CHAPTER 12 | Water Treatment

Water treatment is a necessary adjunct to evaporative systems. Water treatment services are provided by companies that specialize in this type of work. They are charged with keeping the water quality sufficient to prevent scaling, corrosion and biological fouling or attack.

Scale and corrosion are generally thought to be diametrically opposed to one other. Reducing scale build up, for example, exacerbates corrosion and vise versa. Both are addressed in unison. The biological aspect of water treatment comes from living organisms that thrive in the recirculated water and wetted surfaces. Bacteria, slime and algae can foul heat exchanger surfaces and in some cases attack and destroy system components. Chemical treatments address biological issues separately from scale and corrosion.

Scale formation has its root in the evaporation of water. Evaporated water exits the system as pure vapor leaving the solids behind. The replacement (make-up) water introduces more solids which continually increase the solids concentration in the recirculated water. Left unchecked, the system would reach a point where the water could not hold all of the solids in a dissolved state. They would begin to precipitate out of solution as scale.

The common tea kettle demonstrates the build up of scale as water is evaporated.

Someone from Los Angeles, CA vacationing in Olympia, WA may notice a vast difference in the tendency for tap water to build scale. This is because naturally occurring water in some parts of the nation has more dissolved solids than others. Inspection of the solids will also yield different components. Some areas have a preponderance of calcium chloride. Others, silica etc.. The necessary steps to combat scale differ by geographic location; Water treatment services should be performed by people familiar with the local water quality.

Evaporating enough water to make the solids increase to twice their initial value is a two fold increase in solids content. (ex: 60 parts/million becomes 120ppm.) The newly constituted water is said to have "two cycles of concentration". Likewise, it can be increased to 3,4,6,10.5,...cycles depending on how much water is evaporated.

Clearly, water that has few initial dissolved solids can attain a very high number of 'cycles' before the solids precipitate from solution. Conversely, water with high initial solids can only be 'cycled' a small amount before precipitation occurs.

In either case, a saturation point will be reached where the cycles cannot be increased. Every particle that dissolves is offset by another particle that comes out of solution as scale. The water treatment service provider is aware of the dissolved solid content of the water in his/her area and knows how many cycles are acceptable. When in doubt, a make-up water sample can be taken to demonstrate the quality of the water entering the system. Each constituent is examined against a maximum allowable concentration.

Generally, it is calcium carbonate that dictates the maximum allowable cycles. (ex: make-up water tests show Xppm CaCo3 which is known to precipitate from solution at Yppm. The maximum concentration of CaCo3 is, then, Y/X cycles. As a practical matter, the water treater will configure the system to 'cycle' somewhat less than this maximum amount.)

Some locations may require that the cycles be set based on keeping some other constituent- like silica- below a maximum threshold value.

Recognizing the concept that the fewer the initial solids the better, some operators are tempted to use soft water as make-up theorizing that since most of the solids are removed, the cycles can be allowed to reach astronomical levels without scale formation.

This approach is ill advised without the input of a competent water treatment expert who is capable of combating the excessive corrosivity of such water.

Making the scenario more complex is the fact that water treaters add chemicals that allow the water to retain increased solids.

The optimum value for 'Cycles' is a bit elusive... a high value leads to reduced chemical, water and sewage costs but introduces an increased risk of scale formation and vise versa. In the end, a design value is determined.

Next, the water treatment controls are set to maintain the design value. This could be a 'continuous bleed' where a set portion of the recalculated water is intentionally wasted to the drain. The make-up float valve (sort of like a toilet tank float and valve) introduces fresh water to replace that which is evaporated and bled. The new water mixes with system water diluting the solids concentration.

While initially inexpensive, the continuous bleed is rarely used and, in fact, is often illegal. Here, a valve is set to waste water to the drain at a rate necessary to maintain the design cycles at maximum evaporation. The problem is, towers typically aren't called upon to evaporate water at the maximum rate most of the time allowing the cycles to plummet, wasting water.

More common is the use of a 'conductivity monitor' which operates on the principle that the conductivity of water increases in direct proportion to its solids concentration. The device is first used to measure the conductivity of the make-up water and then set to initiate a bleed cycle when the system conductivity reaches a value equal to this initial reading x cycles. The duration of the bleed cycle is also set by the technician. Scale and corrosion inhibitors are typically injected into the system as it bleeds. Liquid chemicals are introduced by small, adjustable, plastic, positive displacement pumps that meter precise dosages.

It is helpful to imagine the entire system surrounded by a bubble and examine the amount of water that enters and that which exits- they must be equal.

V = water that exits as vapor
B = water that exits as bleed
D = water that exits as drift (tiny water particles flung from the tower)
M = Water that enters as make up

Then, M = V + B (When we ignore the insignificant drift losses) (eq.1)
Recognizing that in order to keep from making scale, all of the solids that enter as make- up must exit as bleed, it follows that:

M = cycles x B (eq.2)
And that:
M = V [(cycles)/(cycles-1)] (eq.3)
Combine (eq.2) and (eq.3) to get:
B = V/ (cycles-1) (eq.4)
Combine (eq.1) and (eq.3) to get:
B = M [1/cycles] (eq.5)


So, when cycles = 2, Bleed = 1/2 Make-up
= 3, Bleed = 1/3 Make-up
= 4, Bleed = 1/4 Make-up
etc.

Consider the case of two cycles... The bleed rate works out to half the make-up rate. Assuming good mixing and that bleed occurs prior to solid particles precipitating out of solution (i.e.: all the dissolved solids in the make-up water exit in the bleed water and do not stay behind as scale) the solids concentration of the bleed has to be twice that of the make-up if the flow is half.

Similar arguments can be made for 3,4,5...cycles.

The tendency for any system to grow biological material depends on several factors.

Cooling tower design is one. Crossflow towers and Counterflow towers without louvers, for example, tend to grow more algae due to the increased amounts of sunlight in contact with the system water. Water quality also comes into play. Make-up water that is reclaimed from a sewage treatment plant, for example, can be rich in nutrients. Also, some food processing operations where beer, tomato paste, milk, sugar, etc. enter the cooling system can have severe corrosion and biological problems.

Another potential lies with air quality. Cooling towers located near bakeries, for example, show an increased tendency to grow biological material due to the molds and yeast. Biological concerns run the gamut from nearly zero to very substantial. Whatever the case, an appropriate solution must be developed.

The traditional approach is for the operator to alternate between two liquid biocides adding them at a predetermined frequency. Two different formulations are often used to avoid an immunity being developed to just one. Unlike scale and corrosion chemicals that are metered into the system frequently, biocides are typically administered every few days to 'shock' the system.

Other chemicals for biological control include chlorine, iodine, bromine, hydrogen peroxide and ozone. These are not rotated with other chemicals; They are fed continuously.

Ozone is also used to prevent scale. Be certain to specify Viton pump seals when using ozone; Standard seals will fail quickly.

General: The cooling tower has the misfortune of being a handy receptacle for the addition of water treatment chemicals which are almost always corrosive in concentrated form. The point of injection of scale and corrosion chemicals is important. They should not be dripped into the top of a cooling tower where the air can blow them against metal components. They should be introduced into the piping where they will disperse quickly.

Similarly, pelletized chlorine tablets tossed into the sump can burn holes through the basin floor.

The foregoing 'bleed and feed' (bleed water while feeding chemicals) for scale and corrosion prevention plus dual biocides constitute the classic water treatment approach. Other, more modern, chemical free systems are starting to be marketed.

The desire to save water costs by increasing cycles draws some to unreasonably jeopardize system performance and longevity.

There is also a general failure to recognize that cooling water quality can be very dynamic. Do not, for example, make the mistake of installing a new tower, placing it into operation, and ignoring the water treatment for a few days. Some closely coupled systems with small water volumes (evaporative condensers and fluid coolers lending the best examples) can be scaled in a matter of hours.

Removing scale is akin to chemotherapy...The patient can die from the cure. Having to descale is an indication that the water treatment program has failed. Steps must be taken to prevent a reoccurrence. Every descaling episode strips valuable life from the equipment and connecting pipework.

Cooling towers do not suffer from the accumulation of scale as quickly as their evaporative condenser and fluid cooler counterparts. Scale can be allowed to remain in a cooling tower if it is not so thick as to inhibit airflow. Just descale the device being cooled- typically a condenser bundle- by chemical or mechanical means. When cleaning chemically, isolate the heat exchanger and circulate the chemical solution through it with an auxiliary pump following manufacturer recommendations.

Evaporative condensers and fluid coolers have been successfully descaled using a five percent (by weight) solution of inhibited sulfamic acid. It must be monitored very carefully- ask the manufacturer for recommendations.

One final comment... Do not view water treatment as mere chemicals and hardware- a one time expense that forever ends concern for the subject. Indeed, these items tend to be generic but it is the quality and dedication of the people who monitor the system, make adjustments, etc. that make or break a system. Proper installation of components is essential with injection points being most critical. Added connections must be permanent and leak free. Avoid, for example, merely sticking a pvc pipe or plastic hose through a side wall. All it takes is for someone to trip on the hose or for water movement in the tower to allow leaks. Blow through towers are pressurized during operation and can leak a lot of water with the fan running. Towers installed indoors pose a great risk for water damage from leaks.

Filters: A 100 ton cooling tower processes about 40 tons of air in an eight hour period retaining all the air borne debris in the tower water. Ideally, this debris would stay in suspension and be removed by the bleed but a good portion typically manages to settle in the tower basin and on heat exchange surfaces. Such debris tends to reduce the effectiveness of water treatment chemicals.

Full flow filtration is generally limited to devices that operate with low pressure drop and those that have an ability to purge debris while in operation. Strainers that swing into the flow stream while others backwash and centrifugal separators can be used full flow.

The centrifugal separator directs the water flow tangentially into a cylinder causing the water to rotate. Debris particles that are heavier than water migrate to the outer wall area and slither down into a collection bowl that is periodically blown down. Very little water is lost. Typical performance is 97.8% of solids with specific gravity exceeding 1.2 down to 45 microns.

Side stream filtration is more common in cooling tower applications. The scheme uses a small dedicated pump to draw dirty water from the sump, direct it through a filtration device and sent it back to the basin. In addition to being smaller and less expensive, side stream filtration has the advantage of allowing the return water to be routed through a perforated PVC distribution pipework or- better yet- nozzle jets that agitate the water at the basin floor placing the debris in suspension and increasing its chances of being drawn into the filter suction piping.

The actual side stream filtration device can be a centrifugal separator, strainer device or sand bed filter.

The sand bed filter is especially effective in removing particulate matter- even light weight biological material killed by biocides. Filtration to ten microns is commonplace.

When the sand begins to get clogged, a pressure differential switch signals a backwash cycle that lasts about three minutes. The valves reposition to reverse the flow through the sand bed lifting it and carrying off debris to the drain. Side stream filtration devices applied to open cooling towers are typically sized to pass the entire system volume at least once each hour for good water clarity.

A second filter selection method can be used when the system volume isn't known- and on all evaporative condensers and fluid coolers- that relates the filter size to the amount of air ingested:

Filter GPM = Cooling Tower CFM x .0004

Use both methods when possible and use the largest as a starting point. Then, apply a judgment factor increasing the size to accommodate blowing dust or sand.

Sand bed filter sizing example: An industrial process cooling tower is certified to cool 3,000gpm from 115 DegF to 90 DegF at 72 DegF wet bulb temperature. Catalogued air flow is 260,000cfm. System volume 4,000 gallons.. The operator wants to hold 4 cycles of concentration and use a sand bed filter with city water backwash.

-How big should the filter be?
-How big should the make-up and drain lines be?
-How much water will the system consume?

Side stream filter flow:


Based on system volume = 4,000gal / 60mim = 67gpm
Based on airflow = .0004 x 260,000 = 104gpm
use 104gpm; Select 30" diameter filter catalogued for 98gpm.

The make-up line should be sized to flow a steady 100gpm at about 7ft.sec. for a standard mechanical float. For any type of float that fills periodically such as a pilot operated hydraulic or electric type utilizing a solenoid valve, the line size must be enlarged since it operates at a higher flow rate for only a portion of the time. If filter backwash water is tapped off the same line, size for an additional 98gpm. Just make it generously sized in order to fill the system quickly after draining.

The drain line should be sized to handle the larger of:

- 2x the blow down (assuming bleed occurs half the time) which is 25gpm x 2 = 50gpm. or,
- filter backwash flow = 98gpm. (assumes blow down is inactivated during backwash.)
- Ideally, the drain would be sized to handle both simultaneously... Blow down plus backwash. or, 50 + 98 = 148gpm. If the drain cannot be easily made to accommodate the filter backwash, consider a holding tank (@ 98gpm x 3min = 300gpm size) arranged to drain slowly.

Since backwash typically occurs approximately once in 24 hours and lasts for about three minutes, the example demonstrates the blow down volume far exceeds the back wash volume... (25gpm x 24 hrs x 60min/hr vs 98gpm x 3min; or, 36,000gal vs 294gal.)

One complaint water treatment personnel have with filters is they discharge system chemicals during backwash and therefore prefer backwash be accomplished with city water.

The city water backwash example is actually rare. It is far more common to employ system water for backwash. City water backwash is really only necessary when sumps have insufficient volume to keep from running out of water during the backwash cycle. From the filter supplier's side, city water pressure can be too high or too low causing the backwash cycle to be too vigorous or too lethargic. They feel more comfortable with the known pressure generated with their own pump delivering 'system' water for backwash.

Filter backwash doesn't diminish the residual chemical level appreciably when system volumes are large; However, closely coupled systems with 'small basin' cooling towers can suffer severe chemical depletion from backwash. There should be no reason why controls can't be provided and set to prevent backwash until it is appropriate to bleed and to feed the appropriate chemical dosage after bleeding.