouring the filter prior to backwashing should be 0.9 to 1.5 m³/(min·m²) (3 to 5 ft³/min·ft²) of filter area when the air is introduced into the underdrain. A lower air rate should be used when the air scour distribution system is placed above the underdrains. The air should be free from contamination. Oil-free compressors should be used. Air scouring should be followed by a fluidization wash sufficient to restratify the media.

Air scour distribution systems should be placed below the media and supporting bed interface; alternatively, if placed at the interface, the air scour nozzles should be designed to prevent media from clogging the nozzles or entering the air distribution system. Air scour systems or nozzles should provide for even air distribution. Piping for the air distribution system should not be flexible hose which will collapse when not under air pressure and should not be a soft material which may erode at the orifice opening with the passage of air at high velocity.

Air delivery piping should not pass down through the filter media nor should there be any arrangement in the filter design which would allow short circuiting between the applied unfiltered water and the filtered water. Consideration should be given in the design to maintenance and replacement of air delivery piping.

5.6.3 Direct Filtration

For surface water treatment plants using chemically assisted granular media filtration, the designer may consider omitting the clarification process where raw water average turbidity is about 5 NTU and does not exceed 20 NTU during storm excursions, colour is below 40 TCU, and algae is below 2,000 areal standard units (ASU). This decision should be based on a risk analysis in consideration of the multiple barrier approach (Section 3.2.2 Risk and the Multi-Barrier Approach) and an economic evaluation of the process with and without clarification. A clarification process provides a partial pathogen barrier and enhances the overall disinfection efficiency and stability of the treatment train by allowing extra time for treatment process adjustment and control.

Coagulant aided pre-filtration solids removal using roughing or coarse media filters can extend the application of direct filtration to a slightly broader range of source waters.

5.6.4 Rapid Rate Pressure Filters

Pressure filtration with chemically assisted coagulation and flocculation (in-line pressure filtration systems) is recognized as 'direct filtration' for disinfection removal credits as described in the Disinfection Procedure. It should only be considered for low colour raw water with turbidity not exceeding 20 NTU and where incoming water quality is very consistent and rapidly forms a robust floc in cold conditions.

Pressure filtration systems usually include coagulant (and polymer or coagulant aid) injection in the pipe under pressure, a static mixer, a pressure vessel with or without large grain media where flocculation occurs, followed by rapid rate pressure filters.

Pilot testing should be conducted covering periods of low and high temperature as well as periods of high and low turbidity and colour (high colour is often the cause of system failures). The pilot study should use the allowable flow and its variations, a site specific travel time from coagulant injection point to flocculation (flocculation beyond that vessel may result in failure of the sand media filter), type of coagulant and polymer and the need for automatic coagulant dose in proportion to the turbidity and flow, where applicable.

Recommendations relative to filter media provided for rapid rate gravity filters such as rate of filtration, and structural details and hydraulics also apply to pressure filters where appropriate.

The filters should be designed to provide:

• Pressure gauges on the inlet and outlet pipes of each filter;
• Flow indicators on each filter;
• Backwash flow indicators and controls that are easily readable while operating the control valves;
• Air release valve on the highest point of each filter;
• Access hatches to facilitate inspection and repairs for filters 900 mm (36 in) or more in diameter, or handholds for smaller filters; and
• Cross-connection control.

5.6.5 Slow Sand Filtration

Slow rate gravity filtration should be limited to source water (or influent water after pre-treatment) having a maximum turbidity of 10 NTU (the turbidity should not be attributable to colloidal clay), maximum colour of 15 TCU and low algae counts. Where dissolved organic carbon (DOC) in the influent water is greater than 10 mg/L, pilot testing is essential.

Modified or enhanced slow sand filtration systems use additional pre-treatment processes prior to the slow sand filter. The water quality limitations outlined above describe the water being applied to the slow sand component of the process. Water quality limitations for raw source water prior to pre-treatment will depend on the pre-treatment processes employed.

Where organic material in the raw water is not easily biodegradable, the application of ozone (up to 1 mg/L) upstream of the slow sand filter can promote biological activity by making the natural organic matter (NOM) in the water more amenable to biological removal. It coincidentally increases dissolved oxygen, which is beneficial to microbial activity. Additional information is provided in Section 5.8.2 Biological Filters.

Slow sand filtration rates are generally in the 0.04 to 0.4 m/h range (0.02 to 0.16 USgpm/ft²). The design rate of filtration should be determined by pilot testing of the water to be treated.

Each filter unit should be equipped with a main drain and an adequate number of evenly spaced lateral underdrains to collect the filtered water. The underdrains should be so spaced that the maximum velocity of the water flow in the underdrain will not exceed 0.23 m/s (0.75 ft/s). The maximum spacing of laterals should not exceed 1 m (3 ft) if pipe laterals are used.

Media depths should typically be in the 0.75 to 1.5 m (2.5 to 5 ft) range. The effective particle size should be between 0.15 mm (0.006 in) and 0.30 mm (0.012 in) and the uniformity coefficient should not exceed 2.5. The supporting gravel should be similar to the size and depth distribution provided for rapid rate gravity filters.

The filter design should maintain a depth of water of 1.8 to 2.1 m (6 to 7 ft) above the sand. Influent water should not scour the sand surface. Each filter should be equipped with an indicating loss of head gauge for confirming active biological filtration. Slow rate filters should be designed to provide:

• Means for cleaning and/or scraping of sand;
• A cover with headroom to permit normal movement by operating personnel for scraping and sand removal operations, and to prevent exposure to sunlight;
• Adequate access hatches and ports for handling of sand and for ventilation;
• Filter-to-waste (a minimum two days flow should be considered);
• An overflow at the maximum filter water level; and
• Protection from freezing.

5.6.6 Diatomaceous Earth Filtration

Diatomaceous earth (DE) filters can provide effective pathogen removal and can be used for polishing following other filtration processes. The use of DE filters should be limited to source water (or influent water after pre-treatment) having a maximum turbidity of 20 NTU and maximum colour 15 TCU. Filtration rates should be determined through pilot testing. The designer should refer to the Disinfection Procedure and the AWWA Manual of Water Supply Practices M30 Precoat Filtration for more information on the design criteria for DE filtration processes.

5.7 Straining filtration processes

5.7.1 General

The basic principles and design factors described in Section 5.6 Granular Media Depth Filtration should be considered for designing filtration processes using other than granular media. The designer should always refer to the Disinfection Procedure for more information on the design and operation criteria of specific types of filtration processes and applicable pathogen removal credits.

5.7.2 Membrane Filtration

5.7.2.1 General

The terminology used for membrane system components and processes is not consistent among all suppliers/manufacturers. Refer to Appendix A - Glossary for clarification of the terms used in this section. Membrane filtration systems are proprietary and the manufacturer should be consulted for specific design requirements. The following are general design considerations for microfiltration (MF) and ultrafiltration (UF) membrane systems most commonly used in Ontario. Nanofiltration has not yet found wide application in Ontario, although there are some installations across Canada, primarily for organics removal. Reverse osmosis membrane systems are mainly used as point-of-entry or point-of-use systems in Ontario. Experience with nanofiltration and reverse osmosis membrane systems to date has not been sufficient to provide specific design guidelines; the designer should therefore consult the manufacturer if considering either of these systems.

5.7.2.2 General Design Considerations

The design flow rate for membrane systems is the net filtered output desired from the membrane system. The designer should take into account the loss of feed water used for backwashing and/or reject stream (waste stream) and the lost production while a unit or train is out of service for chemical cleaning. The designer should consider the capability of the source to provide enough water to allow for losses and the capacity of the disposal system to handle waste volumes while meeting system demand (Section 3.6 Plant Capacity Rating).

The overall cost of the delivered water is strongly affected by membrane life. Membrane life depends on membrane fouling rates, among other factors such as oxidant exposure or transmembrane pressure. Generally the higher the design flux for a given water quality, the shorter the membrane life. This trade-off should be selected by the designer in consultation with the municipality/owner and the membrane manufacturer and included in the manufacturer performance guarantee.

Operation and useful life vary depending on type of membrane selected, quality of feed water and process operating parameters. Specific issues the designer should consider include membrane flux, water quality and temperature, cross-connection control and system reliability. It is important for the designer to test a range of different manufacturers' membranes for a given raw water source as substantially different and unpredictable performances may occur.

For most raw water sources either pressure or submerged (vacuum) type membrane systems can provide satisfactory performance. A factor that may influence the choice between vacuum and pressure systems is the performance of the system during pilot testing. In currently available equipment, submerged systems tend to accommodate larger modules than pressure systems, and have fewer valves and piping connections. Pressurized cartridge systems in contrast are compact and more easily accessible for service and can make effective use of smaller quantities of more concentrated and potentially reusable cleaning solutions.

The ability to obtain qualified operators [in accordance with the requirements of Certification of Drinking-Water System Operators and Water Quality Analysts (O.Reg. 128/04) under the Safe Drinking Water Act, 2002] should be evaluated in selection of the treatment process. The necessary operator training and the required operator certification should be obtained prior to plant start-up.

Pilot Testing

Site specific pilot testing is usually needed to select the membrane and to determine particulate removal efficiencies, fouling potential, flux, pressures and pre-treatment requirements. The specific flux for any particular membrane is a strong but predictable function of water temperature and viscosity and will be well known to the manufacturer. Cold weather performance therefore is predictable provided organic and inorganic contaminant levels and types do not change substantially with the seasons.

Fouling potential and pre-treatment needs are site specific and can be difficult to predict without pilot testing. On occasion, it may be possible without pilot work to accurately predict membrane system performance from that of nearby plants which treat very similar quality raw water. However, the membrane fouling rate is an important design consideration that can be strongly dependent on the make up of raw water organics and is difficult to accurately characterize.

A pilot study allows the designer to determine the optimal combination of flux, pressure, recovery rate and cleaning interval. Membrane systems should be designed to operate within the specific range of pressures and fluxes of the membrane.

Membrane Flux

The sustainable flux dictates the membrane area necessary to achieve the desired system capacity. To minimize the number of modules required, and associated capital costs of the membrane filtration system, the designer should strive to maximize the membrane flux without inducing excessive fouling.

The optimum flux should be based on pilot test results. Cold water significantly reduces the flux of a membrane system; hence, cold season demand may govern the required design membrane area.

System Reliability

The designing engineer should provide a system that reliably meets the water demand and pathogen removal credits for the design period without the need for major reconstruction or refurbishment.

Scheduled membrane replacement represents a major cost component in the overall water production costs. The designer should consider that membrane replacement frequency may significantly affect the overall cost of operating the treatment facility.

Designers selecting membrane systems should be aware that this is a fast developing field of technology. Production of membrane system parts may be discontinued and other manufacturers' parts may not be compatible or interchangeable. As a result the designer should consider providing ample space for expansion and installation of alternative equipment, and inform the municipality/owner of the inherent risks.

A minimum of two independent membrane filter trains should be provided. When determining the total amount of membrane area and number of membrane trains to meet system demands, the effect of having one train off-line should be included in the plans. When a train is off-line for cleaning, the remaining trains will need to be capable of operating at a higher flux rate for the duration of the cleaning cycle in order to meet system demands. Where possible, this should be avoided as operating at high flux rates may significantly accelerate deterioration of the membrane performance.

The need for redundant trains and equipment should also be considered when selecting the number and size of trains (Section 3.29 Reliability and Redundancy). The designer should consider the balance between true redundant trains and those used for balancing flow and minimizing the instantaneous peak factor of the system, which may affect the sizing and operation of upstream and downstream processes. Where a system has been designed with a fully redundant train, operational hours should be shared equally between all trains on a rotating basis.

For staged expansion of plant capacity, the appropriate infrastructure for the projected future demands should be installed, and guarantees of future replacement pricing geared to a price index, if required, together with replacement product availability assurances from the manufacturer. Stored membranes may have a limited shelf life.

Cross Connection Control

Cross connection control considerations should be incorporated into the system design, particularly with regard to chemical feeds and waste piping used for membrane cleaning, the waste stream and concentrate.

5.7.2.3 Pre-Treatment

Pre-treatment may be required prior to UF and MF systems to prevent potential fouling of the membranes and/or to reduce disinfection by-product formation, particularly where raw water TOC levels are high.

The pre-treatment unit processes that may be evaluated for integration with membrane filtration systems include:

• Pre-screening of any membrane system to protect the membranes from damage by debris. The required screen size and/or strainer should be dictated by the requirements of the membrane manufacturer;
• Oxidation to be integrated with membrane processes to assist with organics [total organic carbon (TOC) and dissolved organic carbon (DOC)] and taste and odour reduction. It is recommended that the oxidation process be introduced as far upstream of the membrane process as possible. The designer should obtain concurrence for the use of a particular oxidant from the membrane manufacturer;
• Adsorption processes are normally used downstream of the membrane process for removal of organics (TOC and DOC) and taste and odour causing compounds. When carbon adsorption processes are considered upstream of membrane processes, pilot tests should be carried out for the specific membrane and carbon grade to assess the potential reduction in membrane life due to the presence of abrasive carbon fines. When biological filtration is planned upstream of membrane filtration, optimization of pilot biological filtration over a minimum period of 4 months should be conducted before commencing membrane pilot work on the biological filter effluent; and
• Coagulation upstream of the membrane process is not needed for effective pathogen removal. Coagulant use may lead to additional expense and residuals handling costs and may require conditioning of the raw water to be effective. Where coagulant is needed, for example, for colour/dissolved organic substances removal, the coagulation process will add additional solids loading and may plug the membrane pores leading to extra cleaning requirements. On occasion, however, the use of coagulant may improve the agglomeration characteristics of raw water solids leading to longer duration filter runs and reduced chemical cleaning requirements. Utilizing one of the two processes to reduce the solids loading on the membranes can improve membrane performance which could in turn lead to less membrane surface area required and extended life of the membranes. Where coagulation is needed for a substantial part of the operating year, conventional treatment should be evaluated and cost compared.

An additional factor for the designer to consider is the effect of upstream processes on the membrane system. For example, many membranes have restricted tolerance for exposure to chlorine the use of which may be needed for mussel or taste and odour control. Other processes such as biological filters and contactors can significantly affect membrane cleaning requirements and life. Many of these processes are difficult to pilot accurately. The designer should consider the applicability/reliability of pilot results and/or consider placing these processes downstream in the treatment train.

Membranes can be effective for the removal of divalent iron and/or manganese. However, the use of a rapid acting oxidant and sufficient contact time should be allowed for chemical reaction completion and some precipitate/floc formation upstream of the membrane filter.

5.7.2.4 Backwash & Chemical Cleaning

The designer should keep in mind that chemical cleaning of membranes is required at regular intervals of a few weeks or less to slow the deterioration of available flux to acceptable levels. Failure to apply cleaning chemicals in the optimum manner can lead to rapidly declining membrane performance and a need for early membrane replacement. Backwashing and chemical cleaning frequencies, durations and procedures should be obtained from the membrane manufacturer, or be determined based on pilot study data or similar application data.

The membrane should be periodically cleaned with chemicals (both chemically enhanced backwash and recovery clean). Membrane cleaning chemicals may be highly aggressive and excessive cleaning may shorten effective membrane life. The membrane manufacturer should provide appropriate cleaning instructions to balance performance and performance degradation and the costs of cleaning for each specific installation. Care must be taken in the cleaning process to prevent contamination of both the raw and finished water system.

5.7.2.5 Integrity Testing

To routinely evaluate membrane and housing integrity and overall filtration performance, the designer should consider methods for periodic integrity testing. There are five key aspects for achieving an integral membrane system:

• Performance requirements;
• Type of integrity test;
• Integrity test criteria and settings;
• Frequency of integrity testing; and
• Management of the process and information.

There are two basic types of integrity testing: continuous indirect integrity testing and periodic direct integrity testing.

Indirect integrity testing includes on-line particle counting used as a continuous indication of the membrane integrity. In general, sustained particle counts in the filtrate should remain below 20 counts/mL. If filtrate particle counts exceed 20 counts/mL for an extended period of time, this may be an indication that a membrane fibre has been breached and should be isolated and checked for integrity. An alarm should be provided with continuous indirect integrity testing.

The designer should provide means for direct integrity testing, including such measures as pressure decay, vacuum hold, bubble point or sonic testing. The required frequency of direct integrity testing will depend on the quality of the influent raw water and the robustness of the membranes.

Integrity testing is a requirement if the membrane process is to be used for disinfection removal credits (refer to the Disinfection Procedure). Whatever integrity monitoring technique is adopted, it should be capable of confirming numerically that the required log disinfection credit is being achieved and that process train leaks are repaired to consistently achieve this performance. For those systems that have to be tested on-line during production, a filter-to-waste option should be considered in the event of a membrane integrity breach.

5.7.2.6 Ancillary Equipment

The designer should specify or ensure that the membrane manufacturer contractual commitment includes the following ancillary equipment:

Feed Water or Permeate Pumps, Blowers & Compressors

Where pumps, air blowers and compressors are employed, the number of duty pumps, air blowers and compressors required will depend on the number of process trains selected and the anticipated range of flows. A standby unit should be available for any process train in the event one of the duty units is out of service for maintenance or repair. For small systems with adequate storage, the use of "shelf-spares" in place of standby units may be considered acceptable. The designer should also consider the efficiency of pumping and blower equipment, as these are energy intensive processes and operation may be continuous or semi-continuous.

Isolation Valves & Unions

Isolation valves are required for each individual membrane assembly. The size of the individual modules is such that it is often impractical to isolate individual membrane modules. Instead, isolation valves are to be provided to isolate individual trains and membrane assemblies, or subsections of the membrane assemblies.

Piping & Automated Valves

Some membrane systems operate over a wide range of pressures and have a significant number of automated valves. Select piping materials, restraints, and actuator speed controls suitable for the intended materials, service and to prevent water hammer. The designer should ensure that the valves and piping are suitable for a wet and chlorine heavy environment. Valves and actuators should be suitable for multi-cycle operation rather than modulation/shut-off only, where required.

Chemical Feed Systems

Chemical feed systems should have standby pumping units. Refer to Chapter 6 - Chemical Application for storage and safe handling of the chemicals.

The designer should also consult the manufacturer regarding the design of HVAC systems, the provision of means for access to, removal of and repair of membrane modules, valves and instrumentation.

5.7.2.7 Monitoring Equipment

Flow Metering Systems

Flow meters should be provided to directly or indirectly continuously monitor:

• Main raw water supply line (or individual train raw water supply lines) to measure the feedwater volume entering the membrane system and for flow pacing of any pre-treatment chemicals;
• Individual permeate lines from each membrane train to measure the filtration rate and volume of each train and pace post disinfection chemicals;
• Individual reject or concentrate lines from each train to measure the flow rate and volume of waste stream water for calculating the overall recovery rate of the train;
• Individual backwash lines (or use of the permeate flow meters) to measure the backwash flow rate and volume;
• Combined filter effluent line and/or the distribution main header leaving the plant; and
• On-line means for confirming flow paced chemical additive dosing.

On-line Metering/Monitoring Systems

An on-line turbidimeter should be provided on the common feed water line to the membrane trains. On-line turbidimeter instruments should be provided on the permeate discharge from each membrane train. The provision of particle counters may be considered on a per train basis. Sample point connections should be provided at each rack or cassette for connection of a portable particle counter to aid in troubleshooting in the event of a fibre breakage.

Provisions should be made for pH and residual measurement, either on-line or at convenient sample points, on each membrane clean-in-place (CIP) tank to monitor the cleaning solution concentrations, typically citric acid. When protein fouling from biofilm on the membrane requires the use of protease enzyme solutions, strength measuring techniques, as recommended by the manufacturer, should be applied. Pressure gauges and transmitters should also be provided on each membrane train to measure transmembrane pressures for monitoring the rate of fouling and to initiate chemical cleaning, and backpulse pressures to avoid over pressurization and damage to the membrane fibres.

5.7.2.8 Residuals

The ministry should be consulted, as early as possible, when considering the use of membrane technologies, to determine environmentally acceptable options for disposal of waste streams from both pilot scale and full scale membrane plants.

Neutralization of the cleaning solutions should be provided, either directly in the process tank where the CIP has taken place, or the solutions should be transferred into a holding tank to ensure sufficient time for neutralization and monitoring prior to disposal.

Disposal of reject water and waste from chemically enhanced backwashing and recovery cleaning is also discussed in Section 11.2.3 Membrane Filtration.

5.7.3 Bag & Cartridge Filtration

This technology is designed to meet the low flow requirements of small systems. The use of bag or cartridge filters should be limited to source water (or pre-treated influent) having a maximum turbidity of 5 NTU and maximum colour of 5 TCU. Typically, cartridge and bag filters are used in series with decreasing pore size. In order to claim the 2.0 log Cryptosporidium oocyst removal credit, the cartridge/bag filtration process should meet the following criteria:

• Filter elements and housing should be certified for surrogate particle removal evaluation in accordance with testing procedures and manufacturing quality control specified in NSF/ANSI Standard 53: Drinking Water Treatment Units - Health Effects or equivalent;
• Filtrate turbidity from each filter is monitored continuously;
• Alarming is provided if differential pressures across the filter medium exceed the manufacturer rating; and
• Materials coming in contact with water conform to NSF/ANSI Standard 61: Drinking Water System Components - Health Effects.

The particulate loading capacity of these filters is low, and once reached, the bag or cartridge filter should be discarded. The operational and maintenance cost of bag and cartridge replacement and disposal should be considered when designing a system.

The design flow should be determined in consultation with the manufacturer specifications and confirmed with a pilot test. The pilot testing should be carried out with the actual size of the filter element at the design maximum flux and set up so as to provide an assurance of practical filter element life. The design should consider provisions for adding chlorine or another disinfectant at the head of the treatment process to reduce or eliminate the growth of microorganisms such as algae and biofilm on the filters. The impact on disinfection by-product formation should be considered.

Any pre-treatment filters should have filter-to-waste capability or other means to prevent blinding of the bag or cartridge filter at start-up or premature breakthrough. Filter-to-waste should also be provided for the final filter(s) and a pre-determined amount of water should be discharged to waste after changing the filters. A by-pass-to-waste line is also needed during start-up after long periods of non-use, especially on groundwater systems where biofilm forming bacteria are present. It may be necessary to adopt automatic flow redirection to a second filter train and/or discharge to waste for a short period on start-up to avoid rapid plugging of filter elements.

Pressure gauges should be installed before and after the bag/cartridge filter. The designer should ensure that either a differential pressure gauge with shut down alarm is provided or that the design pump maximum pressure is below the cartridge maximum rated differential pressure. An automatic air release valve should be installed on top of the filter housing. The flow through the treatment process should be controlled with a flow control valve and the flow measured.

A slow opening and closing valve should be included upstream of the filters to reduce flow surges. Frequent start and stop operation of the bag or cartridge filter should be avoided. To avoid frequent start and stop cycles, the following options are recommended:

• Low filtration rates which will lengthen filter run times; and
• Installation of a recirculation pump that pumps treated water back to a point ahead of the bag or cartridge filter, complete with reduced pressure principle backflow preventer to ensure there is no cross-connection between the finished water and raw water.

5.8 Soluable contaminant removal processes

5.8.1 Granular Activated Carbon Contactors

GAC contactors can be used for the removal of organic compounds including those producing taste and odour, disinfection by-products, trace contaminants, pesticides and DOC removal. GAC may be used in granular media filters or in separate contactor units.

Designers should be aware that although available grades of GAC can show good adsorption performance in bench testing, the adsorptive capacity of GAC will diminish rapidly over time due to background TOC loading. This should be taken into account when designing GAC contactors or replacing anthracite with GAC in filters, particularly for the removal of taste and odour compounds.

The designer should also be aware of the potential for desorption of organics originally absorbed onto the carbon due to competition by highly adsorbing organics in the influent, resulting in higher concentrations in the effluent. In addition, sloughing of biofilm in biologically active GAC beds may cause high concentrations of bacteria in the filtered water.

A GAC bed installed (capped) over existing filter beds can function as a filter by removing suspended matter but will also remove organic compounds. In most cases, GAC contactors are installed downstream of filters for taste and odour control. Where ozone pre-filtration treatment is applied, carbon contactors can provide effective removal of any remaining ozone traces and therefore protect downstream biological filtration steps.

Stand-alone GAC contactors (separate from filtration) may be gravity fed contactors, pressure contactors or upflow contactors. Gravity-fed contactors are similar to granular filters (Section 5.6 Granular Media Depth Filtration) but are deeper than conventional granular filters. Pressure contactors are similar to pressure filters. The designer should consider that GAC is relatively fragile and easily crushed and subsequently will be lost as fines during backwash flow; therefore, fluidized bed contactors are not recommended.

Design factors to be considered for GAC filters or contactors include mass transfer zone (MTZ), empty bed contact time (EBCT), specific throughput, carbon usage rate, GAC effective particle size, uniformity coefficient and GAC bed porosity. EBCT typically ranges from 5 to 25 minutes. Typical loading rates for GAC contactors are in the range of 10 to 25 m/h (4 to 10 USgpm/ft²). Piloting testing is strongly recommended to determine the required EBCT, the specific throughput or the loading rate with respect to site-specific raw water quality. In addition, the performance of different GACs for the removal of specific contaminants of concern should be evaluated.

For the use of activated carbon specific applications, refer to Section 5.13 - Natural Organic Matter Control and Section 5.14 Taste and Odour Control.

5.8.2 Biological Filters

In a biological filter, the filter medium develops a microbial biofilm that assists in the removal of dissolved organic materials. Intentional biologically active filtration often includes the use of ozone as a pre-oxidant to break down natural organic materials into more easily biodegradable organic matter. Granular activated carbon filter media is often used to support denser biofilms because it has more surface area than other traditionally employed media, and may be referred to as biologically activated carbon. Biological dissolved organics removal and particle removal may occur in the same filter or in separate processes. Non-chlorinated backwash water is usually needed to preserve high levels of biological activity after backwashing. Consideration should be given to the likelihood of increased heterotrophic plate counts in the filter effluent.

The design of biologically active filters should ensure that aerobic conditions are maintained at all times. The final filter design should be based on pilot studies and should be designed as rapid rate gravity filters. Pilot studies should be of sufficient duration to stabilize the biological activity in the bed and establish the effectiveness and impact of the backwash procedures. Pressure filtration should not be used for biological filtration. Designers should be aware that carbon is electrically conductive and can strongly accelerate corrosion of exposed contacting materials.

5.9 Disinfection

5.9.1 General

The design of the drinking water disinfection processes must conform to the Disinfection Procedure. This Disinfection Procedure provides guidance for disinfection (primary disinfection), including any pre-disinfection treatment necessary to achieve the required level of removal and/or inactivation of pathogens potentially present in the source water, maintenance of a disinfectant residual in a distribution system (secondary disinfection) and control of disinfection by-products.

Drinking water disinfection and any pre-disinfection treatment requirements in Ontario are specific to the type of raw water supply. All water supplies should be individually assessed by measuring relevant water quality parameters. Design of the treatment processes should consider the characterization, variability and vulnerability to contamination of the raw water supply.

5.9.2 Disinfection By-Products

Disinfectants may be capable of producing disinfection by-products (DBPs) in concentrations that may present long-term health risks to drinking water consumers. The by-products produced are specific to the disinfectant and the raw water quality, and the concentrations are related to the contact time and the dosage and residual of the disinfectant.

The designer should consider alternative disinfection strategies for inclusion in the process train selected where optimized pilot or jar test filtered samples results indicate trihalomethane formation potential (THMFP) greater than the current THM standard in the Ontario Drinking-Water Quality Standards (O.Reg. 169/03) under the Safe Drinking Water Act, 2002. This test may be supplemented by simulated distribution system (SDS)-THM measurement, as specified in the most recent version of APHA/AWWA/WEF Standard Methods for the Evaluation of Water and Wastewater, where results of regular THMFP testing indicates that the standard may be exceeded. The potential formation of haloacetic acids (HAA) and other disinfection by-products should also be considered when evaluating disinfection alternatives.

The designer should refer to the Disinfection Procedure for design options and measures to control DBP formation.

5.9.3 Inactivation

Disinfection (primary disinfection) is a process or a series of processes intended to remove and/or inactivate human pathogens such as viruses, bacteria and protozoa that are potentially present in the water before the treated water is delivered to the first consumer.

Effective inactivation of pathogens in adequately filtered influent water, or groundwater of suitable quality, can be accomplished by either chemical or physical means such as the use of chlorine, monochloramine, chlorine dioxide, ozone or ultraviolet light. The designer should refer to the Disinfection Procedure for design options and specific requirements related to the use of different disinfection processes, the need for and use of different pre-inactivation treatment processes, as well as the pathogen removal or inactivation credits given to these processes and the criteria the processes must meet to qualify for these credits.

5.9.4 Chemical Inactivation Agents

For inactivation processes using free chlorine residual (chlorination), combined chlorine residual (chloramination), chlorine dioxide and ozone, the designer should refer to the Disinfection Procedure for design options including dosing, contact time and residual concentrations.

5.9.4.1 Free Chlorine

Refer to Section 6.4.2 Chlorine Gas and Section 6.4.3 Sodium Hypochlorite for information on chlorine feed systems.

5.9.4.2 Combined Chlorine

Sources of ammonia for chloramine production are either ammonia gas or water solutions of ammonia (aqua ammonia or ammonium hydroxide) or ammonium sulphate. Refer to Section 6.4.6 Ammonia for information regarding ammonia feed systems.

Addition of ammonia gas or ammonia solution will increase the pH of the water and addition of ammonium sulphate will depress the pH. The actual pH shift may be small in well buffered water but the effects on the disinfecting power and corrosiveness of the water should be considered.

For practical purposes, chloramines generally should not be used for primary disinfection due to the unacceptably high CTs required.

5.9.4.3 Ozone

Ozone must be generated on-site by means of an electrical discharge in oxygen or dry air. Although ozone is a highly effective disinfectant, it does not produce a lasting residual, and is therefore not suitable for secondary disinfection.

As a minimum, bench scale studies should be conducted to determine minimum and maximum ozone dosages for disinfection compliance and oxidation reactions. More involved pilot studies should be conducted when necessary to document benefits and DBP precursor removal effectiveness. Pilot studies should be conducted for all surface waters. Particularly sensitive measurements include gas flow rate, water flow rate and ozone concentration. Consideration should be given to multiple points of ozone addition.

Use of ozone may result in an increase in biologically available organics in the ozonated water and the need for biologically active filtration to stabilize the ozonated water should be evaluated. Ozone use may also lead to increased chlorinated by-product levels if the water is not stabilized and free chlorine is used for distribution protection. When considering ozone, the potential formation of bromate should also be evaluated.

A higher degree of operational skill is required, except in the case of very small packaged generators that provide only a few grams of ozone per day. The ability to develop operator skills should be evaluated in selection of the treatment process. The necessary operator training should be provided prior to plant start-up.

The production of ozone is an energy intensive process; substantial economies in electrical usage, reduction in equipment size and waste heat removal requirements can be obtained by using oxygen enriched air or 100% oxygen as feed and by operating at increased electrical frequency. Refer to Section 6.4.9 Ozone for information regarding ozone generation and feed systems.

5.9.4.4 Chlorine Dioxide

Chlorine dioxide must be generated on-site through the reaction of sodium chlorite with chlorine gas, hypochlorous acid or hydrochloric acid, or through the use of an electrochemical process. Chlorine dioxide is a powerful disinfectant that does not form chlorinated DBPs, however, the formation of chlorite and chlorate by-products should be evaluated. The chlorine dioxide residual from primary disinfection may also be used to maintain a residual through part or all of the water distribution system. Refer to Section 6.4.4 Chlorine Dioxide for information regarding chlorine dioxide generation and feed systems.

5.9.5 Ultraviolet Light Inactivation

5.9.5.1 General

Ultraviolet (UV) radiation may be used for primary disinfection but, since it does not produce a residual, it is not suitable for secondary disinfection. The designer should refer to the Disinfection Procedure for basic UV disinfection design requirements. UV facilities should be designed taking into account reliability and redundancy. At least one extra parallel UV system should be provided to ensure a continuous treated water supply when one unit is out of service. Parallel systems may not be required when there is adequate redundancy of treatment and supply through multiple sources (e.g., more than one well).

UV systems are proprietary. The designer has a choice between different UV technologies such as low pressure, medium pressure (MP) and low pressure high output (LPHO) lamps. Reactors using LPHO lamps are much more energy efficient in producing germicidal light and thus may have lower lifecycle costs. However, LPHO reactors require substantially more space than reactors housing the more intense and much higher temperature MP UV lamps.

The designer should support the choice of lamp/reactor technology by estimating life cycle costs, including the initial cost of purchase, installation costs and on-going lamp replacement costs, and by taking into consideration site specific conditions, such as plant layout, space availability and available head.

The designer should make provisions for water flow delays upon start-up of the UV reactors to allow the lamps to come up to the manufacturer recommended design operating temperature. In smaller installations, it may be economical to select reactors designed for air cooling and "always-on" illumination. Always-on reactors have an added benefit of lengthened lamp life, as frequent on-off cycles substantially reduce lamp life. Where there are extended no-flow periods and fixtures are located a short distance downstream of the UV unit, consideration should be given to UV unit shutdown between operating cycles to prevent heat build-up in the water due to the UV lamp. For medium pressure based reactors, cooling may be required during lamp start-up.

In selecting a UV system, the designer should consider the following major factors (additional information is also available in the USEPA Ultraviolet Disinfection Guidance Manual for the Final Long Term 2 Enhanced Surface Water Treatment Rule).

5.9.5.2 UV Dosages

The UV dosage required for water disinfection is set by the Disinfection Procedure as a pass-through UV energy of 40 mJ/cm² for groundwater that is not under the influence of surface water (where standard SI convention is used this is referred to as a fluence or energy flux of 4 J/m²).

Laboratory testing of cryptosporidium and giardia confirms 20 mJ/cm² is sufficient to provide adequate disinfection of these protist microbes⁵. The designer may provide UV disinfection at that dose where other treatment barriers ensure pathogens such as viruses and bacteria will be adequately disinfected in other treatment barriers.

Each flow path through a reactor receives a different UV dose and as a result it is not possible to estimate dose by using dwell time and UV intensity in an analogous way to chemical disinfection. To confirm the calculated design performance UV reactors need to be bioassay tested by an accredited third party testing organization, to a protocol accepted by the ministry and the manufacturer should provide assurances that reactor supplied conform to the design of the bioassayed sample reactor. The designer should ensure that bioassay calibration data are passed on to the municipality/owner and system operator for operational adjustments.

The designer should ensure that the reactors specified cannot be operated outside the bioassay established range of flows, UV transmittances and the corresponding photometer measured UV intensity. Care should also be taken to ensure that flow disturbances upstream of UV reactors are minimized and conform within the specified limits of the equipment manufacturer.

5.9.5.3 Water Quality

The absorption of light by water contaminants has a major influence on equipment selection and the cost of UV treatment. The designer should assess the available UV transmittance (UVT) data and select a well supported minimum transmission design value for the equipment supplier. UVT is defined as the percentage of 254 nm wavelength UV light that passes through 1 cm (0.4 in) of the water. Many common water contaminants can dramatically reduce the transmittance at UV wavelengths, resulting in higher costs of UV treatment. Since UVT is an important design consideration, the designer should endeavour to obtain data that supports a realistic estimate of the site specific lowest likely UVT.

Representative influent water quality should be evaluated and pre-treatment equipment, if necessary, should be designed to handle water quality changes. Particulates in water may shield microorganisms, affecting the UV inactivation performance. The scale formation/fouling potential in the UV reactor specific to the quality of the raw water supply should also be considered. Calcium, alkalinity, hardness, iron, pH, UV absorbing organics that adhere to quartz optical surfaces and water temperature are parameters that typically impact sleeve and sensor fouling. The manufacturer should be consulted regarding influent raw water quality requirements.

Chemicals added upstream of the UV system and the natural organic matter content in the influent water affect UVT. Therefore, it is recommended that UV systems be installed downstream of filtration for surface water sources. The designer should refer to the Disinfection Procedure for UV dose and pre-treatment requirements for groundwater and groundwater under the direct influence of surface water.

It is important to select an appropriate UV pass-through dose, based on the Disinfection Procedure. It is also important to ensure that the UV system selected for the design has obtained accredited testing agency bioassay validation to demonstrate that the UV pass-through dose can be achieved under the design conditions, and that the monitoring and control components are in agreement with the reactor bioassay dose over the range of water quality and operating conditions anticipated.

To determine the appropriate cleaning methods, a pilot test should be conducted if the fouling potential of the water is unknown (fouling potential may be difficult to predict based solely on water quality data). Alternatively, both chemical and mechanical cleaning systems may be included in the design.

5.9.5.4 Hydraulics

Headloss through the UV units should be considered where there may be periods of limited hydraulic head available. Typical headlosses range from 150 to 900 mm (0.5 to 3 ft). If the headloss through the UV system (reactor and associated piping, valves, flow control devices) is greater than the available head, modifications to the design or installation of booster pumps may be needed.

5.9.5.5 Operational Control Strategies

Several different control strategies are used to operate UV systems. The designer should consider the control strategies unique to various manufacturers and equipment, and select equipment consistent with the operating philosophy, disinfection strategy and energy efficiency objectives for the treatment plant.

The UV pass-through dose should at all times exceed the minimum dose required for disinfection. To ensure that the actual dose always exceeds this target, the design should be such that the reactor is operated within the bioassay validated range at all times and that the dose delivery and operation conditions are monitored at all times using appropriate strategies that, depending on the technology and bioassay validation protocol under which it was validated, might involve flow rate monitoring, UV lamp intensity monitoring, UV transmittance monitoring and UV dose delivery algorithms that integrate these parameters into a delivered dose.

5.9.5.6 Performance Validation

UV treatment devices should have proof of performance, demonstrating that they can deliver a sufficient UV dose with both an aging and fouling factor applied. The validation testing should include the following operating, design and equipment factors:

• Flow;
• UV intensity as measured by the UV sensor;
• UV lamp status;
• UVT of the water;
• Lamp ageing;
• Lamp sleeve fouling;
• Measurement uncertainty of on-line sensors;
• Failure of lamps or other critical system components; and
• Reactor inlet and outlet configurations.

It is important for UV units to have undergone performance validation by an accredited testing agency and for the manufacturer to certify that the selected reactor is identical to the bioassay tested unit. The validation testing should be to an internationally recognized protocol such as:

• German DVGW (German Technical and Scientific Association for Gas and Water) Technical Standard W 294;
• Austrian ON/ÖNORM M 5873;
• National Water Research Institute-American Water Works Association Research Foundation (NWRI-AwwaRF) Ultraviolet Disinfection Guidelines for Drinking Water and Water Reuse;
• USEPA Ultraviolet Disinfection Guidance Manual; and,
• NSF/ANSI Standard 55: Ultraviolet Microbiological Water Treatment Systems.

Class A of NSF/ANSI Standard 55 covers UV treatment systems for point-of-entry applications with flow rates of 1.9 L/s (30 USgpm) or less, therefore, their use in municipal drinking-water systems is limited. However, several of these small units can be used in a bank installation to meet the demands of very small systems if hydraulic and dosing issues are properly addressed, or when using a point of entry treatment design option in accordance with Drinking-Water Systems regulation (O.Reg. 170/03) under the Safe Drinking Water Act, 2002.

5.9.5.7 Alarms

Many UV reactor signals and alarms are specific to the UV facility and the level of automation used. The following alarms⁶ should be considered at each installation:

• Lamp Age - Lamp cumulative run-time and number of starts to guide scheduled replacement;
• Calibration Check of UV sensor - UV sensor requires calibration check based on operating time;
• Low UV Validated Dose - Indicated validated UV dose (based on UV reactor parameters, i.e., flow rate, UV intensity, and UVT) falls below required UV dose;
• Low UV Intensity - Intensity falls below validated conditions;
• Low UV Transmittance - UVT falls below validated conditions;
• High Flow Rate (if applicable; may rely on flow meters) - Flow rate falls outside of validated range;
• Mechanical Wiper Function Failure (if applicable) - Wipe function fails;
• Lamp/Ballast Failure - Lamp or ballast failure identified;
• Low Liquid Level - Liquid level within the UV reactor drops and potential for overheating increases; and
• High Temperature - Temperature within the UV reactor or ballast exceeds a setpoint.

The alarms provided for the unit(s) may vary depending on the specific validated conditions, type of UV reactor, manufacturer, dose-monitoring strategy and disinfection requirements.

5.9.5.8 Other Considerations

Water systems using UV should have ready access to a bench-top UVT meter which measures transmission at a wavelength of 254 nm in a path length of 1 cm of water sample to an accuracy of ± 2%. On-line water UVT sensing may be useful where lamp output modulation is practiced, but is not needed where on-line photometry is used to measure light penetrating through the water in the reactor.

The UV assemblies should be accessible for visual observation, cleaning and replacement of the lamps, lamp jackets and sensor window/lens.

The power supply for UV systems should be free from voltage variations exceeding the power supply design range and also free from frequent interruptions. Where disinfection performance must be maintained continuously, an uninterruptible power supply (UPS) is recommended.

The supply pump should be shut down, or alternatively, an automatic shutdown valve should be installed in the water supply line upstream of the UV treatment system that will be activated, whenever the water treatment system loses power or is tripped by the monitoring device when the dosage is below its alarm point. When power is not being supplied to the UV unit, the valve should be in a closed (fail safe) position. Pass-through light intensity monitoring without on-line flow confirmation may be acceptable where flow is restricted to not exceed a bioassay validated design maximum.

The UV housing should be 304L or 316L (low carbon) grade stainless steel.

5.10 Aeration & Air stripping

5.10.1 General

Aeration and air-stripping are gas-liquid contact processes. The designer should consider contaminant transfer efficiency, off-gas disposal issues, available hydraulic head, ease of operation, and capital and operating/maintenance costs in evaluating a gas-liquid contact process.

Air stripping is most commonly used in Ontario for the control of methane and/or hydrogen sulphide in groundwater. Due to the relatively low solubility of methane in water, even a splash plate type aerator may provide sufficiently effective treatment. Appropriate ventilation should be provided to ensure that methane concentrations do not reach the Lower Explosive Limit (LEL).

Air stripping for hydrogen sulphide removal has specific limitations, as carbon dioxide is also stripped, leading to pH increases and the potential for scaling. Pilot testing should be conducted to ensure sulphide levels can be sufficiently reduced without causing scaling issues. If air stripping can not be used, alternative chemical processes to reduce sulphide concentrations to inoffensive levels should be considered.

The feasibility of aerators or air strippers should be evaluated through pilot studies. The pilot test should evaluate a variety of loading rates and air-to-water ratios at the peak contaminant concentration. Consideration needs to be given to removal efficiencies, oxidation rates or scaling due to incidental carbon dioxide stripping, when multiple contaminants are present.

The materials of construction, including all material in contact with the water, should be corrosion resistant (Section 3.26 Chemicals and other Water Contacting Materials).

Aeration and air-stripping processes include multiple tray, spray aerators or towers, pressure aerators and packed towers, spraying, diffused air, cascades and mechanical aeration. Since these processes are not common in Ontario, specific design guidelines are not included in this document. The designer should consult other guidelines such as Recommended Standards for Water Works⁷ (Ten-State Standards) and with the equipment manufacturer for all gas-liquid contact processes.

5.11 Softening

5.11.1 General

The softening process selected should be based upon the mineral qualities of the raw water and the desired finished water quality in conjunction with requirements for disposal of sludge or brine waste, capital and operating/maintenance costs. Methods of hardness reduction other than lime softening should be investigated when the sodium and dissolved solids concentrations are of concern.

5.11.2 Lime or Lime-Soda Process

Design guidelines for rapid mix, flocculation and sedimentation are described in Section 5.4 Coagulation and Flocculation and Section 5.5 Clarification. Additional consideration should be given to the following process elements:

• Hydraulics - When split treatment is used, the bypass line should be sized to carry total plant flow and an accurate means of measuring and splitting the flow should be provided;
• Aeration - Determinations should be made of the carbon dioxide content of the raw water. When concentrations exceed 10 mg/L, the economics of removal by aeration as opposed to removal with lime should be considered if it has been determined that dissolved oxygen in the finished water will not cause corrosion problems in the distribution system;
• Chemical feed point - Lime and recycled sludge should be fed directly into the rapid mix basin;
• Rapid mix - Rapid mix basins should provide not more than 30 seconds detention time with adequate velocity gradients to keep the lime particles dispersed;
• Stabilization Equipment - for stabilization of water softened by a lime or lime-soda process should be provided;
• Sludge collection - Mechanical sludge removal equipment should be provided in the sedimentation basin. Sludge recycling to the rapid mix should be provided;
• Sludge disposal - Provisions should be included for proper disposal of softening sludges (Section 11.2.6 Precipitative Softening); and
• Plant start-up - The plant processes should be manually started following shutdown.

5.11.3 Ion Exchange Process

The iron and/or manganese concentration should not exceed 1.0 mg/L and the turbidity should be less than 5 NTU in the water applied to the ion exchange resin. The ion exchange units may be of pressure or gravity type, of either an upflow or downflow design. Automatic regeneration based on volume of water softened should be used. A manual override should be provided on all automatic controls.

Other design considerations include:

• Exchange capacity - The design capacity for hardness removal should not exceed 46 kg/m³ (22,000 gr/ft³) when resin is regenerated with 0.14 kg (0.3 lbs) of salt per kg of hardness;
• Depth of resin - The depth of the exchange resin should not be less than 900 mm (3 ft);
• Flows - The flow rate should not exceed 17 m/h (7 USgpm/ft²) and the backwash rate should be 14 to 20 m/h (6 to 8 USgpm/ft²). Rate-of-flow controllers or the equivalent should be installed;
• Freeboard - The freeboard will depend upon the size and relative density of the resin and the direction of water flow. Generally, the washwater collector should be 600 mm (24 in) above the top of the resin on downflow units;
• Underdrains and supporting gravel - The bottoms, strainer systems and support for the exchange resin should conform to criteria provided for rapid rate gravity filters (Section 5.6.2 Rapid Rate Gravity Filters);
• Brine distribution - The design should ensure even distribution of the brine over the entire surface of both upflow and downflow units;
• Cross-connection control - Backwash, rinse and air relief discharge pipes should be installed in such a manner as to prevent any possibility of backsiphonage;
• Bypass piping and equipment - A bypass should be provided around softening units to produce a blended water of desirable hardness. Totalizing meters should be installed on the bypass line and on each softener unit. The bypass line should have a shutoff valve and should have an automatic proportioning or regulating device. In some installations, it may be necessary to treat the bypassed water to obtain acceptable levels of iron and/or manganese in the finished water;
• Resins - Silica gel resins should not be used for waters having a pH above 8.4 or containing less than 6 mg/L silica and should not be used when iron is present. When the applied water contains a chlorine residual, the resin should be of a type that is not damaged by residual chlorine. Phenolic resin should not be used;
• Sampling taps - Smooth nosed sampling taps should be provided for the collection of representative samples. The taps should be located at the softener influent, effluent and blended water. The sampling taps for the blended water should be at least 6 m (20 ft), or at a distance where sufficient mixing has occurred, downstream of the point of blending. Sampling taps should also be provided on the brine tank discharge piping;
• Brine and salt storage tanks - Salt dissolving or brine tanks and wet salt storage tanks should be covered and be corrosion-resistant. The make-up water inlet should be protected from backsiphonage. Water for filling the tank should be distributed over the entire surface by pipes above the maximum brine level in the tank. The tanks should be provided with an automatic declining level control system on the make-up water line. Wet salt storage basins should be equipped with hatchways for access and for direct dumping of salt from truck or railcar. Openings should be provided with raised curbs and watertight covers having overlapping edges similar to those required for finished water reservoirs. Overflows, where provided, should be protected with corrosion resistant screens and should terminate with either a turned-down bend having a proper free fall discharge or a self-closing flap valve. Two wet salt storage tanks or compartments designed to operate independently should be provided. The salt should be supported on graduated layers of gravel placed over a brine collection system. Alternative designs which are conducive to frequent cleaning of the wet salt storage tank may be considered;
• Salt and brine storage capacity - Total salt storage should have sufficient capacity to store in excess of 1½ full loads of salt, and provide for at least 30 days operation;
• Brine pump or eductor - An eductor may be used to transfer brine from the brine tank to the softeners. If a pump is used, a brine measuring tank or means of metering should be provided to obtain proper dilution;
• Waste disposal - Suitable disposal should be provided for brine waste (Section 11.2.5 Ion Exchange Processes). The designer could consider using part of the spent brine for a subsequent regeneration;
• Construction materials - Pipes and contact materials should be resistant to the aggressiveness of salt; and
• Housing - Bagged salt and dry bulk salt storage should be enclosed and separated from other operating areas in order to prevent damage to equipment.

5.12 Iron & Manganese control

5.12.1 General

Iron and manganese are frequently encountered nuisance parameters that seriously affect aesthetic water quality. They can cause visible water colour and turbidity and cause brown and black staining of plumbing fixtures and washed clothing. These effects can occur at specific locations in a distribution system even when the concentration of either metal in the treated water entering the distribution system is below the Ontario aesthetic objective stated in the Technical Support Document. This occurs as a result of precipitation and redissolution processes resulting in pockets of local high concentrations.

Elevated iron and manganese concentrations occur most frequently with groundwater sources. Surface water sources may also be contaminated with the metals at anoxic depths in lakes or seasonally under long-duration ice cover. Five control technologies⁸ are in common use in Ontario, including:

• Removal of iron by air or chlorine oxidation followed by sedimentation;
• Masking the impact of iron by "sequestering";
• Removal by ion exchange water softeners;
• Removal by "greensand" processes; and
• Pre-oxidation and regular chemically assisted depth or membrane filtration.

Very high levels of iron and manganese of 5 mg/L or more can be treated by lime softening processes. This is costly and rarely necessary in Ontario where alternative, nearby sources may be found with lower iron or manganese concentrations. Refer to Section 5.11 Softening for lime softening guidelines.

5.12.2 Air/Chlorine Oxidation of Iron

In groundwater treatment systems where iron levels are near the aesthetic objective, a simple sedimentation process may result in acceptable finished water quality. On exposure to active chlorine or oxygen, divalent or ferrous iron is rapidly oxidized to the effectively insoluble trivalent or ferric state. In favourable circumstances the ferric iron may precipitate as a readily separating brown solid without additional complex treatment. Oxygen and chlorine oxidize manganese at too slow a rate for effective removal unless the manganese is present at very low levels relative to the iron concentration. However in some cases the newly formed iron precipitate surface is found to at least be partially effective in adsorbing the manganese.

5.12.3 Sequestering with Silicates or Polyphosphates

Sequestering is an inexpensive and commonly adopted palliative measure for iron control that slows, but does not stop, the perceptible formation of the typical yellow/brown colour. Sequestering temporarily traps and then slowly releases oxidized iron into the water from a complexed/colloidal form. At most, the effect lasts for only a few days. The eventual failure of sequestration is due to calcium ions progressively displacing the iron so that the regular perceptible yellow-brown colour becomes visible. (This happens rapidly in hot water forming sediment in domestic hot water heaters.)

Effective sequestering depends on the sequestering agent, either freshly hydrated silica or polyphosphate ions, intercepting ferric iron ions as they are formed by chlorine oxidation. The forming precipitate of oxidized iron is then physically trapped and kept in a colourless colloidal suspension by the sequestrant. This process of sequestering the iron must be completed during the few seconds before hardness cations such as calcium take up and block any further sequestrant activity. The designer of a sequestering system should first confirm suitability of the source water for sequestering by:

• On-site colorimetric testing of fresh water samples to confirm that the iron content is mostly in the chemically reduced divalent form; and
• Confirming through on-site or regular total metals analytical scans that manganese levels are low relative to the iron content of the water.

Designers are reminded that GUDI sources may show substantial seasonal variations in both iron and manganese speciation and concentrations.

The approximate sequestrant dosage and the likely delay time to perceptible colour development should be confirmed by on-site testing⁹. If a significant delay in colour formation (relative to the water retention time in the distribution system) is confirmed by on-site testing, the following equipment should be provided for full scale silicate sequestering:

• A locally placed day tank with lid for hypochlorite;
• A locally placed day tank with lid for silicate, sized for up to two weeks of water treatment that allows for 1:2 dilution of silicate (for viscosity reduction), preferably with softened water. Greater dilution ratios or longer storage of diluted silicate should be avoided, as these conditions reduce sequestering effectiveness, particularly in warm seasons;
• Feed pumps, preferably of the peristaltic type, or other pumping arrangements, adapted to provide continuous, effectively pulse free addition of the hypochlorite and silicate to the flowing water stream;
• Injectors of the "duck bill" or other scale blockage resistant variety that allow for injection of the silicate and hypochlorite to the centre of a rapidly flowing stream to aid in the necessary rapid dispersion of the two chemicals. Best results are commonly achieved with hypochlorite added a metre or more upstream of the silicate injector location and where mixing is assisted by use of nearby downstream elbows or other means. Several tappings should be made to allow for easy injector relocation and spacing changes; and
• A nearby downstream wide bore sample tap that allows for easy collection of 20 Litres (5.28 USgal) samples for observation and dosage optimization.

Polyphosphate blends are considerably more costly but may be as effective as silicate for sequestering. However they have been documented in some cases to cause lead leaching in domestic plumbing. Polyphosphate is usually most effective when injected close to the hypochlorite injector location. Designers should note that divalent manganese, in the absence of an excess of iron, is not sufficiently rapidly oxidized by hypochlorite for effective sequestering. Often where this is attempted, the manganese precipitates during passage through the distribution system causing the usual black staining. Other oxidants may work more rapidly but because of the additional complications involved, their use is not generally merited. Where manganese concentration approaches the aesthetic objective levels, removal often may be a more appropriate treatment option.

Groundwater sources tend to show increasing iron levels as they age. Designers should consider providing space to accommodate future installation of removal equipment where there is evidence from other local wells that this may eventually become a necessary treatment upgrade.

Refer to Chapter 6 - Chemical Application for guidelines regarding chemical storage, handling and feed systems,

5.12.4 Ion Exchange in Regular Softeners

In small groundwater systems it is frequently attractive to consumers to have the water hardness reduced by having all or a substantial fraction of the water passed through an ion exchange softener. An added feature of ion exchange softening is that dissolved iron and manganese are also removed in the same way as calcium and magnesium on the exchanger resin beads. Regeneration of the resin with salt brine displaces calcium and the divalent iron and manganese into the spent brine which then should be safely disposed (Section 11.2.5 Ion Exchange Processes).

This process is quite effective but operates only on divalent and non-organically bound iron and manganese and only reduces concentration in proportion to the ratio of softened to non-softened/by-passed water. Designers should be aware that as wells age it is frequently observed that increasing contamination occurs with iron and manganese in oxidized form and with sloughed biofilm and other debris from bacterial activity. These materials may plug protecting cartridge filters and the resin causing frequent, premature and costly replacement. In addition, exposure of resin containing adsorbed divalent iron and manganese to oxygen or other oxidants causes effectively permanent loss of resin capacity.

For these reasons, the application of this technology should be restricted to raw water where on-site colorimetric testing shows relatively low levels of iron and manganese with a high proportion present in the divalent form. Refer to Section 5.11.3 Ion Exchange Process for ion exchange guidelines.

5.12.5 Greensand Type Processes

Historically this process depended on using greensand, a manganese ore, as the filter media downstream of the addition of a chemical oxidizing agent. Media made from manganese ore particles has long been replaced commercially by more robust proprietary media, but the name for the overall process retained. The proprietary media used are surface treated to physically adsorb and retain an oxidized iron and/or manganese surface layer. Filter media should conform to NSF/ANSI Standard 61: Drinking Water System Components - Health Effects and the applicable AWWA Standard B101: Precoat Filter Media or AWWA Standard B102: Manganese Greensand for Filters.

The process is selective for the iron and manganese and does not involve removal of other contaminants, such as hardness components. As a result the equipment is usually more compact than for ion exchange softening or for conventional chemically assisted filtration.

Deposited material on the media is periodically partially removed by backwashing. Backwashing should be limited in intensity to avoid completely stripping the essential base catalytic surface from the media.

Both designers and operators of greensand systems should be aware that failure to maintain the activity of the catalytic layer can lead to sudden failure of the process and this will require replacement of the media. In some cases, the maintenance of media activity requires a continuous feed of potassium permanganate solution (often with hypochlorite), while in others a regular soaking of the media in the permanganate solution and operation of the filter with a hypochlorite feed alone is sufficient. If iron alone is to be removed, the media may not require catalytic activity at all, just a capability to adsorb and retain the precipitated iron; therefore permanganate use may not be necessary.

The presence of iron and manganese is often attributed to biological activity in the aquifer near wells bores. A secondary and frequent consequence is the presence of sloughed-off biofilm and particulate bacterial waste in the raw water. The presence of these particulates influences the optimum media bed shape and media selection, with conical beds and dual media more often adopted where there is heavier contamination.

Selecting filter loading rates, media type and depths, run duration to backwash and oxidant regenerant/oxidant and dose control are complex issues dependent on a detailed understanding of how the technology responds to the specific raw water characteristics. On occasion, the manganese may be found to be at least partly resistant to oxidation because of association with organics. In such cases, the use of chlorine dioxide or ozone may be required as they are faster acting oxidants. Specialist media supply companies are commercially active in this field in Ontario and can provide information regarding site/water adapted equipment, media and full scale operating instructions together with performance guarantees. Short term bench and on-site pilot testing lasting no more than a few days is recommended.

Greensand type processes will remove pathogens in most circumstances; however, because coagulant activity only occurs on the catalytic media particle surface, the efficiency may be very low. As a result no disinfection log credits are allowed for pathogen removal under the Disinfection Procedure.

5.12.6 Preoxidation & Chemically Assisted/Membrane Filtration

Where pathogen removal is required, this category of treatment may be preferred as it combines the required pathogen removal and associated log credits for filtration with the control of iron and manganese. The designer may consider the following options following use of a selected oxidant:

• Conventional chemically assisted filtration with granular filtration media of the type required for turbidity control, coupled with appropriate coagulation/flocculation processes for pathogen removal log credits; and
• Membrane filtration with, or in some cases without, coagulant chemical or aid. Time is required for the agglomeration of iron and manganese particles for effective filtration and/or for the action of coagulant. Coagulant use is not required for log removal credits for some pathogens.

In some small systems with iron contamination, a combination of hypochlorite oxidation on granular media, with or without downstream cartridge filtration, may be an effective low cost alternative.

The use of permanganate as an oxidant or a combination of permanganate with hypochlorite is commonly effective with GUDI sources. The oxidant should be added upstream of the coagulant and sufficient time allowed for effective completion of the metal oxidation before coagulation. Oxidation may be slowed in winter conditions as GUDI source temperatures can be close to freezing and also the presence of natural organics may also slow the reactions by forming chemical complexes with manganese. The designer should take account of these factors particularly with pressure filter systems where the scale of the system can have a very significant influence on the cost of pressure vessels required to provide adequate residence time. On occasion, oxidation reactions can be so slow that it may be cost effective to consider the use of more rapidly acting oxidants such as ozone, or providing two stage treatment comprising regular greensand contact upstream of a form of regular filtration.

The designer should make provision to avoid a potential consequence of use of permanganate (or ozone with manganese containing raw water as this combination can produce permanganate). Permanganate is strongly coloured and if present in sufficient amount can have the undesired consequence of a faint permanganate pink colour becoming visible in finished water. While this is not an immediate health issue, it will cause consumer alarm. The use of available on-line colorimetry for permanganate or ozone dose adjustment should be considered.

Some surface water supplies can also require seasonal iron and manganese treatment. Typically, surface sources that require treatment for iron and manganese control have long duration ice cover and/or deep lake intakes and usually require variable and carefully adjusted treatment modification seasonally before spring break up. Provision of means for addition and mixing of permanganate solution some minutes upstream of coagulant addition is desirable for effective removal of the metals in the sedimentation and filtration steps.

5.13 Natural Organic Matter Control

5.13.1 General

Natural organic matter (NOM) has traditionally been partly removed from drinking water for aesthetic reasons as it often imparts colour to the water. The low molecular weight fractions of NOM are mostly responsible for chlorination by-product formation and therefore a reduction of the NOM level is desirable. NOM may also affect water treatment processes including coagulation (dosage and optimum pH), membrane filtration (fouling), disinfection (chemical demand or UVT), activated carbon usage rates and distribution system water quality (biological regrowth potential). Enhanced coagulation can reduce the fraction of the chlorine by-product forming NOM in filtered water, but it is often found that other techniques are easier to use in controlling by-product formation.

5.13.2 Activated Carbon

GAC can be used for reducing the NOM concentration, and may also be operated in a biologically active mode for NOM reduction. However, the use of GAC to adsorb NOM is not generally very effective or economically attractive as the GAC rapidly loses adsorption capacity. This occurs even when GAC is operated as an active biofilter support. However, GAC use may be necessary as the only moderately economic recourse in controlling some taste and odour events, such as those caused by geosmin and methylisoborneol (MIB).

The empty bed contact time (EBCT) required for substantial NOM removal typically varies from 10 to 20 minutes. Several months of pilot testing is needed to determine the EBCT and other variables described in Section 5.8 Soluble Contaminant Removal Processes.

5.13.3 Nanofiltration

A large fraction of the NOM can be removed with membranes having a molecular weight cutoff of 1000 or less. This process can be costly as relatively high pressures are used and pre-treatment is needed to protect the membranes from particulate accumulation, as the maximum backwash flow rates available with such "tight" membranes are too low to dislodge accumulated solids. Nanofiltration has not yet found wide application in Ontario, although there are some installations across Canada. Experience with nanofiltration to date has not been sufficient to provide specific design guidelines; the designer should therefore consult the manufacturer if considering nanofiltration.

5.14 Taste & Odour control

5.14.1 General

Offensive tastes and odours should be controlled at all surface water treatment plants. Plants treating water that is known to have taste and odour problems should be designed with several control processes so that the operator will have flexibility in operation. Pilot-scale and/or in-plant studies are recommended to determine the best treatment process(es).

5.14.2 Microscreening

Microscreens or microstrainers are mechanical screens with very small openings capable of removing suspended matter from the water by straining. Microscreens generally follow immediately after coarse screens. Microscreens are used during periods when the raw water contains nuisance organisms such as algae when heavy loadings may negatively impact downstream processes (e.g., granular and membrane filtration).

Designers should consider the:

• Expected loading and duration of the algae blooms;
• Corrosiveness of the water;
• Effect of chlorination when required as pre-treatment;
• Duplication of units for continuous operation during equipment maintenance;
• Automated backflushing; and
• Alternative technologies such as dissolved air floatation (Section 5.5.5 Dissolved Air Flotation).

The design should provide:

• By-pass arrangements;
• Protection against backsiphonage when treated water is used for washing; and
• Proper disposal of wash water.

5.14.3 Oxidation

Many common taste and odour causing substance can be chemically oxidized to less odorous substances or mineralized to carbon dioxide and water. Chemical oxidants include chlorine, monochloramine, chlorine dioxide, potassium permanganate, hydrogen peroxide and ozone. Advanced oxidation processes (Section 5.14.4 Advanced Oxidation) and aeration (Section 5.10 Aeration and Air Stripping) may also be effective.

Taste and odour control chemicals (e.g., chlorine and potassium permanganate) should be added sufficiently upstream of other treatment processes to ensure adequate contact time for an effective and economical use of the chemicals. Considerations should also include the potential for by-product formation.

Refer to Chapter 6 - Chemical Application for guidelines relating to chemical handling, storage and feeding,

5.14.4 Advanced Oxidation

Advanced oxidation processes (AOPs) are processes that provide powerful oxidizing conditions to mineralize organic water contaminants. AOPs involve the use of any one of several possible combinations of UV, hydrogen peroxide, ozone and titanium dioxide.

AOPs depend on extremely unstable radical chemical species that react very rapidly with any organic material present. Any NOM which may also be present in the water is mineralized at a similar rate to the target contaminants. As a result, AOPs should only be used on very low to trace amounts of specific contaminants such as N-nitrosodimethylamine (NDMA) or 1-4 dioxane, and only in water with low NOM content. Bench and/or pilot scale evaluation using the specific source water and covering seasonal variations is needed to establish effectiveness and costs.

AOPs that use hydrogen peroxide may produce water with a peroxide residual that behaves like chlorine in colourimetric tests, reacts with and destroys free chlorine, and can upset downstream biological processes. Thiosulphates, sulphites or GAC can be used to destroy peroxide residuals.

5.14.5 Activated Carbon

A wide range of water contaminants that cause offensive tastes and odours can be at least partly removed by contact with activated carbon. Activated carbon may be in powdered (PAC) or granular (GAC) form. PAC is used as a continuously fed additive that must be removed following the required contact time, but before primary disinfection, by processes such as chemically-assisted filtration. GAC is used in fixed contactor beds. The selection decision between using PAC or GAC should be based on the nature and concentration of the contaminant to be removed and a wide range of site and process specific considerations.

5.14.5.1 Powdered Activated Carbon

PAC should be added as early as possible in the treatment process to provide maximum contact time. The designer should consider the removal of the PAC and its impact on the filtration process, as well as disposal of the sludge produced by the PAC addition. Activated carbon should not be added near the point of chlorine or other oxidant application, as the adsorption capacity of the carbon decreases due to chemical reactions that convert chlorine to chloride and other oxidants to inactive materials.

The rate of feed of carbon in a water treatment plant depends upon the taste and/or odour reduction needed and the contact time available, but provision should be made for adding up to at least 40 mg/L. Pilot scale testing is recommended to determine contact time and the range of dosages required.

PAC can be added as a pre-mixed slurry or by means of dry-feed equipment as long as the carbon is thoroughly wetted before its introduction to the water to be treated. Refer to Section 6.4.13 Powdered Activated Carbon for more information on storage and feeding of PAC.

5.14.5.2 Granular Activated Carbon

Granular activated carbon (GAC) can be used in place of anthracite in granular filters (Section 5.6 Granular Media Depth Filtration) or in separate contactors (Section 5.8 Soluble Contaminant Removal Processes). When GAC is used as a layer in filters, the GAC cannot be removed from service or by-passed during periods when tastes and odours are not a problem. This potentially shortens the life of the GAC for taste and odour control as other compounds are adsorbed onto the active sites on the carbon. GAC contactors however can be by-passed in winter months to extend the effective bed life.

The empty bed contact time (EBCT) required for taste and odour control depends on the nature of the taste and odour compounds and typically varies from 10 to 30 minutes. Pilot testing is recommended to determine EBCT and expected bed operation life. Where the contaminant to be controlled is present only in short term seasonal excursions, pilot work may be useful to indicate effective bed life and the potential need for off-line contactors.

GAC in filters or separate contactors may be operated in a biologically active mode for taste and odour control (Section 5.8 Soluble Contaminant Removal Processes).

5.15 Nitrite/ Nitrate removal

The following treatment processes are generally considered acceptable for nitrate/nitrite concentration reduction: anion exchange, reverse osmosis, nanofiltration and electrodialysis. Although these treatment processes, when properly designed and operated, will reduce the nitrate/nitrite concentration in the treated water to acceptable levels, primary consideration should be given to obtaining water from an alternate water source or reducing the nitrate/nitrite levels through blending and/or better control of nitrogen fertilizer application, septic systems and waste disposal.

Refer to Section 5.11.3 Ion Exchange Process for ion exchange guidelines. High levels of sulphate, chloride or dissolved solids may interfere with an ion exchange process. In these cases, reverse osmosis, nanofiltration or electrodialysis should be investigated. The equipment manufacturer should be consulted for guidelines for these processes.

5.16 Arsenic removal

The form in which arsenic is present in water is critical in the selection of the treatment technology for arsenic removal. The use of pre-oxidation processes may be necessary to oxidize As(III) to As(V) for optimum treatment performance. Pre-treatment may also be needed to adjust pH and to remove competing ions such as fluoride, sulphate and silicate as well as to reduce total dissolved solids. Issues to be evaluated when considering pre-treatment processes include by-product formation and membrane fouling potential if membranes are used.

At this time, the technologies available for removing arsenic from municipal drinking-water systems include coagulation/filtration, iron based adsorbents, lime softening, activated alumina, ion exchange, reverse osmosis, manganese greensand filtration, adsorption/filtration and electrodialysis. Point-of-entry treatment devices may be practical for small systems (treatment requirements in O.Reg. 170/03 are related to microbiological contaminants; other contaminants, such as arsenic, are outside the scope of this regulation). Blending of water sources or treating a portion of the water (sidestream treatment) to reduce the concentration of arsenic in water delivered to the consumer are other potential arsenic control techniques. Additional guidance in process selection is available from the USEPA and the AWWA.

Pilot testing of any treatment process is recommended. Disposal of residuals is an important consideration in the design of arsenic removal processes.

5.17 Fluoride removal

The designer should refer to the Technical Support Document where naturally occurring fluoride is above the standard established in O.Reg. 169/03. Methods to reduce fluoride include ion exchange, reverse osmosis, coagulation/flocculation processes with high alum dosage, and proprietary technologies.

If fluoride is to be added to the water, the total concentration of naturally occurring fluoride plus added fluoride should be in the range specified in the Technical Support Document. Sodium fluoride, sodium silicofluoride and hydrofluosilicic acid are common chemicals used for fluoridation. Refer to Section 6.4.14 Fluoride for guidelines regarding chemical storage, handling and feeding.

5.18 Internal corrosion control

5.18.1 General

Metals that are in contact with water containing oxygen or chlorine or metals exposed to acidic conditions alone will undergo corrosion. Electrical contact between different metals and stray electric currents also can cause localized corrosion. Distribution system wide corrosion and the corresponding water quality deterioration can usually be reduced through water treatment plant process adjustment or by addition of selected chemicals.

Parameters such as the Langelier Saturation Index¹⁰ that are based exclusively on water solution chemistry, and which do not take into account processes occurring at the metal surface, are presently considered unreliable indicators of changes that may slow corrosion. Also, industrial corrosion coupon testing is not reliable unless long duration procedures are conducted with standardized hydraulic conditions. Simulated distribution testing (pipe loop testing) is an expensive, time consuming procedure and may provide unreliable results.

A suggested approach is to adopt a methodical full-scale testing procedure such as:

• Establishing a baseline measure of the existing rate of corrosion by documenting water quality complaint location and frequency, or preferably by adopting widespread sampling following a selected protocol (e.g., the USEPA Lead and Copper Rule); or,
• Making limited water quality changes such as pH shift of +0.5 units or soda ash addition of up to 10 mg/L, and confirming the effectiveness of the adopted corrosion control method after allowing 6 to 9 months for the corrosion process to adjust.

The most reliable indication of the effectiveness of the above procedure can be obtained by setting aside a section of the distribution system for water quality adjustment and comparing sampling results from this area with that from the rest of the distribution system. This eliminates the effect of seasonal water quality changes that may otherwise obscure indications of effectiveness level.

Increases in pH, alkalinity and carbonate buffer content are the most consistent methods for reducing the rate of corrosion. Increasing the carbonate buffer level is particularly recommended for systems treating very soft water.

It is recommended that corrosion rate estimations be used to validate the efficacy of these water quality changes in reducing corrosion. Corrosion rates are commonly observed to take from 6 months to a year to stabilize after changes in water quality. A reliable testing technique is to make changes to water composition in one area of a distribution system while leaving the water composition unchanged in the remainder of the system, and comparing testing results from the two areas after a period of 6 months to a year.

Very soft and acidic water has been known to be capable of causing health- significant releases of lead from lead service lines, from domestic plumbing that may have used lead containing solder and from lead containing brass fittings. However lead levels in flushed samples from domestic taps rarely exceed the O.Reg. 169/03 lead standard where the water has alkaline pH, moderate alkalinity and carries an adequate secondary disinfectant residual. An exception to this may occur with major changes in water composition such as with uncontrolled and wide pH variations, new mixing with water from another source, new addition of a corrosion inhibiting chemical or changeovers from free chlorine to chloramine as secondary disinfectant. It is prudent to track lead level changes in well flushed samples taken from consumer taps during planned changes to water composition, and if necessary to slow or reverse the changes, and/or plan removing the lead service lines.

5.18.2 Raising pH, Alkalinity & Carbonate Buffer Level

Sodium carbonate addition provides simultaneous and easily controlled increases in pH, alkalinity and carbonate buffer capacity. Addition may be by feeding solid soda ash or by solution addition.

The use of 50% sodium hydroxide solution increases pH and alkalinity but does not increase the buffer capacity and may be of limited effectiveness with very soft water. In addition, pH control can be difficult with sodium hydroxide. 50% sodium hydroxide solution can freeze at moderately depressed temperatures and handling hazards exist.

Combinations of sodium hydroxide with carbon dioxide offer maximum flexibility in adjusting finished water pH, alkalinity and carbonate buffer capacity but relatively complex controls may be needed.

The use of lime to raise pH and alkalinity in the water has been largely discontinued because of the tendency to plug feed lines and to create excursions in finished water turbidity.

5.18.3 Commercial Corrosion Inhibitors

Where adjustments to water quality parameters such as chlorine residual, pH, alkalinity and carbonate buffer strength prove insufficient to control corrosion rates, the use of special additives as listed below should be considered.

Sodium silicate in dosage rates of up to 10 mg/L has been shown to suppress red water in some systems.

The addition of ortho-phosphoric acid has been used at a 3-5 mg/L in some systems with red water problems. Sodium ortho-phosphate may be used in the same way as the acid without the pH depressing effect. Several different grades are available depending on the sodium/phosphate ratio. Some success has been also reported in red water control with polyphosphates and also with ortho-phosphate/polyphosphate blends. The designer should be aware that in some systems, elevated lead levels have occurred with these inhibitors. The addition of phosphorus containing substances will add to the phosphorus load entering sewage treatment facilities and may encourage biofilm growth in distribution systems.

Zinc salts are also known to suppress microbial activity and have provided some reductions in red water problems. Use of zinc compounds may affect downstream sewage treatment.

5.18.4 Carbon Dioxide Reduction by Aeration

The carbon dioxide content of an aggressive and corrosive groundwater may be reduced by aeration. Refer to Section 5.10 Aeration and Air Stripping for a summary of aeration processes. The designer may consult a guideline such as Recommended Standards for Water Works¹¹ (Ten-State Standards) and with the equipment manufacturer for detailed aeration guidelines.

5.18.5 Limestone Chip Contactors

Percolating water through a bed of limestone chips at the end of the water treatment process has been used to neutralize acid and suppress iron corrosion in very small water systems. Bench testing is recommended with selected commercial chips and treated water on site to confirm effectiveness.

Limestone chips can not be certified to regular NSF/ANSI standards but may be considered on a site specific basis at the discretion of the Director

Chapter 6: Chemical application

This chapter provides general information regarding chemical feed systems, equipment, application points, as well as recommendations for storage and handling. Specific information concerning chemical processes and the use of the chemicals described herein is provided in the appropriate sections of Chapter 5 Treatment.

Section 6.4 Specific Chemicals provides more detailed information about some of the chemicals which are commonly used in water treatment plants in Ontario. For all chemicals, the designer should consult the chemical manufacturer/supplier regarding chemical functionality and safety and for guidance in designing the chemical feed system. AWWA Standards also provide information on specific chemical safety, handling and storage.

In addition, the use, storage and handling of any hazardous materials should be in accordance with federal and provincial legislation for Workplace Hazardous Materials Information System (R.R.O. 1990, Regulation 860) under the Occupational Health and Safety Act (OHSA), the Building Code (O.Reg. 350/06) under the Building Code Act, 1992 and the Fire Code (O.Reg. 388/97) under the Fire Protection and Prevention Act, 1997.

6.1 General

6.1.1 Plans & Specifications

Plans and specifications should include descriptions of feed equipment including maximum and minimum feed ranges, dosage capabilities at maximum production rates, location of feeders, piping layout and points of application, storage and handling facilities, specifications for chemicals to be used, operating and control processes, and descriptions of monitoring equipment and procedures. Process flow diagrams (PFD) showing all process components including reactors, pumps, chemical feeders, valves, analyzers and the location of all points of chemical addition, effluent sampling and monitoring should be included. A narrative statement of the intended chemical process control philosophy, treatment ranges, and process and instrumentation diagrams (P&ID) showing the operation control arrangements for all processes should also be included.

6.1.2 Chemicals & Water Contacting Materials

Requirements for chemicals and water contacting materials are discussed in Section 3.26 Chemicals and Other Water Contacting Materials.

6.1.3 Chemical Application

Chemicals should be applied to the water at such points and by such means as to ensure safety for consumers and operators, ensure efficacy of treatment and the ability to respond to changes in water quality, ensure sufficiently rapid and effective mixing of the chemicals with the water and provide maximum flexibility of operation through provision of multiple points of application, when appropriate. Sampling points should be designed for effective, timely and representative monitoring of chemical application.

6.1.4 General Equipment Design

The equipment design should ensure that feeders will be able to accurately supply, at all times, the necessary amounts of chemicals throughout the design range of dosage and water flows. Reliability of chemical feed systems should be considered, and full redundancy is required in any feed system which is needed for the provision of safe drinking water. Applications that can influence primary disinfection should be installed with capability to respond to real-time sensing of flow failure, or have monitoring of actual flow to automatically trigger the redundant feed system or shutdown the process. The standby units should have sufficient capacity to replace the duty units.

The design of chemical feed systems should incorporate suitable drains, flushing line adapters and other necessary appurtenances to enable safe flushing of hazardous chemicals to facilitate maintenance.

Overflow/overpressure relief systems should be piped back to the original storage tank.

Materials and surfaces resistant to the potential for corrosion by the chemical should be selected.

Where there is potential for chemical interactions between streams that may reduce their effectiveness, or cause a degradation of water quality or other harmful health and safety effects, chemicals should be conducted from the feeder to the point of application in separate conduits. Chemical feeders should be located as near as possible to the feed point. All chemical feed systems should be equipped with a means such as a graduated tube to allow calibration of the feed equipment.

Care should be taken to avoid the use of diffusers and static mixers with chemicals which form scale or other solids. Frequently used chemicals that can produce scale in hard water include sodium hypochlorite, ammonia, sodium silicate, sodium aluminate, sodium hydroxide (caustic soda), phosphate and polyphosphate blends, sodium carbonate (soda ash) and calcium hydroxide (lime). This is especially important where variable flow rates can be expected as this greatly reduces mixing power levels produced with static mixing arrangements. Reliance should be placed instead on ensuring high local turbulence through creating a permanent high velocity of the main stream and/or by using high velocity softened dilution/carrier water streams. For chemicals that undergo especially rapid interactions on contact with water such as coagulants, the use of power mixers is preferred in order to ensure rapid dispersion of the chemical. The mixing of coagulants is specifically addressed in Section 6.4.1 Coagulation/Flocculation Chemicals.

6.2 Facility design

6.2.1 General Storage & Handling of Chemicals

Storage for at least thirty days consumption at the maximum anticipated chemical usage rate should be provided, allowing for variations in chemical dosage and flow in that period. Where deliveries of chemicals can be expected to be interrupted by adverse weather conditions or in isolated locations, provision should be made for increased storage capacity. Where deliveries at short notice can be ensured, and the material is not essential to the production of safe water, storage requirements may be reduced.

Except where impractical in case of small water systems or where significant decay in chemical quality may be expected, sufficient storage should be provided to permit full load deliveries. The minimum recommended storage for truckload delivery is 1 ½ truck loads, or one truckload plus the quantity of chemical consumed in seven days, whichever is greater.

All chemical storage areas should be designed for containment of chemical spills. The minimum containment should be equal to 110% of the volume of the largest storage unit, or combination of units if interconnected, less the volume remaining in the container(s). Consideration of common header design and isolation for multiple tank installations should be made with respect to spill containment. Dissimilar chemicals should not share the same containment area. Containment walls should not be tied to the building walls, especially where floating slabs are used. Containments should undergo documented testing to demonstrate that they are leak proof.

Chemical or other process residuals should be handled in accordance with the requirements of Chapter 11 Waste Residuals Management.

Storage of chemicals inside buildings is recommended to avoid problems of materials freezing or becoming too viscous to pump, vandalism, high costs of providing fully weatherproof equipment, and the difficulty of containing gaseous spills.

The chemical storage area should be segregated from the main areas of the treatment plant, and separate storage areas should be provided for each chemical. Where chemicals in storage could react dangerously with other materials in storage (e.g., chlorine and ammonia, strong acids and bases, oxidants and fuels) segregated storage is required. The storage and feed equipment areas should be arranged for ease of restocking of chemicals, process operation and monitoring.

Off-loading areas should be clearly labeled to prevent accidental cross-contamination. Chemicals should be stored in covered or unopened shipping containers, unless the chemical is transferred into a storage unit made of NSF/ANSI certified materials. Bags, fibre drums and steel drums should be stored on pallets.

All chemical storage should be at or above the surrounding grade. Where sub-surface locations for chemical storage tanks are proposed, these locations should be free from sources of possible contamination, assure positive drainage for ground waters and provide for containment of chemical spills and overflows. Where above grade storage is provided, due consideration should be given to the method of unloading chemicals; for example, there is a limit on the allowable pressures to be used for air-padded trucks. Where drums or dry bagged chemicals are used, a loading dock or ramp should be provided.

Storage areas should be arranged to avoid and contain chemical spills, or liquid from clean-up operations, from entering the water under treatment. Floor surfaces should be smooth and impervious, slip-proof, and sloped so as to drain rapidly. Drains should be equipped with a normally closed valve to prevent accidental discharge of spilled substances.

When necessary, ventilation systems should be arranged so that air is exhausted outside the building and slight negative pressures are maintained where dry chemicals are in use as a dust control measure. Where large amounts of dust are anticipated, appropriate local exhaust systems and filters, scrubbers or dust separators should be provided in the ventilation system. Ventilation systems should be designed specifically for use in a corrosive environment and special measures taken in dust systems to prevent static build-up or explosion potential.

The designer should note that special precautions may be necessary in the design of air emissions control systems to prevent chemical concentrations at the point of impingement from exceeding limits permitted within the building or site under Air Pollution - Local Air Quality Regulation (O.Reg. 419/05), made under the Environmental Protection Act (EPA), or which might be hazardous. An approval under Section 9 of the EPA is required if a contaminant may be discharged into the air in the course of normal operation. For details, see the ministry document Guide for Applying for Approval (Air and Noise) (PIBS 4174e).

Chemical buildings or storage areas must be provided with eye-wash and/or deluge showers, adequate facilities for cleaning up chemical spills, space for cleaning and storage of the recommended protective equipment.

All doors in chemical buildings should open outward, and corridors or space between storage areas should be a minimum 1.5 m (5 ft) wide to permit the use of hand trucks or other equipment for safe movement of materials.

Carts, elevators and other appropriate mechanical means should be provided for lifting chemical containers to minimize excessive lifting by operators.

Where chemical solutions are prepared in batches by the operator, provision should be made for measuring quantities of chemicals used to prepare the feed solutions. For liquid chemicals, graduated cylinders or other calibrated containers, transfer pumps, load cells for use under storage tanks and flow meters should be provided. For dry chemicals weigh scales, volumetric feeders, calibrated solution tanks and flow meters should be provided.

6.2.2 Safe Handling Considerations

Chemical buildings or storage areas must be provided with adequate warning signs, conspicuously displayed where identifiable hazards exist, and a storage area for filing Material Safety Data Sheets (MSDS) as set out under the federal Hazardous Products Act and associated Controlled Products Regulations. An MSDS must be available for each chemical. All storage containers should be conspicuously labeled in accordance with the Workplace Hazardous Materials Information System (R.R.O. 1990, Regulation 860) under the Occupational Health and Safety Act (OHSA).

The WHMIS label includes: the product name, the supplier name, hazard symbol(s), risk, precautionary measures and first aid measures.

6.2.3 Liquid Chemicals

6.2.3.1 Fill Line

All storage tanks should be provided with an adequately sized fill line, minimum 50 mm (2 in) in diameter, sloped to drain into the tank. The fill line should be properly identified at the end remote from the tank, and provision should be made to drain this fill line.

6.2.3.2 Vent

Each tank should have an adequate vent line, minimum size 50 mm (2 in), with a down-turned end. Where venting outside is required, the vent should be provided with an insect screen. Securing vents that are externally accessible should be considered to minimize potential for contamination of the tank contents. The potential for moisture build-up resulting in vent freezing should also be considered.

6.2.3.3 Overflow

All tanks should have an overflow appropriate for the rate of fill proposed for the tank, sloped down from the tank, with a down-turned end and free discharge, located where it can be readily observed, within the containment area. Overflow pipes should not connect directly to the sewer and an air gap to prevent backflow/ siphonage should be provided.

6.2.3.4 Drain

Each tank should be provided with an accessible, valved drain, which should not discharge directly to a sewer, and should terminate at least two pipe diameters above the overflow rim of a receiving sump.

6.2.3.5 Level Indicator

Each tank should be provided with means to indicate the level of liquid in the tank, and where an external level gauge is provided, a shut-off valve at the tank connection is recommended. Low level and high level alarms, enunciated where an operator is present, should be provided for process chemical day tanks, where applicable.

6.2.3.6 Covers

Tanks should be provided with removable lids or covers where the contents are such that venting indoors is permitted. In the case of tanks which are to be vented outside, the covers should be constructed so as to be air tight, or with a slow stream continuously exhausted.

6.2.3.7 Lined Tanks

Where lined tanks are proposed, weep holes in the outer shell should be provided to give an indication of liner leakage.

6.2.4 Dry Chemicals

6.2.4.1 Dust

Granular materials are preferred to powders. Particular care should be taken to protect mechanical and electrical equipment from fine dust. Where exhaust fans, filters, and conveying systems are used, grounding should be provided to prevent the build-up of static electricity.

Floor drains should be provided for wash down of floors in the transfer/storage area.

6.2.4.2 Bulk Storage Silos & Feeders

Bulk storage silos should be provided with adequately sized fill openings. Fill lines, where necessary, should be smooth internally with long radius elbows. Silos should be provided with suitable level indicating devices, such as load cells. They should include a pressure relief valve when pneumatic fill systems are provided. Silo vent and exhaust systems should be provided with dust filters and/or cyclone type separators to prevent the release of dust into the atmosphere. Air exhausted from the handling areas should be directed away from air intakes.

The designer should take into account material characteristics such as flowability, tendency to pack tightly, angle of repose in the design of the silo bottom and method of removal of material to a feeder. Provision should be made to relieve bridging or rat-holing of the stored material, either by manual, mechanical or other means of rapping or agitating the hopper bottom or improving flowability of the material, for example by air fluidization.

6.2.4.3 Transfer

Provision should be made for the transfer of dry chemicals from shipping containers to storage bins or hoppers, in such a way as to minimize the quantity of dust which may enter the room in which the equipment is installed. Dust control should be provided by use of vacuum pneumatic equipment or closed conveyor systems, facilities for emptying shipping containers in special enclosures and/or exhaust fans and dust filters which put the hoppers or bins under negative pressure.

6.2.4.4 Disposal

Provisions should be made for disposing of empty bags, drums or barrels by a procedure approved by the ministry which will minimize exposure to dust.

6.2.5 Gaseous Chemicals

6.2.5.1 Storage Areas

Gas storage areas should be separated from the other areas, with separate outside accesses, and arranged to prevent the uncontrolled release of spilled gas to other areas of the plant and surrounding environment.

6.2.5.2 Measuring Contents

Means of measuring the contents of gas containers should be provided, and where necessary for the proper operation of the feed system, means of adjusting and indicating gas pressure/vacuum and flow rates should be provided. A system for automatic changeover (for disinfection related gases) when gas cylinders are empty is required.

6.2.5.3 Feed Rates

Where high feed rates are required by evaporation from liquefied gas, it may not be possible to withdraw the required gas quantity from a single cylinder due to evaporative cooling and the consequent reduction in gas vapour pressure. The designer should consider either using multiple cylinders on-line or the use of an evaporator to meet higher withdrawal rates.

6.2.5.4 Moving Cylinders

The designer should allow sufficient space in the storage area for convenient moving of cylinders from full storage to on-line to empty storage.

For chlorine, sulphur dioxide, ammonia and carbon dioxide gas systems, the designer is referred to the Chlorine Institute and gas equipment suppliers.

6.2.6 Chemical Feed Equipment & Control

The design and capacity of feed equipment should be such that it can supply the required quantity of chemicals at a continuous rate. Equipment should be capable of proportioning the chemical feed rate to flow, and dosage adjustments should be available through the anticipated ranges of both dosage and water flow. This requirement may not apply to installations where flow is essentially constant. As a minimum, the additive flow rate should be within plus or minus 10% of the optimum flow rate over the full range of expected water flow rates.

Feeders may be either manually controlled or automatically controlled with manual override. Feed systems should be equipped with means to confirm delivery rate. As a minimum, calibration tubes should be provided. For more critical applications, continuous flow failure sensing or delivered volume confirmation should be provided, including alarms and where appropriate, capability for automatic shutdown.

Continuous monitoring equipment should be provided for critical processes and as required by O.Reg. 170/03. Sample locations and the configuration of the sampling system should be representative of the actual conditions. Issues relating to reaction time and proper mixing should be considered.

Dosing devices should be appropriate for the chemical feed range and the precision needed. Continuous feed (avoiding pulsed feed with time delays between pulses which are significant in relation to pulse duration) is required when the chemical application is required for an immediate chemical reaction, as in the application of chlorine and sodium silicate for sequestering. Special consideration should be given when selecting equipment for very small dosing systems.

All positive displacement pumps should be equipped with adequately sized pressure relief valves. If the pumped fluid is relieved through this valve, it should pass to a safe location, preferably back to the storage tank. Where liquid filled diaphragm pumps are in use, the over-pressure should be relieved by discharge of the motive fluid to a safe location. Where oil-filled diaphragm pumps are used, the oil must be of a grade suitable for use in drinking water supplies (food grade). Pressure relief valves should be set not greater than 20 per cent higher than the pump discharge pressure in normal operation. Double diaphragm pumps are recommended for corrosive chemical pumping. Positive displacement rotary pumps should be used for chemical slurries.

Turndown capability should be considered when specifying pump capacity. If necessary, higher turndown ratios can be achieved through independent adjustment of motor speed and stroke length or application of drives that vary motor speed during the stroke cycle to adjust capacity. Multiple pumps should be considered where demand or flow varies widely.

Where reciprocating type pumps are in use, flexible connections should be provided on the pump suction and discharge to prevent the transmission of vibrations to the feed line. These flexible sections should be sufficiently rigid to withstand both the pump suction and discharge pressures, and reinforced hose is recommended.

The pump, in combination with its suction piping and valving arrangement, should be such that the pump discharge rate is not affected by fluctuations in storage tank level or a suction line calibration tube.

Volumetric or gravimetric feeders are suitable for dry chemicals and should provide effective means of dissolving or dispersing the material prior to addition to the water under treatment.

The use of remote ejectors and transmission under vacuum is recommended for gaseous chemicals to avoid pressurized lines passing through the plant.

Where solution tanks are in use, means should be provided to maintain a uniform strength solution and continuous agitation should be provided to maintain slurries in suspension. Make-up water for the solution tank should enter the tank with an air gap providing a complete physical separation [not less than 150 mm (6 in) or two pipe diameters, whichever is greater] between the free flowing discharge end and the flood level of an open tank, unless the make-up water supply has an approved backflow preventer.

Where the design of the chemical feed system includes day tanks, consideration should be given to the decay of the chemical, whether the chemical will be diluted and the decay rate of diluted solution. Sizing of day tanks should also consider the level of staffing (i.e. 24-hour operation) and the degree of automation. A single day tank providing 72 hours of chemical storage may be appropriate for hypochlorite storage at small waterworks where the combination of needed dilution and storage time combined with available space would make the duplication of storage tanks unnecessary.

Day tanks should either be scale mounted or have a calibrated level gauge. The piping arrangement for refilling the day tanks should be such that it will prevent over-filling of the tanks. In all other respects the requirements for day tanks should conform to the requirements for larger storage tanks.

Chemical feed lines should be kept as short as practical, especially suction lines, protected from freezing, and located to be readily accessible. Feed lines should be sized in accordance with flow. Chemical feed lines discharging to a pressurized system should be equipped with a backflow prevention device just upstream of the chemical injection point to prevent back-mixing. Consideration should be given to solution feed lines that empty during zero process flow conditions (if the solution feed point is at a higher elevation than the application point and drains out by gravity).

The designer should also consider the potential formation of scale and gas bubbles when sizing feed lines. The designer should allow for line flushing and chemical cleaning where hard water supplies, which could promote scaling, are used in solution preparation. Consideration may also be given to flexible lines in a carrier pipe which can be easily replaced.

Where chemical strainers are employed, the design should allow for easy cartridge removal, as particulate from liquid chemical storage may impact feed pump functionality.

Where feed lines are provided from multiple feeders or distributed to multiple application points, adequate valving should be provided to isolate appropriate sections of the supply system.

6.2.7 Chemical Application Points

Potentially corrosive chemicals should not be applied immediately preceding screens or pumping equipment, nor should solids-producing materials be applied prior to pumping. Thorough and timely mixing of chemicals into the water flow may be critical in many cases, and application points should be through a suitably designed diffuser or selected for high turbulence. The designer should be aware of the potential for the accumulation of scale or other solids on the diffuser and should make provision for its removal and cleaning. Application points should not be located where water flow splits. Chemical application points should take into account deposition due to interactions with the flow stream that may block the application ports.

The sequence of addition of chemicals should be evaluated for potential interactions (reactions) that may decrease or eliminate the intended process effect. For example, the use of activated silica and alum for coagulation will have more successful results if the activated silica is added downstream of the alum, but would be ineffective if both are added at the same time since they react together.

Backflow or siphonage between multiple feed points should be prevented. Where chemicals are added to pressurized lines, isolating valves should be provided.

6.2.7.1 Fluoride

The conventional application point for fluoride solutions is in filter effluent lines or in the clearwell. At plants using well supplies, it is usually advantageous to inject the solution into the discharge side of the well pump. The application point of hydrofluosilicic acid, if in a pipe, should be in the lower half of the pipe. Fluoride should be added downstream of the application point(s) of any coagulant, coagulant aid and/or softening chemicals, lime-soda softening or ion exchange softening processes. The designer should take into account the aggressive action of hydrofluosilicic acid (if used) on concrete structures at the point of application.

6.2.7.2 Chlorine

The designer should consider the potential formation of disinfection by-products when selecting chlorine application points. For minimizing disinfection by-product formation, chlorine should be applied as far downstream in the treatment process as possible or practical. Pre-chlorination may not provide reliable disinfection with low quality raw water. The designer should also refer to applicable regulations and procedures for more information regarding disinfection requirements.

If chlorine is to be piped to a pre-oxidation point in the plant intake (for zebra mussel control or other reasons), the transport pipe should be located within the intake pipe and precautions should be in place to prevent leaks.

6.2.8 In-plant Water Supply

In-plant water supply should be designed to satisfy treatment plant water demand and pressure needs. Means for measurement should be provided when preparing specific solution concentrations by dilution. Treatment for hardness should be considered for the water supply, due to the scale-forming potential of mixing alkaline chemicals with hard water. The water supply should be obtained from a location sufficiently downstream of any chemical feed point.

Where service water is used for ejector feed pressure regulation, the design should allow for impacts from fluctuating pressures on critical chemical feed systems that require steady service water delivery for accuracy of chemical application.

6.2.9 Backflow Prevention/ Cross-Connection Control

Backflow or siphonage protection and cross-connection control should be incorporated for all applicable chemical feed systems. The service water lines discharging to solution tanks should be protected from backflow. Provisions should be made to ensure that liquid chemical solutions cannot be siphoned through solution feeders into the water supply. The in-plant water supply should be protected against backflow by a reduced pressure principle backflow preventer on pressurized lines or by an air gap in other applications. No direct connection should exist between any sewer and a drain or overflow from the feeder, solution chamber or tank. The designer should refer to the Canadian Standards Association (CSA) standards CAN/CSA-B64 SERIES-01 Backflow Preventers and Vacuum Breakers, CAN/CSA-B64.10-01/B64.10.1-01 Manual for the Selection and Installation of Backflow Prevention Devices/Manual for the Maintenance and Field Testing of Backflow Prevention Devices, and B64.10S1-04/B64.10.1S1-Supplement #1 to CAN/CSA-B64.10-01/CAN/CSA-B64.10.1-01, the AWWA Manual of Water Supply Practices M14 Recommended Practice for Backflow Prevention and Cross-Connection Control and USEPA Cross-Connection Control Manual, 2003.

6.3 Operator safety

6.3.1 Legislation & Regulations

The safety of workers and workplaces is governed by the Occupational Health and Safety Act (OHSA) and the Workplace Safety and Insurance Act (WSIA), as well as the regulations made under these acts. The design of water systems must include provisions to protect operator and other worker safety and health. For further information, consult the Act and regulations, specifically the Industrial Establishments (R.R.O. 1990, Regulation 851) and the Workplace Hazardous Materials Information System (R.R.O. 1990, Regulation 860) regulations made under the Occupational Health and Safety Act (OHSA). The design of water systems should also take into consideration other applicable regulations such as building, electrical and fire codes.

6.3.2 Protective Equipment

Personal protective equipment should be provided for operators handling chemical compounds as required by applicable health and safety legislation. Deluge showers and eye wash stations should be provided where appropriate.

6.4 Specific chemicals

6.4.1 Coagulation / Flocculation Chemicals

Rapid mixing should be provided for all systems which utilize chemical addition in the form of coagulation and flocculation in the treatment process. Agitation may be provided through mechanical in-line mixers, static mixers or paddle-type mechanical agitators.

Chemicals injected to rapid mix units should be injected at a point close to the inlet of the rapid mix unit. Coagulant/flocculant aids should not be injected into the rapid mixing unit unless an additional rapid mixing unit for the coagulant/flocculant aid is provided. Coagulant and coagulant/flocculant aid addition should be derived from jar and/or pilot testing.

In-line static mixers are recommended for rapid mixing of primary coagulants, provided that the flow is mainly constant and near the design maximum flow rate; alternatively, powered mixers should be used. Primary coagulants should not be mixed using in-line devices such as pumps, weirs, valves or other such appurtenances, as they do not provide controlled mixing. High intensity mixing reduces the performance of coagulant aids, and these should be added only with low shear mixing after a delay period of ideally one or two minutes.

6.4.2 Chlorine Gas

All chlorination facilities should be designed according to the recommendations of the Chlorine Institute. Chlorine gas (Cl₂) feed and storage should be enclosed and separated from other operating areas. The chlorine room should be provided with a shatter resistant inspection window installed in an interior wall, constructed in such a manner that all openings between the chlorine room and the remainder of the plant are sealed, and provided with doors equipped with panic hardware, assuring ready means of exit and opening outward only to the building exterior.

Full and empty cylinders of chlorine gas should be isolated from operating areas, restrained in position to prevent upset and stored in rooms separate from ammonia storage. Cylinders should not be stored in areas exposed to direct sunlight or excessive heat.

Where chlorine gas is used, the room should be constructed to provide the following:

• Each room should have ventilation sufficient to produce 30 air changes per hour under emergency conditions and three air changes per hour under normal conditions when the room is occupied; where this is not appropriate due to the size of the room a lesser rate may be considered
• The ventilating fan should take suction near the floor as far as practical from the door and air inlet, with the point of discharge so located as not to contaminate air inlets to any rooms or structures;
• Air inlets should be through louvers near the ceiling;
• Louvers for chlorine room air intake and exhaust should facilitate airtight closure;
• Separate switches for the fan and lights should be located outside of the chlorine room and at the inspection window. Outside switches should be protected from vandalism. A signal light indicating fan operation should be provided at each entrance when the fan can be controlled from more than one point;
• Vents from feeders and storage areas should discharge to the outside atmosphere, above grade;
• The room location should be on the prevailing downwind side of the building away from features such as entrances, windows, louvers and walkways;
• Floor drains are discouraged. Where provided, the floor drains should discharge to the outside of the building and should not be connected to other internal or external drainage systems; and
• The need to install an absorption scrubber or any other device should be evaluated by the designer based on the site specific conditions including: volume of chlorine storage, type of containers, rate of chlorine gas withdrawal, and distance to public buildings or to the nearest point of impingement. The designer should consider the release of chlorine gas via safety relief systems to the environment. The design of absorption scrubbers should be based on environmental protection criteria rather than personnel safety criteria or facility protection. Confinement and local adsorption is an alternative that may also be considered.

Chlorinator rooms should be heated to 15°C (59°F) and be protected from excessive heat. Cylinders and gas lines should be protected from temperatures above that of the feed equipment. Pressurized chlorine feed lines should not carry chlorine gas beyond the chlorinator room.

6.4.2.1 Chlorination Equipment

Solution feed gas chlorinators or hypochlorite feeders of the positive displacement type should be provided.

The chlorinator capacity should be such that a free chlorine residual of at least 2 mg/L can be maintained in the water after a contact time of at least 30 minutes at the anticipated maximum flow rate. The equipment should be designed to ensure that it will operate accurately over the desired feeding range.

Standby equipment of sufficient capacity should be available to replace the largest unit. Spare parts should be made available to replace parts subject to wear and breakage. If there is a large difference in feed rates between routine and emergency dosages, a gas metering tube should be provided for each dose range to ensure accurate control of the chlorine feed.

Automatic switch-over of chlorine cylinders is needed to ensure continuous disinfection.

Automatic proportioning chlorinators are recommended where the rate of flow or chlorine demand varies by more than plus or minus 10%.

Each eductor should be selected for the point of application with particular attention given to the quantity of chlorine to be added, the maximum injector water flow, the total discharge back pressure, the injector operating pressure and the size of the chlorine solution line. Gauges for measuring water pressure and vacuum at the inlet and outlet of each eductor should be provided.

The chlorine solution injector/diffuser should be compatible with the point of application to provide a rapid and thorough mix with all the water being treated. The centre of a pipeline is the preferred application point.

The chlorinator water supply piping should be designed to prevent contamination of the treated water supply by sources of questionable quality. At all facilities treating surface water, pre- and post-chlorination systems should be independent to prevent possible siphoning of partially treated water into the clearwell. The water supply to each eductor should have a separate shut-off valve. Master shut-off valves are not recommended.

The pipes carrying liquid or dry gaseous chlorine under pressure, as well as all chlorine solution piping and fittings, must be constructed of materials recommended by the Chlorine Institute. Nylon products are not recommended for any part of the chlorine solution piping system.

6.4.3 Sodium Hypochlorite

Sodium hypochlorite storage and handling procedures should be arranged to minimize decay either by contamination or by exposure to more extreme storage conditions. In addition, feed rates should be regularly adjusted to compensate for this progressive loss in chlorine content.

Sodium hypochlorite should be stored in the original shipping containers or in sodium hypochlorite compatible containers. Storage containers or tanks should be sited out of the sunlight in a cool area and vented to the outside of the building.

Wherever reasonably feasible, stored hypochlorite should be pumped undiluted to the point of addition. Where dilution is unavoidable, deionized or softened water should be used. Injectors should be selected to be resistant to scale blockage or should be made removable for regular cleaning where hard water is to be treated. For small ground water supplies which are typically served by a hypochlorite metering pump controlled by the well pump in an on-off operation, the dilution of hypochlorite to lower concentrations may be necessary for proper operation of a regular diaphragm metering pump unless a small volume dosing system is used.

Pipe system design should allow for adequate flushing of components primarily for operator/maintenance personnel safety considerations.

Storage areas, tanks and piping should be designed to avoid the possibility of uncontrolled discharges, and a sufficient amount of appropriate spill absorbent should be stored on-site.

Reusable hypochlorite storage containers should be reserved for use with hypochlorite only and should not be rinsed out or otherwise exposed to internal contamination.

Positive displacement pumps with hypochlorite compatible materials for wetted surfaces should be used. To avoid air locking in smaller installations, small diameter suction lines should be used with foot valves and degassing pump heads. In larger installations, flooded suction should be used with piping arranged to ease escape of gas bubbles.

6.4.3.1 On-Site Generation of Sodium Hypochlorite

Proprietary equipment is available for the production of dilute hypochlorite solution by electrolysis of sodium chloride brine. The design should provide for a supply of softened water for brine makeup in order to protect the electrode life and to safely discharge an off-gas stream of hydrogen. A day tank for storage of the hypochlorite solution should also be provided. This process may be more advantageous in remote locations where transportation/delivery is an issue. The designer should consult the manufacturer for specific design requirements.

6.4.4 Chlorine Dioxide

Chlorine dioxide gas, even in mixtures of over 10% in air, is highly unstable and as a result, it must be generated on-site. The gas should be handled only in water solution with feed lines arranged to avoid gas pocket formation, be maintainable under moderate pressure and be easily water purged. The gas is toxic; therefore, the designer should make appropriate provisions to protect operations staff from excessive exposure.

6.4.4.1 Chlorine Dioxide Generators

Chlorine dioxide generation equipment should be factory assembled and pre-engineered units with a minimum efficiency of 95%. The excess free chlorine should not exceed 3% of the theoretical stoichiometric concentration required.

Common continuous generators require a stream of sodium chlorite solution and a carefully proportioned stream of chlorine. Where chlorine is not available, hypochlorite solution and an acid may be used in a three-feed reactor. When feed streams are correctly proportioned, these generators can show an efficiency in generating chlorine dioxide of over 95%.

A variant generator that uses only hydrochloric acid and chlorite solution is simpler to feed but operates at lower conversion efficiency. It produces chlorine dioxide that does not contain any elemental chlorine contamination; as a result it does not form any THMs or HAAs when used as a disinfectant.

The design of chlorine dioxide equipment should conform to all applicable design criteria relating to chlorination equipment (Section 6.4.2.1 Chlorination Equipment) as well as to the manufacturer recommendations.

6.4.5 Sodium Chlorite

Sodium chlorite is used for chlorine dioxide generation. Chlorite may also be effective in inhibiting nitrifying bacterial activity in some chloraminated distribution systems.

Sodium chlorite is available in concentrated solution or granular form. The liquid form is preferred for ease of handling.

Sodium chlorite should be stored by itself in a separate room and preferably in an outside building detached from the water treatment facility. It should be stored away from organic materials because many materials may catch fire and burn violently when in contact with chlorite.

The storage structures should be constructed of non-combustible materials. If the storage structure must be located in an area where a fire may occur, water should be available to keep the sodium chlorite area cool enough to prevent explosion. The design should take into consideration explosion proof equipment requirements (if applicable).

Positive displacement feeders should be provided. Tubing for conveying sodium chlorite or chlorine dioxide solutions should be Type 1 PVC, polyethylene or materials recommended by the manufacturer.

Where appropriate, check valves should be provided to prevent the backflow of chlorine into the sodium chlorite line.

6.4.6 Ammonia

Production of monochloramine for secondary disinfection can be achieved by the addition of ammonia or ammonium salts, usually to prechlorinated water streams. Correct proportioning and effective mixing are required to avoid the formation of odorous dichloramines and trichloramines. Care should also be taken to avoid operating at the breakpoint, where there may be no residual chlorine. Ammonia for chloramine formation may be added to water either as a water solution of ammonium sulphate, as aqua ammonia (ammonium hydroxide), or as anhydrous ammonia (purified ammonia in liquid or gaseous form). Special provisions required for each form of ammonia are listed below. The designer should consult with the chemical manufacturer/supplier regarding additional specific requirements for handling, storage and feeding. Continuous gentle extractive ventilation of the ammonia area is advised to protect copper wiring and other metallic items from accelerated corrosion.

6.4.6.1 Ammonium Sulphate

A water solution is made by mixing ammonium sulphate solid with water. The tank and dosing equipment contact surfaces should be made of corrosion resistant non-metallic materials. Provision should be made for removal of the agitator after dissolving the solid. The tank should be fitted with a lid and vented outdoors. AWWA Standard B302: Ammonium Sulfate provides additional safety, handling and storage information.

6.4.6.2 Aqua Ammonia

Ammonia solutions may be obtained from suppliers at concentrations under 20% w/w. This more dilute product minimizes the ammonia vapour pressure over the liquid and the associated handling difficulties. Aqua ammonia feed pumps and storage should be enclosed and separated from other operating areas.

The aqua ammonia should be conveyed directly from storage to the injector without the use of a carrier water stream unless the carrier water is softened. Provision should be made for easy access for removal of scale deposits from the injector.

The designer may consider the installation of a small capacity scrubber capable of handling occasional ammonia emissions which may occur during tank filling.

6.4.6.3 Anhydrous Ammonia

Anhydrous ammonia is readily available as a pure liquefied gas under moderate pressure in cylinders or as a cryogenic liquid boiling at -15°C (5°F) at atmospheric pressure. The liquid causes severe burns on skin contact.

Anhydrous ammonia and storage feed systems (including heaters where required) should be enclosed and separated from other works areas and constructed of corrosion resistant materials. Pressurized ammonia feed lines should be restricted to the ammonia room.

An emergency air exhaust system, with an elevated intake, should be provided in the ammonia storage room. Leak detection systems should be fitted in all areas through which ammonia is piped.

Special vacuum breaker/regulator provisions should be made to avoid potentially violent results of backflow of water into cylinders or storage tanks.

Carrier water systems of soft or pre-softened water are essential for transporting ammonia to the finished water stream and to assist in mixing. The ammonia injector should use a vacuum eductor or should consist of a perforated tube fitted with a closely fitting flexible rubber tubing seal punctured with a number of small slits to delay fouling by lime deposits. Provision should be made for the periodic removal of scale/lime deposits from injectors and carrier piping.

Where storage units are housed in enclosed areas, consideration should be given to the provision of an emergency gas scrubber capable of absorbing the entire contents of the largest ammonia storage unit whenever there is a risk to the public as a result of potential ammonia leaks.

6.4.7 Chemicals Used in Dechlorination Facilities

Dechlorination is practiced to partially or totally remove chlorine residual. The most commonly used reducing agent is sulphur dioxide. Other common dechlorination chemicals include sodium sulphite, sodium bisulphite, sodium metabisulphite, sodium thiosulphate and hydrogen peroxide. Table 6.1 shows the theoretical amount of each dechlorination chemical needed to remove one part of chlorine residual.

Gaseous storage facilities may house gaseous sulphur dioxide for dechlorination requirements. Most precautions for gaseous chlorine (Section 6.4.2 Chlorine Gas) will also apply for sulphur dioxide. The designer should consult the manufacturer for specific details regarding storage and handling requirements for sulphur dioxide and other reducing agents used for dechlorination.

Guidance on dechlorination can be found in the AWWA Manual of Water Supply Practices M20 Water Chlorination/Chloramination Practices and Principles, and the AwwaRF report Guidance Manual for Disposal of Chlorinated Water (Project #2513).

6.4.8 Acids & Bases

Strong acids and bases should be stored in separate areas. Concentrated acids and bases should be kept in closed corrosion-resistant shipping containers or storage units. Acids and bases should not be handled in open vessels, but should be pumped in undiluted form from original containers through suitable hose to a dilution tank, or where dilution is not required to aid process control, to the point of treatment or to a covered day tank. Separate ventilation systems may be required for areas where concentrated acids or bases are stored.

6.4.9 Ozone

6.4.9.1 General

Ozone is an unstable and very powerful oxidant and needs to be generated electrically on-site.

For small systems, very simple, low maintenance pre-engineered treatment units producing ozone from ambient air at up to a very few grams of ozone per hour are used.

For taste and odour control, and for primary disinfection, larger outputs of ozone are generally required using proprietary equipment. The designer should consult the manufacturer for power, air-feed, cooling and temperature requirements, as well as contactor design and ozone destructor requirements. Additional guidance for the design of large scale ozonation systems is provided in the Ten States Standards.

High purity oxygen used as feed gas can be purchased and stored as a liquid (LOX), or it can be generated on-site through either a cryogenic process, vacuum swing adsorption (VSA) or pressure swing adsorption (PSA). Cryogenic generation of oxygen is a complicated process and should only be considered for large systems. Storage of LOX is governed by regulations in building and fire codes. These regulations will impact the space requirements and may dictate the construction materials of adjacent structures.

The generators can be low, medium or high frequency type. Specifications should require that the transformers, electronic circuitry and other electrical hardware be proven, high quality components designed for ozone service. Appropriate ozone generator backup equipment should be provided.

Only low carbon 304L or 316L grade stainless steel piping should be used for ozone service with 316L being the preferred material. Connections on piping used for ozone services should be welded, where possible. Connections with meters, valves or other equipment are to be made with flanged joints with ozone resistant gaskets, such as Teflon™ or Hypalon™. Screwed fittings should not be used because of their tendency to leak.

Ozone monitors should be installed to measure ozone concentration on both the feed gas and off-gas from the contactor and in the off-gas from the destruct unit. For disinfection systems, monitors should also be provided for monitoring ozone residuals in the water. The number and location of ozone residual monitors should be such that the amount of time that the water is in contact with the ozone residual can be determined.

A minimum of one ambient ozone monitor should be installed in the vicinity of the contactor and a minimum of one should be installed in the vicinity of the generator. Ozone monitors should also be installed in any areas where ozone gas may accumulate. Ambient ozone exposure levels are governed by the Occupational Health and Safety Act (OHSA).

6.4.9.2 Ozone Contactors

The specific process objective should dictate contact basin design. Reactions that are rapid relative to the ozone mass transfer rate from gas to liquid phase are best served by contactors that promote the maximum transfer of ozone in the shortest period of time. For these applications, such as oxidation or iron, manganese or simple organics, contact time is often less important and contactors that rely on single points of application may be suitable. For reactions that are slow relative to the ozone mass transfer rate, such as disinfection or oxidation of complex organics (such as herbicides or pesticides), contact time is critical and favours contactors with extended detention time and multiple application points, such as the conventional multi-stage fine bubble diffuser design.

The types of contactors that are commonly used include:

• Conventional fine bubble
• Turbine
• Packed column
• Injectors
• Deep U-tube

Bubble Diffusers

The most common design is a multichamber over-under baffled contactor with ozone addition to the first one or two chambers via diffusers situated at the bottom of the chambers. Where disinfection is the primary application, a minimum of two contact chambers each equipped with baffles to prevent short circuiting and induce counter-current flow should be provided. The water depth in the contactor is typically between 4.6 and 6 m (15 and 20 ft) to achieve high transfer efficiency of the added ozone.

Once the immediate or initial ozone demand has been satisfied and a residual ozone level has been maintained in contactor stages, the designer may consider the provision of a "passive" stage without ozone diffusers, which will serve to increase the required retention time without the cost of the additional diffusers and associated piping.

The contactor should be designed to ensure good flow dispersion and to avoid short-circuiting. The diffusion system should optimize liquid/gas contact and maximize mass transfer. For optimum ozone transfer efficiency, gas bubble size should be between 2 and 5 mm (0.08 and 0.2 in).

Ozone should be applied using porous tube or dome diffusers. For ozone applications in which precipitates are formed, such as with iron and manganese removal, porous tube diffusers are not recommended. Dome diffusers are generally preferred as they produce finer bubbles. Due to head loss limitations, proprietary diffusers typically have a maximum gas flow rating that should not be exceeded. The chamber floor area should therefore be large enough to accommodate the minimum number of diffusers.

Contactors should be separate closed vessels that have no common walls with adjacent rooms. The contactor should be kept under negative pressure and ozone monitors should be provided to protect worker safety. Placement of the contactor where the entire roof is exposed to the open atmosphere is recommended.

Large contact vessels should be made of reinforced concrete. All reinforcement bars should be covered with a minimum of 38 mm (1.5 in) of concrete. Smaller contact vessels can be made of stainless steel, fibreglass or other material which will be stable in the presence of residual ozone and ozone in the gas phase above the water level.

Multiple sampling ports should be provided to enable sampling of effluent water in each compartment and to confirm CT calculations.

Contactors should be designed with approximately 1 m (3 ft) of headroom to provide for unimpeded gas flow to the off-gas exit. Where necessary, a system should be provided between the contactor and the off-gas destruct unit to remove froth from the air and return the froth to the contactor or other acceptable location. If foaming is expected to be excessive, then a potable water spray system should be placed in the contactor head space.

All contactors should have provisions for cleaning, maintenance and drainage of the contactor. The basic arrangement of the contactor will establish the provisions required for personnel access. Stainless steel hatches with ozone-resistant gaskets should be provided for each contactor stage to minimize difficulty and delay of entry and exit from the contactor under regular and emergency conditions. The design should also provide for exhaust ventilation prior to the entry of maintenance personnel

Other Contactors

Other contactors may be acceptable, provided adequate ozone transfer is achieved and the required contact times and residuals can be met and verified.

6.4.9.3 Ozone Destruction Unit

A system for treating the final off-gas from each contactor should be provided in order to meet safety and air quality standards. Acceptable systems include thermal destruction and thermal/catalytic destruction units.

In order to reduce the risk of fires, the use of units that operate at lower temperatures is encouraged, especially where high purity oxygen is the feed gas.

The maximum allowable ozone concentration in the air discharge is 0.1 mg/L (by volume).

At least two units should be provided which are each capable of handling the entire gas flow.

Exhaust blowers should be provided in order to draw off-gas from the contactor into the destruction unit.

Heat exchangers and catalysts should be protected from froth, moisture and other impurities which may harm the units.

The catalyst and heating elements should be located where they can easily be reached for maintenance.

6.4.9.4 Alarms

The following alarms/shutdown systems should be considered at each installation:

• Dew point shutdown/alarm This system should shut down the generator in the event the system dew point exceeds -60°C (-76°F);
• Ozone generator cooling water flow shutdown/alarm This system should shut down the generator in the event that cooling water flow decreases to the point that generator damage could occur;
• Ozone power supply cooling water flow shutdown/alarm This system should shut down the power supply in the event that cooling water flow decreases to the point that damage could occur to the power supply;
• Ozone generator cooling water temperature shutdown/alarm This system should shut down the generator if either the inlet or outlet cooling water exceeds a certain preset temperature;
• Ozone power supply cooling water temperature shutdown/alarm This system should shut down the power supply if either the inlet or outlet cooling water exceeds a certain preset temperature;
• Ozone generator inlet feed gas temperature shutdown/alarm This system should shut down the generator if the feed gas temperature exceeds a certain preset value;
• Ambient ozone concentration shutdown/alarm The alarm should sound when the ozone level in the ambient air exceeds 0.1 ppm or a lower value chosen by the municipality/owner. Ozone generator shutdown should occur when ambient ozone levels exceed 0.3 mg/L (or a lower value) either in the vicinity of the ozone generator or the contactor; and
• Ozone destruct temperature alarm The alarm should sound when temperature exceeds a preset value.

6.4.10 Hydrogen Peroxide

Advanced oxidation processes using a combination of peroxide and ozone or peroxide and UV have been developed which require a hydrogen peroxide feed system and an ozone generation or UV system. For AOPs using ozone, hydrogen peroxide can be added upstream or downstream of ozone, or simultaneously. Where UV is used, peroxide is added either upstream or simultaneously with UV. Ozone generation systems are discussed in Section 6.4.9 Ozone. Design considerations for UV systems are discussed in Section 5.9.5 Ultraviolet Light Inactivation. Peroxide removal should be accomplished before the application of chlorine.

Peroxide is a strong oxidant and contact with personnel should be avoided. Secondary containment should be provided for storage tanks to contain any spills. Dual containment piping should be considered to minimize the risk of exposure to plant personnel.

Peroxide can be stored in high density polyethylene or 304L or 316L grade stainless steel drums or tanks. Peroxide can be stored on-site but deteriorates gradually over time. Peroxide deteriorates rapidly if contaminated and with heat or exposure to certain materials. Excessive heat may cause a tank rupture due to gas generation if the tank is not vented properly. Tanks should be vented according to the manufacturer/supplier specifications. Peroxide has a lower freezing point than water, however, housing or heat tracing should be provided for storage tanks and exterior piping if extended periods with temperatures below freezing are anticipated. Hydrogen peroxide, at concentrations of 35% and 50%, freezes at temperatures of -40°C (-40°F) and -45°C (-49°F) respectively.

Pipes, gaskets and metering pumps should be constructed of peroxide resistant materials. The designer should ensure that all wetted stainless steel components are passivated using industry accepted passivation procedures. Pumps should be designed to prevent potential air binding of peroxide off-gas. Adequate mixing should be provided. It is recommended that all peroxide chemical dosing systems be provided with safety relief valves in areas where hydrogen peroxide can become trapped.

6.4.11 Potassium Permanganate

Potassium permanganate solution decomposes slowly and, as a result, is better purchased as a granular solid. Potassium permanganate may be supplied in dry form in buckets, drums and bulk. A concentrated potassium permanganate solution (1 to 4%) can be generated on-site for water treatment applications. Depending on the amount of permanganate required, these solutions should be made up in batch modes, using storage tanks with mixers and a metering pump for small feed systems. Larger systems should include a dry chemical feeder, storage hopper and dust collector configured to automatically supply permanganate to the solution storage tank.

In conventional treatment plants, potassium permanganate solution is added to the raw water intake, or as far upstream of coagulant addition as possible. Adequate mixing should be provided. In all cases, potassium permanganate should be added prior to filtration.

Potassium permanganate solution should be pumped from the solution tank to the injection point. If the injection point is a pipe, a standard injection nozzle protruding midway into the pipe section should be used. Injection nozzles can also be used to supply the solution to mixing chambers and clarifiers. Powered activated carbon (PAC) and potassium permanganate should not be added concurrently. PAC should be added downstream of potassium permanganate because it may adsorb permanganate, rendering it unavailable for the oxidation of target organics.

6.4.12 Phosphates & Polyphosphates

Stock phosphate solutions should be kept covered and disinfected by maintaining approximately 10 mg/L chlorine residual. Phosphate solutions having a pH of 2.0 or less may be exempt from this requirement.

Polyphosphates should not be applied ahead of iron and/or manganese removal treatment. The point of application should be prior to any aeration, oxidation or disinfection if no iron or manganese removal treatment is provided.

Feed and storage equipment should conform to the requirements specified by the manufacturer.

6.4.13 Powdered Activated Carbon (PAC)

The designer should consider the possibility of powered activated carbon (PAC) addition at several points within the treatment process. PAC addition should take place as far upstream of coagulant addition as possible, preferably with mechanically aided mixing. The designer should avoid feeding chlorinated water to any form of carbon.

PAC can be added as a premixed slurry or by means of dry feed. Continuous agitation should be provided to ensure that the PAC does not deposit in the slurry storage tank.

PAC should be considered as a potentially combustible material and should be stored in a separate fire retardant building or room equipped with explosion proof lighting and electrical systems. Wet activated carbon may create an oxygen-deficient environment in enclosed spaces, therefore, appropriate safety precautions should be provided. The manufacturer recommendations regarding storage and handling should be followed.

Provision should be made for adequate dust control. Provisions should also be made to scrub or filter the carrier air when dry PAC is off-loaded into silos.

6.4.14 Fluoride

Sodium fluoride, hydrofluosilicic acid and sodium silicofluoride may be used for fluoridation. These compounds are highly corrosive and require specific considerations. In addition to these guidelines, the designer should refer to a fluoride manual such as AWWA Manual of Water Supply Practices M4 Water Fluoridation Principles and Practices.

Water used for sodium fluoride dissolution should be softened if hardness exceeds 75 mg/L as calcium carbonate.

For smaller systems, the electrical outlet used for the fluoride feed pump should have a non-standard receptacle and should be interconnected with the well or service pump;

Saturators should be of the upflow type and be provided with a meter and backflow protection on the make-up water pipe.

Construction should be of corrosion resistant material. The use of explosion proof motors and electrical components should be considered. Light and fan switches should not be located within the fluoride room.

6.4.15 Carbon Dioxide

Where carbon dioxide is added for pH adjustment, the contact chamber should provide a detention time sufficient to ensure effective absorption. The recarbonation basin design should provide:

• A total detention time of 20 minutes;
• A minimum of two parallel compartments, with a depth that will provide a diffuser submergence of not less than 2.3 m (7.5 ft) nor greater submergence than recommended by the manufacturer, as follows:
  • A mixing compartment having a detention time of at least three minutes; and
  • A reaction compartment.
• The practice of on-site generation of carbon dioxide is discouraged;
• Where liquid carbon dioxide is used, adequate precautions should be taken to prevent carbon dioxide from entering the plant from the recarbonation process;
• Recarbonation tanks should be located outside or be sealed and vented to the outside with adequate seals and adequate purge flow of air to ensure worker safety; and
• Provisions should be made for draining the recarbonation basin and removing sludge.