Cooling Tower Calculations

The purpose of a cooling tower is to conserve water. The heat picked up in the heat exchanger is returned to the cooling tower where it is rejected to the atmosphere by evaporative and convective cooling. The water that is evaporated at the cooling tower is pure, that is, it doesn’t contain any of the dissolved minerals present in the makeup water. As the evaporation process continues, these mineral solids concentrate in the recirculating water. If left unchecked, the solids eventually concentrate to the point of saturation. Here the dissolved solids precipitate to form a mineral scale or sludge deposit in the system This normally occurs in areas of high heat transfer such as in a heat exchanger.

The cycles of concentration or concentration ratio is defined as the ratio between the impurity levels in the recirculating water to the same impurity level in the makeup water. Generally the chloride ratio, conductivity ratio, or magnesium ratio is taken as the indicator of cycles of concentration, since these impurities are relatively soluble as compared to calcium carbonate.

Cycles of concentration = ClB = MgB = CondB

                                          ClMU MgMU CondMU

The cycles of concentration are controlled by the deliberate bleeding of water from the system. The bleed off water discharges the concentrated solids in the cooling water to drain. The water lost by evaporation and bleed off is replaced by fresh makeup water. The relationship between the required bleed rate and the cycles of concentration is given by the expression:

Bleed, gpm = Evaporation, gpm

                       (Cycles – 1)

Makeup rate, gpm = Evaporation + Bleed

These relationships indicate that the higher the cycles of concentration, the lower the bleed rate and, therefore, the lower the makeup rate. Since the purpose of the cooling tower is to conserve water, it is desirable to operate at the highest concentration ratio, while, at the same time, staying below the solubility limits of the dissolved minerals in the makeup.

Determining Cycles of Concentration

Several “rules of thumb” have been developed to determine the optimum cycles of concentration. Unfortunately, because of the numerous variables involved in cooling water chemistry, there is no universally accepted method for determining the maximum cycles of concentration. The Langelier and Ryznar Indices are often cited as the best indicator of the scaling or corrosive tendency of the recirculating water. These indices use the total dissolved solids, temperature, calcium hardness, and total alkalinity of the cooling water to compute the pH of saturation or pHs. The pHs is the theoretical pH at which calcium carbonate is in equilibrium with the calcium hardness and total alkalinity. The actual water pH, which we’ll indicate as pHa, and the pHs are used to calculate the index numbers according to the following relationships.

Langelier Index = pHa – pHs

Ryznar Index = 2pHs – pHa

In the case of the Langelier Index, positive index numbers indicate a scaling condition and negative numbers a corrosive or non-scaling condition. The Ryznar index uses the same operating variables, but the index value is always positive. Ryznar indices less than 6 indicate that calcium carbonate is likely to precipitate from the water, and values greater than 6 suggest the water will dissolve calcium carbonate, i.e. the water is corrosive. In either case, the objective is to set the bleed off rate to limit the cycles of concentration such that the cooling water chemistry is maintained on the non-scaling side of the index. However, more recently, several cooling water treatment programs have been marketed that permit the operation of the cooling tower within the scaling range of the index.

In many cases the value and usefulness of the Langelier and Ryznar indices is overstated. According to James McCoy, author of The Chemical Treatment of Cooling Water, “the Langelier Saturation Index applies only to the equilibrium between CO2 and CaCO3. Neither the pH of saturation nor the values derived from it are significant in industrial cooling systems.” More over, the LSI and RSI are not accurate indicators of the corrosion potential in the system unless you are concerned with the deterioration of concrete pipe.

More recently, the Practical Scaling Index (PSI) has been advanced as more accurate and useful than the LSI or RSI values. With the PSI value, the same operating parameters of total dissolved solids, temperature, calcium hardness and total alkalinity are used to compute the pH of saturation. To determine the PSI value, however, a pH of equilibrium, pHeq, is calculated from the total alkalinity, TA, of the recirculating water according to the following equation.

pHeq = 1.465log(TA) + 4.54

In 1977, Kunz and others published a method for predicting cooling water pH using an empirical formula derived from 400 data points obtained from actual operating systems. His equation is as follows:

pH = 1.6log(TA) + 4.40

In reviewing these equations for pH calculation, Jack Matson, PhD from the University of Houston, says in his paper “Precise Prediction of Cooling Water pH,” that neither equation is very precise in determining pH from the total alkalinity because they do not take into consideration the partial pressure of the carbon dioxide in the atmosphere.

With regard to determining the PSI value, the calculated pH of equilibrium, pHeq, is used instead of the actual pHa in the Ryznar formula to determine the PSI index value as follows:

PSI = 2pHs – pHeq

From this we can conclude that the LSI, RSI and PSI indices are imprecise indicators of the scaling or corrosive tendencies of the cooling water. Nevertheless, they are one of the few tools available to the water chemist to determine the optimum concentration ratio in the system.

Other Limiting Factors

The LSI, RSI and PSI indices are useful predictors of the solubility of calcium carbonate. With waters high in calcium hardness and total alkalinity, this is the primary scale-forming impurity.

Scale-forming impurities other than calcium carbonate are known to cause problems in cooling water systems. Calcium sulfate, tricalcium phosphate, silica, suspended solids, and process contaminants often limit the maximum permissible cycles of concentration. Solubility charts and related equations are available to determine the maximum concentration ratio of these impurities. Some useful guidelines are as follows:

Calcium Carbonate Deposition


Without Treatment

With Treatment



0 to +2.5



4.0 to 4.6

Calcium Sulfate Deposition

Calcium sulfate is more soluble than calcium carbonate. However, waters high in sulfate pose significant scaling problems. Once formed, calcium sulfate (gypsum) is more difficult to remove than calcium carbonate.

As a rule of thumb, the product of the calcium concentration times the sulfate concentration should be maintained at or below 500,000 to prevent calcium sulfate deposition.

[Ca] x [SO4] = less than 500,000

Tricalcium Phosphate Deposition

Polyphosphate is used in cooling water treatment programs to control scale deposition and corrosion. Over time, polyphosphate reverts to form orthophosphate. Orthophosphate, in turn, reacts under the right temperature and pH conditions with calcium hardness to cause the precipitation of tricalcium phosphate. The pH of saturation of tricalcium phosphate can be estimated from the following equation.

pHs = [11.755 – log(CaH) – log(o-PO4) – 2log(T)]


Actual cooling water pH’s above the pH of saturation for tricalcium phosphate will cause phosphate precipitation in the absence of chemical treatment.

As a rule of thumb, follow these guidelines for phosphate and calcium hardness levels.



o-PO4, ppm

Calcium, ppm

< 110 F


5 to 10

750 to 800

110 to 129 F


5 to 10

650 to 700

130 to 149 F


5 to 10

550 to 600

Silica Deposition 

Silica reacts with magnesium to form adherent scale deposits in cooling water systems. Like other scales, silica solubility is influenced by temperature and pH. The solubility increases with increasing pH and decreasing temperature. As a general rule, maintain silica levels below 150 ppm in the recirculating water to guard against this deposit.

Chemical Treatment

Chemical additives such as organophosphonates (HEDP and AMP) have been shown to increase the solubility of the common cooling water scales at low, threshold dosages. The following chart summarizes the impact these additives have on the solubility of calcium carbonate, tricalcium phosphate and calcium sulfate.

Phosphate versus Scale Solubility

Dosage, ppm

LSI Value

RSI Value

Calcium Phosphate

Calcium Sulfate








































Determining the Water Balance

Open cooling water systems must maintain a proper balance between evaporation, bleed, windage and makeup to control the cycles of concentration within the desired range. Once the desired cycles of concentration have been determined, the mass flow of water in and out of the system can be calculated from a few additional operating parameters.


Cooling Range is a measure of the difference between the cold water in the tower basin and the warmer cooling water return. This temperature differential is normally between 10 and 20 degrees F.

Approach is the difference between the cold cooling water temperature and the wet bulb temperature of the air. A cooling tower cannot cool water below the wet bulb temperature. Normally, the approach is with 7 to 10 degrees of the wet bulb temperature.

Recirculation Rates:

Obtain the rated capacity of the recirculation pumps in the cooling system. For comfort cooling systems, using centrifugal or absorption chillers, the recirculation rate can be estimated based on the refrigeration capacity of the chiller.

Centrifugal Machines require 3 gpm per rated ton.

Absorption Machines require 4 gpm per rated ton.

Water Quality

Obtain water analyses for the recirculating cooling water and makeup.

Evaporation Rate Calculations

The amount of water evaporated from the cooling tower is a function of the recirculation rate and cooling range. Cooling towers cool water by evaporating a small percentage of the recirculating water flow. In general, 0.1% of the recirculating water is evaporated for every 1 degree of temperature drop across the tower. It takes approximately 1000 Btu’s to evaporate 1 pound of water. If 1 pound of water is evaporated from 1000 pounds of water (0.1%), 1000 Btu’s are removed from 999 pounds of water, or 1 Btu per pound. 1 Btu removed from 1 pound of water lowers the temperature by 1 oF. (A Btu is the amount of heat required to raise (or lower) the temperature of 1 pound of water by 1 oF.) Therefore:

Evaporation = 0.001 X R x dT X f


R = recirculation rate, gpm

dT = temperature range across tower or “delta T”

f = evaporative cooling factor

Not all of the temperature drop across the cooling tower is a result of evaporative cooling. Some of the heat loss occurs by convective cooling, whereby heat is transferred by direct contact between the cooler air and the warmer water. On average, convective cooling accounts for 25% of the heat loss in a cooling tower. This will vary seasonally, however. In the Midwest, for example, 15% convective cooling occurs in the summer months, increasing to 35% in the winter. For most areas of the US, an average of 25% is reasonable. The evaporation rate must be adjusted for the amount of convective cooling taking place. The “f” factor accomplishes this. Use 0.75 for most estimates.

For mechanical refrigeration machines, the evaporation rate can be estimated from the refrigeration capacity and the heat rejection factor of the machine using the following formulat

Evaporation = Tons X Hfr X 24



Evaporation = evaporation rate, gpm

Tons = refrigeration capacity, tons

Hfr = Heat rejection factor of the machine

Compression machines = 1.25

Absorption machines = 2.6

Hfg = Heat of vaporization of water, 1050 Btu/pound

Cycles of Concentrati

As water evaporates from the cooling tower, the mineral impurities in the makeup are concentrated in the recirculating water. The cycles of concentration or concentration ratio is determined by calculating the ratio between an impurity in the cooling water and the same impurity in the makeup. Normally, chlorides are used for this purpose since they are very soluble and unlikely to precipitate to form scale or sludge in the system. If sodium hypochlorite (bleach) is used, however, the chloride level in the tower will be artificially high. Other impurities such as magnesium hardness or conductivity may be used to check the concentration ratio.

Cycles = ClB = MgHB = CondB

                ClMU MgHMU CondMU

Cycles of concentration can also be determined by calculating the ratio between the makeup and bleed off rates. A water meter installed on the makeup and bleed off lines is helpful in determining the average gpm for each parameter.

Cycles = Makeup rate

               Bleed rate

Bleed Rate

Cycles of concentration are controlled by discharging a percentage of the recirculating water to drain and replacing this concentrated cooling water with fresh makeup. Increasing the bleed rate decreases the cycles of concentration. Decreasing the bleed increases cycles. The bleed rate required to maintain a desired cycle of concentration is determined by the following equation.

Bleed = Evaporation Rate

                (Cycles – 1)

The bleed rate can also be determined by measuring the makeup rate and dividing by the cycles of concentration.

Bleed = Makeup rate



The air passing through a cooling tower frequently blows small droplets of water out of the tower. This mist is called windage or drift. In some cooling systems windage can account for a significant amount of water loss from the system. These losses are reported as a percentage of the total recirculation rate. Although difficult to measure, the impact of windage can be estimated from the following chart.

Windage Losses


Windage Loss, %

Spray Pond

1.0 to 5.0

Atmospheric Tower

0.3 to 1.0

Mechanical Draft Tower

0.1 to 0.3

Evaporative Condenser

0.0 to 0.1

Windage = %Windage X Recirculation rate

Water lost by windage has the same effect on cycles of concentration as does bleed off. Water losses in cooling towers with excessive windage must be included in the total bleed off rate when calculating the cycles of concentration. Increased windage lowers cycles, whereas decreased windage increases cycles. For modern mechanical draft cooling towers with drift eliminators, the impact of windage on the overall bleed rate is minimal and often ignored in cooling tower calculations.

Theoretically, if the controlled bleed is shut off, the maximum cycles of concentration achievable in a cooling tower is limited by the percent evaporation and percent windage

Cycles max = %Evaporation + %Windage



Fresh makeup water is added to the cooling system to replace water lost by evaporation, bleed off, windage, and leaks. Unless systems leaks are significant, they are generally ignored altogether or included in the windage loss estimates. The makeup rate is then determined by adding all of the water losses from the system.

Makeup = Evaporation + Bleed + Windage

The makeup rate can also be estimated by rearranging the bleed off and cycles of concentration equations identified previously. Some useful formulas are:

Makeup = Evaporation + Makeup


Makeup = Evaporation + Evaporation 

                                           (Cycles – 1)                   

Makeup = Cycles X Evaporation

                     (Cycles – 1)



Boiler Layup Procedures

Boilers must be stored under carefully controlled conditions during non-operating periods to avoid corrosion damage that can occur in the absence of proper lay-up procedures. Improper lay-up and storage will result in rust formation, pitting-type corrosion and general deterioration of boiler metal surfaces. This damage can occur on both the waterside and fireside of the boiler.

If the waterside of the boiler is exposed to the atmosphere, corrosion will occur at the liquid-to-air interface. Corrosion damage is also possible in the preboiler and afterboiler sections. Once formed, the by-products of corrosion can then be transported to the operating boiler when the system is returned to service. These corrosion products may deposit on critical heat transfer surfaces, increasing the potential for localized corrosion or overheating during system operation.

The two major factors which determine the corrosion rate on boiler metal are moisture and dissolved oxygen. Under completely dry conditions, the corrosion of steel is negligible. In a moist or wet environment, however, the amount of dissolved oxygen in the water determines the severity of the corrosion. Conditions that increase the oxygen concentration in the water, or allow the continued addition of oxygen, will increase the corrosion rate.

The fireside of the boiler is also subject to corrosion damage. Like the waterside, corrosion damage to the fireside will occur if the metal surfaces are wet and exposed to oxygen. Sulfurous and sulfuric acid residues, which are by-products of the condensation of acidic flue gases, will also promote corrosion attack.

For these reasons, proper boiler storage procedures must be followed to protect and prolong the useful life of the steam generating equipment.

Removing the Boiler from Service

Pre-Shutdown Procedures

Preparation for boiler shutdown should begin 10 days prior to the scheduled shutdown date. The following procedures will help remove accumulated boiler sludge prior to draining the boiler.

  • Increase the blowdown rate to maintain the boiler conductivity at the low end of the normal control range.

  • Keep the steaming rate as high as possible during this period to maintain optimum boiler water circulation

  • Maintain the boiler water chemicals at the normal operating levels

Boiler Shutdown Procedures

The boiler should be brought down in rating slowly. Raise the water level as high in the glass gauge as is consistent with safe operating practice while still unloading some steam to the line. After the pressure has dropped, do the following four operations:

  • Shut off the continuous blowdown system

  • Blow down all manual valves including the sidewall and waterwall headers

  • Open the steam drum vent, close the steam non-return valve and the head stop valve.

  • Blow down all manual valves once for every 25 psig decrease in boiler pressure

Draining the Boiler

The last blowdown should be made at 25 psig. Allow the boiler to cool to approximately 120 o F. This will allow for uniform cooling of the boiler tubes and drum. Drain the boiler as follows:

  • Start the draining process by opening all manual blowdown valves.

  • When the water level is below the steam drum manhole, remove the manhole covers and start washing the boiler with a fire hose at high pressure.

  • Wash down the boiler for at least three hours

  • Allow the boiler to drain completely. Revove the manhole covers from the mud drum and all handhole plates from the sidewall or waterwall headers. Wash the headers and the mud drum until all loose material has been removed.

Chemical Cleaning

The waterside surfaces of the boiler and superheater must be free of any deposits, sludge, oils, corrosion by-products, or other debris. Any deposits that exist on the boiler surfaces will promote under-deposit corrosion. This type of corrosion is characterized by pits which form under loose, porous deposits on the metal surface.

A typical chemical cleaning procedure involves the use of strong cleaners such as inhibited hydrochloric acid. The mineral acid dissolves inorganic scales and corrosion by-products. The acid cleaner is then drained, the boiler flushed with water, and then passivated with an alkaline passivating solution.

The decision to chemically clean the boiler prior to storage is a subjective one. If the visible portions of the boiler internals are clean and free of foreign material, then the chemical cleaning step can be eliminated. Boilers that have significant scale deposits, however, should be chemically cleaned prior to storage.

Boiler Storage Options

Two basic boiler lay-up procedures are in use:

  • Dry lay-up

  • Wet lay-up

These two basic methods have several variations, including nitrogen or steam blanketing, which helps insure the complete exclusion of air from the boiler during lay-up and storage.

Dry lay-up is recommended for boilers that will be out of service for 1 month or more. Here the boiler is drained, dried and stored in a moisture-free environment. This includes installing trays of desiccant, such as silica gel, in the boiler drums to maintain a constant low humidity atmosphere. The boiler is then closed to minimize oxygen ingress. If the boiler is to be stored open, the boiler and superheater are thoroughly dried and a positive dry air flow is maintained from bottom to top during the storage period. Superheaters can be stored under a 5 psig nitrogen blanket for added protection.

Wet lay-up is recommended for boilers that must be maintained in an emergency standby mode. This procedure involves filling the boiler, feedwater heaters, and deaerators with demineralized water, treating the water with a chemical oxygen scavenger, and adjusting the pH between 10 and 11. For boilers without superheaters or for boilers with drainable superheaters, sodium sulfite and caustic soda are generally used for this purpose. For high pressure boilers, or boilers with non-drainable superheaters, volatile chemicals must be used such as hydrazine and ammonia. Alternatively, neutralizing amines such as morpholine, cyclohexylamine, or diethylaminoethanol may be substituted for the ammonia.

Generally, boilers should not be stored in the wet lay-up mode for more than 6 months. For extended periods beyond 6 months, the dry lay-up procedure provides better corrosion protection.

Nitrogen blanketing is used to provide an inert, corrosion-free environment. Nitrogen is an odorless, colorless gas with an extremely low dew point. It is used routinely to purge oxygen from enclosed vessels. Corrosion can not occur in an inert nitrogen environment. Under wet lay-up conditions, nitrogen may be connected to the steam vent to provide a low pressure nitrogen blanket to prevent oxygen ingress. Alternatively, nitrogen may be used during dry lay-up conditions to provide a positive nitrogen pressure (5 psig) in the closed boiler vessel to prevent oxygen and moisture intrusion. It may also be used to inert superheaters, and provide a nitrogen blanket in deaerators and feedwater heaters.

Steam blanketing is used to store boilers and auxiliary equipment under a positive pressure to prevent the ingress of dissolved oxygen. The water temperature keeps the metal surfaces above the dew point, which helps protect the fireside from corrosion. Deaerators and feedwater heaters may also be stored under a steam blanket. Because of the high energy costs associated with steam blanketing, however, it is not recommended for storage periods of more than 6 months.

Dry Lay-up Method

The objective of dry lay-up is to maintain the boiler metal surfaces in a moisture-free condition. Corrosion cannot occur in a dry, oxygen –free environment.

Storing Under a Nitrogen Blanket

  • Drain the boiler as described previously

  • Clean the boiler, if necessary

  • Completely dry all circuits of the boiler with a positive air flow from the bottom to the top

  • If nitrogen blanketing is to be used for dry storage, close up the boiler, purge all of the air from the boiler with nitrogen, and then store under a 5 psig nitrogen pressure.

Storing Dry with Chemical Desiccant

  • Dry the boiler out by blowing hot dry air through the boiler, or apply low auxiliary heat to dry the metal surfaces

  • Place bags of desiccant in wooden trays in the steam and mud drums

Amount of Desiccant Required


Pounds per 30 ft3

Telltale silica gel




Note: Silica gel is easier to use than quicklime and can be dried and re-used.

  • Close all manholes and blank or close all connections on the boiler as completely as possible to prevent the intrusion of humid air

  • Inspect the waterside of the boiler every 2 to 3 months for active corrosion

  • Inspect the condition of the desiccant and replace, if necessary

Wet Lay-up Method

Wet lay-up is the preferred method for storing boilers that must be maintained in an emergency standby mode. However, boilers should not be stored for periods longer than 6 months in the wet lay-up condition. During this period the boiler should be fired twice per month to bring the boiler water to 160 o F. This circulates the boiler water and improves corrosion protection.

Two methods for chemical treatment of the boiler water during wet lay-up are recognized. For boilers with drainable superheaters, an inorganic treatment program is used consisting of sodium sulfite for oxygen scavenging and sodium hydroxide (caustic soda) for pH adjustment. For boilers with non-drainable superheaters, a volatile treatment program is required to avoid the formation of inorganic deposits in the superheater at the time of startup. In this case, hydrazine is used for oxygen scavenging and ammonia or a volatile neutralizing amine, such as morpholine or DEAE, for pH adjustment.

Alternatively, the non-drainable superheater must be blanked off to prevent the entrance of treated water and then blanketed with nitrogen.

Boiler with Non-drainable Superheaters

  • Drain the boiler as described previously

  • Clean the boiler, if necessary

  • Backfill the boiler through the superheater with demineralized, deaerated and chemically treated water. Disconnect the high water alarms and fill the boiler completely to the top. The economizer and feedwater heaters may also be filled and treated in a similar manner.

Chemical Dosage



Hydrazine (35%

4.8 lbs per 1000 gallons


0.10 lbs per 1000 gallons

(or) DEAE

0.5 lbs per 1000 gallons

  • After the boiler and superheater are filled with treated water, close or blank all connections. Install a 55 gallon tank with a tight-fitting cover and sight glass at a location above the steam drum. Connect the tank to the vent line on the boiler to create a hydrostatic head. A net loss or gain of water from the boiler will be indicated by the water level in the drum.

  • As an alternative to installing a 55 gallon expansion tank, connect low pressure (5 psig) nitrogen to the boiler vent to pressurize the system.

Boilers with Drainable Superheaters

  • Drain the boiler as indicated above.

  • Clean the boiler, if necessary.

  • Backfill the boiler through the superheater with demineralized, deaerated feedwater. Since the superheater is drainable, the water may be treated with sodium sulfite and caustic soda.

Chemical Dosages



Sodium sulfite

1.5 lbs/1000 gallons

Sodium hydroxide

3.0 lbs/1000 gallons

  • Fill the boiler all the way to the top and cap with a 55 gallon expansion tank, or pressurize with 5 psig nitrogen.

Recirculating the Boiler Water

During the wet storage period the boiler should be fired twice per month to bring the water temperature to 160 oF and then allowed to cool to ambient temperature. This insures good circulation of the treatment chemicals.

An alternate method is to install a small recirculating pump on a bottom connection to the boiler mud drum such as a T-connection off the bottom blowdown line. The boiler water can be circulated from the mud drum back to the steam drum at a rate sufficient to give one turnover of boiler water every 8 hours. The pump can also be operated to mix the boiler lay-up chemicals or circulate the lay-up solution through the system.

Testing the Boiler Water

The boiler lay-up water should be tested once per week. This will confirm that adequate corrosion protection is being maintained. The control limits for these chemical levels are indicated below:

Chemical Control Limits

Volatile Program for Non-drainable Superheaters


Control Range


200 to 400 ppm


pH 10 to 11

Inorganic Program for Drainable Superheaters


Control Range

Sodium sulfite

100 to 200 ppm

Caustic soda

OH alk = 400 ppm

If the chemical residuals drop below the indicated minimums, drain some water from the boiler, add additional lay-up chemical, fire the boiler to 160 oF to circulate and retest.

Startup After Wet Lay-up

The following guidelines should be followed when returning the boiler to service after wet lay-up.

  • Purge the nitrogen from the system, or disconnect the hydrostatic tank from the boiler vent connection.

  • Drain the boiler water to normal operating levels. Drain the deaerator and feedwater heater, if stored in a similar manner.

  • Reset the high water alarm.

  • Open all valves and disconnect the recirculating pump, if used.

  • The boiler can be fired with the lay-up chemicals. If a non-drainable superheater is stored wet, waste steam to the atmosphere to purge the treated water from the superheater. For drainable superheaters, completely drain it prior to startup and blow with dry compressed air prior to firing the unit.

  • Re-start the normal chemical treatment program and water quality control practice.

Startup After Dry Lay-up

  • If stored under a nitrogen blanket, purge all nitrogen from the system. Do not enter the boiler. Nitrogen will not support life, and will cause death by suffocation.

  • If stored with trays of desiccant, remove them from the steam and mud drums.

  • Refill the boiler and return to service in a normal manner.

Fireside Lay-up Procedure

Provisions should also be made for protecting the boiler fireside from corrosion during the boiler lay-up period. To minimize the corrosion caused by acidic flue gases, the fireside of the boiler, air heater, and ID fan should be washed down with a 5% solution of sodium carbonate. Drain all wash water from the boiler. Dry the fireside and keep it dry by continuous circulation of dry air through the furnace. Seal the furnace as completely as possible to minimize the entry of humid air.

The fireside should be inspected once per month. Any signs of active corrosion should be noted and corrective action taken to eliminate the cause of the problem.

Safety Considerations

As with all chemical products, the boiler lay-up chemicals should be stored and handled according to recommended procedures as stated in the Material Safety Data Sheets.

Nitrogen is an inert gas that does not support life. If nitrogen is used to inert boilers and auxiliary equipment, do not enter the equipment until all nitrogen is purged from the system and tests show that sufficient air is present to support life.

Hydrazine is toxic. It has been classified as a suspected carcinogen. Skin or eye contact can cause permanent injury or dermal sensitization. Inhalation of hydrazine can be irritating and can cause health problems. Do not enter equipment until the concentration of hydrazine in the air is below 1 mg/l.

Wear safety equipment such as rubber gloves, eye goggles, aprons and boots when handling lay-up chemicals. Follow all other precautions as stated in the Material Safety Data Sheets.

A Summary of Boiler Lay-up Procedures

Boilers must be stored under carefully control conditions to prevent rust formation, pitting-type corrosion and general deterioration of boiler metal. Two methods for boiler lay-up are recognized as effective: (1) dry lay-up and (2) wet lay-up.

Dry lay-up is recommended for boilers that will be out of service for 6 months or longer. With this procedure the boiler is drained, dried and inerted with a nitrogen blanket. As an alternative to nitrogen blanketing, trays of telltale silica gel are installed in the steam and mud drums, and the boiler is closed up to minimize the entry of humid air.

Wet lay-up is intended for boiler that must be stored under emergency standby conditions. Here the boiler is backfilled through the superheater. The boiler is filled completely to overflowing and then pressurized with a nitrogen blanket to prevent air ingress. As an alternative to nitrogen capping, the steam vent can be connected to a 55-gallon drum of treated water to provide a hydrostatic head. Or the boiler can be pressurized under a steam blanket. If the boiler has drainable superheaters, sodium sulfite and caustic soda are used to protect the system. If the superheaters are not drainable, an all volatile treatment program is required using hydrazine and ammonia (or alternatively a neutralizing amine) for corrosion protection.

These basic lay-up procedures can be adapted to meet individual plant requirements. Overall, however, the goal is to maintain an oxygen-free, non-corrosive environment to protect the boiler and auxiliary equipment while in short or long-term storage.

Chemical Treatment Requirements for Condensate Systems

Corrosion damage in steam-condensing equipment and systems that collect, return and store condensate for reuse as boiler feedwater is a continuing problem at industrial plants, government installations, institutions, hotel, office and commercial buildings. Power engineers long have recognized that steam condensate is corrosive to iron and steel piping systems, and will attack many non-ferrous alloys as well.

Material costs and labor for replacement of corroded piping or repair of damaged equipment are appreciable and growing. Where condensate lines are buried in concrete floors or installed in walls of permanent structures, replacement costs will be far greater than those of the original construction. Failure of corroded condensate piping not only damages building and equipment, but also may cause finished product losses. Interruption of steam service to repair corrosion damage causes inconvenience, lost time and production losses; shutdown of major equipment may be required during repairs.

Corrosion products carried back to boilers in condensate returns may deposit on steam generating surfaces causing over-heating failures or localized corrosion of boiler tubes. Buildup of metal oxides in condensate systems fouls traps, strainers and piping, reducing flow capacity and increasing pumping costs.

The growing demands for steam to be used in absorption air conditioning, process and humidification systems have intensified corrosion problems previously related to heating seasons. Repair of damage caused by condensate system corrosion no longer can be handled as a summer maintenance project. Instead, engineers responsible for steam plant operation and building maintenance now must recognize the need for year around corrosion control programs and improved procedures for monitoring corrosion in critical steam/condensate systems.

Origins of Corrosion Agents – Carbon dioxide and Oxygen

Carbon dioxide (CO2) and oxygen are the main corrosive agents in steam. Carbon dioxide dissolves in condensate to form corrosive carbonic acid. This, in turn, lowers the condensate pH and makes it corrosive to steel and other metals.

H2O + CO2 —– H2CO3

Dissolved oxygen also is corrosive to system metals, and when present along with CO2 in the system, greatly intensifies the corrosion activity of carbonic acid. Mechanical factors of the system design and operation also contribute to condensate system problems. This must be considered in the overall corrosion control program.

Much of the oxygen in steam enters the boiler in poorly deaerated feedwater. Deaerators may be operated with inadequate steam supply, at temperatures well below those needed for effective oxygen removal, or experience mechanical problems and malfunctions. This includes faulty venting of non-condensable gases. Oxygen enters the deaerator both in makeup water and in returning condensate. Gravity and vacuum return systems may draw in air at receiver vents, radiator relief valves, at vacuum pump packing glands, around valve stems, and similar in-leakage points. Even deaerators and deaerating heaters may create partial vacuums that draw in air. This occurs in the event of steam supply failure or a drop in steam pressure under light load conditions. Process heat coils, dryers and sterilizers equipped with air relief vents will draw in air when the heating cycle is interrupted and partial vacuums are created by condensation.

The internal use of chemical oxygen scavengers (sodium sulfite and hydrazine) in boiler water should neutralize most of the dissolved oxygen that is not removed by feedwater deaeration. But the fact remains that some oxygen still is released from the boiler in steam and dissolved in condensate to cause corrosion. More oxygen is added to condensate by the various in-leakage sources mentioned earlier.

The major source of carbon dioxide in steam is the natural carbonate-bicarbonate alkalinity of boiler makeup water. Although small amounts of dissolved carbon dioxide gas are naturally present in many water supplies, the bicarbonate alkalinity is a far greater source of CO2. Alkaline salts such as calcium and magnesium bicarbonate in feedwater readily decompose at boiler temperatures, liberating carbon dioxide into the steam.

HCO3 + heat —– CO2 + OH

The addition of soda ash (Na2CO3) to feedwater also increases the CO2 potential of the steam. About 80% of the soda ash dissociates at boiler pressures of 100 to 200 psig to release CO2 and hydroxide alkalinity. The decomposition reactions of carbonate alkalinity and sodium carbonate (soda ash) are as follows:

CO3-2 + H2O —– CO2 + 2(OH)

Na2CO3 + H2O —– CO2 + 2NaOH

The carbon dioxide potential of the steam is increased by sodium zeolite softening of the makeup water. This process removes calcium and magnesium hardness from the raw water, converting calcium and magnesium bicarbonates into sodium bicarbonate. This sodium salt completely decomposes in the boiler to release CO2. The CO2 released from 100 ppm carbonate alkalinity is about 35 ppm as CO2. The total CO2 produced by 100 ppm sodium bicarbonate is much higher – about 79 ppm as CO2.

Reducing the Carbon Dioxide Potential of Steam

All pretreatment methods that reduce the carbonate and bicarbonate alkalinity of the makeup water will reduce the CO2 release from the boiler feedwater during steam generation. Pretreatment processes such as lime or lime-soda ash softening (cold or hot), reverse osmosis, demineralization, or split stream ion exchange all reduce makeup water alkalinity and simplify corrosion control in return condensate systems. The sodium zeolite softening process does not reduce the carbonate – bicarbonate alkalinity of the makeup water. This tends to aggravate condensate system corrosion problems rather than relieve them.

Characteristic Forms of Corrosion – Nature and Identification

Carbonic acid corrosion usually shows up in the form of distinct channeling or “grooving” along the lower sides of condensate lines and extending up the pipe walls to the average condensate depth. Sometimes the groove will be almost as uniform as if it had been machined. It may cause uniform thinning of pipe walls when lines run full of condensate. Attack usually penetrates and shows up first in the thinnest, most highly stressed, parts of pipe fittings such as short nipples, threaded ends of joints, unions and elbows.

Corrosion products are usually absent at the attack location, but may appear downstream. The iron is dissolved, forming ferrous bicarbonate, which stays in solution initially then decomposes to release CO2 and reacts with dissolved oxygen to form various types of iron oxides.

Iron reacts with carbonic acid to form ferrous bicarbonate and hydrogen gas.

Fe + 2H2CO3 —– Fe(HCO3)2 + H2

(Ferrous bicarbonate)

Ferrous bicarbonate decomposes to form ferrous oxide and releases carbon dioxide gas.

Fe(HCO3)2 —– FeO + 2CO2 + H2O

(Ferrous oxide)

Ferrous bicarbonate reacts with oxygen to form ferric oxide and releases more carbon dioxide gas.

4(Fe(HCO3)2 + O2 —– 2Fe2O3 + 8CO2 + 4H2O

(Ferric oxide)

Other compounds found in condensate system deposits include magnetic iron oxide (magnetite) and ferrous carbonate (FeCO3). Note that decomposition of the iron bicarbonate and its reactions with oxygen both release CO2, which is then free to redissolve in the condensate and start the corrosion cycle again. The iron bicarbonate also raises the condensate pH by virtue of tying up carbonic acid. This often leads to a false sense of security when only pH tests are used for condensate monitoring. The condensate from active corrosion areas often show pH within the desired range because of the presence of dissolved iron bicarbonate. If the condensate had been tested for dissolved iron or had corrosion coupons been installed in the system, they would have shown serious corrosion in progress.

Dissolved oxygen attack appears as irregular, scattered pitting usually accompanied by oxide deposits or “tubercles” near the corrosion location. But if both oxygen and carbonic acid are in the condensate, the oxygen corrosion products may stay in solution until carried further downstream. The combined effects of oxygen and carbonic acid show up as irregular pitting joined together by carbonic acid grooves to form a continuous but irregular flow pattern. This is called combined pitting and grooving or sometimes “worming.” Since dissolved oxygen increases the reaction rate between carbonic acid and iron, the combined attack proceeds more rapidly than that caused by either corroding substance working alone.

Other corrosion types encountered in return line systems include impingement or wire drawing. Here wet steam or corrosive condensate contacts metal surfaces at high velocity, causing erosive loss of protective oxides and inhibitor films. This action continually exposes fresh metal to attack. This can occur at elbows and pipe bends carrying high velocity wet steam (turbine exhaust lines, for example) or where steam traps discharge into larger return lines.

Steam collapse, causing cavitation effects, also occurs where high pressure traps discharge into low pressure, colder return lines. The resulting shock forces strip away oxide layers and inhibitor films, causing severe localized attack.

Most probable locations for concentrated corrosion attack are inside space heater tubes, in process heating coils, dryer rolls, etc. Where steam condenses continuously the attack shows up as thinning in the constantly wetted sections. Just downstream from condensing units the attack is likely to be severe in condensate lines and fittings. Poorly drained condensate lines that “pool” condensate or run continuously at a steady liquid level will develop “grooving” problems, particularly in areas subject to oxygen in-leakage. Other trouble spots are found near remote ends of large low-pressure heating systems where CO2 tends to build up in the remaining steam.

Influence of Flow Rates, Concentration and Solution Rate

Since carbonic acid attack is proportional to the pounds of acid contacting the metal surface in a given period, both the amount of carbon dioxide in solution and the condensate flow rate control the corrosion rate. In oxygen attack, the corrosion rate mainly is a function of oxygen concentration.

Not all of the carbon dioxide in steam goes into solution immediately when the steam condenses. At the entrance to a steam condensing system, the amount of carbon dioxide may only be a small fraction of the total vapor. As the steam condenses, however, the percentage of carbon dioxide in the remaining vapor increases and more will go into solution. This explains why severe carbonic acid attack often occurs near remote ends of heating systems or in isolated condensing units that operate intermittently even though carbon dioxide content of the steam leaving the boiler is low. When steam condensation begins, the first condensate contains little dissolved carbon dioxide, since the initial partial pressure of CO2 in the steam is low. The remaining carbon dioxide accumulates in the vapor phase. With continued steam condensation, the equilibrium content of the accumulated vapor may be several hundred times the concentration of carbon dioxide in the steam leaving the boiler or dissolved in discharged condensate.

When incoming steam is shut off, pressure drops as the unit or system cools. Accumulated carbon dioxide dissolves into the residual condensate causing the pH to drop sharply and accelerating corrosion. Proper drainage from intermittently operated condensing units is vital. Continuous venting with positive pressures in such units also helps reduce damaging accumulations of carbonic acid.

Sampling and Testing Condensate Samples

Condensate samples should be taken from systems in continuous use, and as close as possible to the point of condensation. For example, a sampling connection immediately downstream from the trap serving a large space heater or process heating coil is an excellent location. It is important to sample freshly-form condensate rather than the condensate mixture accumulated in receiver tanks. The condensate taken from close to the condensing unit will be representative of the most serious corrosion potential to be expected and should have the highest CO2 content and lowest pH. If we wait until this aggressive condensate has returned to the receiver, the condensate tests may be misleading. This is because of corrosion reactions that occur in the return system, the formation of iron bicarbonate and other corrosion products, and the resulting alteration of condensate pH plus carbonic acid and iron concentrations.

In addition to sampling and testing condensate, methods are needed for monitoring system corrosion rates. It is desirable to employ test specimens or sections that will furnish visual indications as to type, location and severity of corrosion attack. Corrosion test specimens also must be properly installed in steadily flowing condensate lines so as to be continuously immersed in condensate. In some cases a drop leg is needed to keep the specimen immersed. A bypass line with proper shutoff valves and unions may also be needed so specimens can be inserted and removed without having to close down the system.

Corrosion coupons are widely used for measuring the corrosion rate in condensate systems. While acceptable for corrosion weight loss measurements and indications of penetration rates, they give little or no indication as to the nature of attack in the system other than pitting. The weighed coupons are inserted into the condensate system by means of a pipe plug drilled to accommodate a short section of phenolic rod. This rod acts as an insulating mounting for the coupon. The piping arrangement for the corrosion coupon must be arranged to provide continuous immersion during the scheduled period of test.

Regular sampling and testing of condensate for dissolved metals such as iron and copper is an effective means for monitoring the corrosion activity in the system. For these tests, samples collected at condensate receivers are acceptable, since they should show the maximum amount of iron and copper corrosion products returning to the boilers. They also give an indication as to how much return line plugging can be expected from iron oxide and other corrosion product deposits.

Corrosion Inhibitors

Corrosion inhibitors added to the boiler or steam header will aid in reducing corrosion caused by carbonic acid and dissolved oxygen in the condensate. Two general classes of amines are used for condensate system protection – neutralizing amines and filming amines.

Neutralizing amines volatilize with the steam and condense with the steam condensate. Here they neutralize carbonic acid and raise condensate pH.

Commonly used neutralizing amines include morpholine, cyclohexylamine and diethylaminoethanol (DEAE). Other, chemically similar, amine compounds are also used. In some cases, mixtures or blends of neutralizing amines have advantages when dealing with large steam/condensate systems using both high pressure and reduced pressure steam, or those including turbines and low pressure steam heating loads. Ammonia is used as a neutralizing agent for carbonic acid also, but generally is restricted to large all-iron systems because of its corrosivity to copper alloys when oxygen is present.

The dosages of neutralizing amines are based on the CO2 content of the steam. Since these inhibitors function only by neutralizing carbonic acid, they must be added in proportion to the average CO2 level in the steam as estimated from the carbonate-bicarbonate alkalinity of the feedwater. If concentrated (100% active) amines were practical, it would take about 2 ppm of amine to neutralize one ppm of dissolved CO2 in the condensate. For practical reasons, amine solutions of 20 to 40% are used. They are easier to handle and are non-flammable at these concentrations. With the 20% to 40% strength amines, it takes from 5 to 10 ppm to neutralize each ppm of CO2 in the condensate. In addition to the amounts needed to neutralize dissolved CO2, small extra dosages are added to raise the pH of the treated condensate to the usual operating range of pH 7.5 to 8.5.

After the initial starting dosage is determined, further adjustments in amine dosages are made to accomplish the treatment objectives.

  • pH adjustment to get the pH up to at least 7.0 at the discharges of all steam-condensing units and within the desired range elsewhere through the return system

  • To keep the dissolved iron and copper at very low levels, i.e. well below 0.1 ppm

  • To maintain low corrosion rates on corrosion coupons installed in steam condensing and return systems.

The criteria by which specific amines are selected to accomplish these objectives are determined by several factors. Among these is the CO2 potential of the feedwater, the presence or absence of dissolved oxygen, the pressure and temperature relationships in steam-using and return systems, and the overall economics.

Since the only function of neutralizing amines is to neutralize carbonic acid in condensate, they have no utility in preventing oxygen corrosion. Also, they must be used in stoichiometric amounts; that is, the dosage goes up in proportion to the carbonic acid concentration. Consequently, neutralizing amines would not be an acceptable choice when the CO2 potential of the feedwater is very high and dissolved oxygen is a problem. Simple economics may rule out neutralizing amine use when the feedwater alkalinity is high.

Distribution ratios characteristics of neutralizing amines also introduce an important consideration. The distribution ratio determines the concentration of amine required in the steam to get one ppm dissolved in the condensate. This characteristic differs widely for morpholine, cyclohexylamine and diethylaminoethanol (DEAE). For cyclohexylamine about 4 to 5 ppm are needed in steam to get one ppm dissolved in the condensate. DEAE is more efficient, having a steam/condensate distribution ratio of about 2. Morpholine has the lowest distribution ratio of about 0.4. This indicates the highest tendency to dissolve in first-formed condensate.

Cyclohexylamine is very effective in protecting low pressure and reduced pressure steam systems and their condensate return lines. It reaches remote ends of the system, which morpholine cannot do because of its low distribution ratio. DEAE is intermediate in distribution ratio and solubility between morpholine and cyclohexylamine, indicating utility in protection of mixed pressure systems with some high pressure steam use along with reduced pressure heating steam demand.

In summary, morpholine is the most effective in condensate systems where pressures average at least 50 psig or above. Cyclohexylamine is the material of choice for systems operating at pressures below 50 psig, especially those including reduced pressure steam heating systems. DEAE falls between these two extremes in characteristics, so should be useful for mixed-pressure systems with both high pressure and low pressure steam and return systems.

Filming amines offer an alternative approach to condensate system protection. A filming inhibitor does not neutralize carbon dioxide; instead it establishes a non-wettable, absorbed film on the metal surface. This wax-like protection provides a barrier between the metal and the condensate. Dosages required are low at 1 to 3 ppm and are substantially independent of the concentration of carbon dioxide and dissolved oxygen. In contrast to neutralizing amines, filming amines prevent attack by both carbon dioxide and dissolved oxygen.

The dosages of filming amine products are adjusted to produce an amine residual of 0.1 to 0.5 ppm in the returned condensate. (They have little effect on condensate pH, so they are not regulated on the basis of pH control. As with neutralizing amines, the objective is the same, to keep the corrosion rates low as measured by coupons testers and to keep iron and copper concentrations as low as possible.

Field experience and laboratory studies show that boiler water carryover destroys octadecylamine protective films. The sulfate ion is particularly critical. As little as 8 ppm sulfate is sufficient to nullify the value of the inhibitor. Oil contamination also reduces the effectiveness of octadecylamine performance. Therefore, precautions to minimize carryover and keep oil out of the steam system are essential to accomplishing effective corrosion control with filming amines.

A useful characteristic of the filming inhibitors is their ability to displace corrosion deposits and fouling layers of all types. When starting filming amine treatment in an old, corroded system, initial concentrations should be well below the desired final level. Raise the dosages very gradually to help minimize plugging of traps, strainers, etc caused by too-rapid removal of corrosion products and other foulants.

The removal of fouling layers naturally improves heat transfer efficiency at condensing surfaces. In addition to the elimination of fouling factors, octadecylamine and other filming inhibitors improve heat transfer by promoting drop wise condensation of steam and better runoff of condensate. In many large paper mills, marked improvement in drum drier heat transfer efficiency can be obtained by filming amine use.

Overall, filming amines should be used where oxygen corrosion is the major problem and where high CO2 potential of the boiler feedwater rules out neutralizing amines from economic considerations.

FDA, USDA and OSHA Restrictions on the Use of Amines

The toxicity aspects of steam system corrosion inhibitors suggest than in concentrated solution both morpholine and cyclohexylamine are strongly alkaline and are classed as toxic or hazardous chemicals under provisions of the Occupational Safety and Health Act of 1970 (OSHA) or the Hazardous Substances Act. They are rated toxic if ingested and exhibit irritant and caustic effects when brought in contact with the lungs, skin or eyes. In the 20% to 40% dilutions commonly marketed for industrial water treatment use, toxicity aspects are considerably diminished. However, contact with the eyes or excessive skin contact must be avoided.

Commercially available dilutions of neutralizing amines usually have no flash point or flash points that are well below the minimum of OSHA “flammable” category. Emulsions of octadecylamine generally are non-flammable.

Morpholine was assigned a Threshold Limit Value (TLV) for vapor toxicity of 20 ppm by volume (70 mg/cu meter by weight) at the 1968 meeting of American Conference of Government Industrial Hygienists in St. Louis. For cyclohexylamine vapor, the maximum allowable concentration is usually accepted at 10 ppm by volume.

The possibility of accidental poisoning from either amine is remote because of their strong, unpleasant (fishy) odors and bitter taste. However, because of the alkaline, irritant nature of these amines to the skin on prolonged contact and the possibility of eye damage in case of contact with strong solutions, appropriate safety precautions must be taken when handling the concentrates or making solutions. Protective gloves and goggles or face shield are recommended to prevent skin or eye exposure. Likewise, adequate ventilation should be provided in areas where strong solutions are handles or in case of accidental spills.

Steam humidification – The use of neutralizing amines for steam treatment and other volatile boiler chemicals is limited and controlled as regulated by OSHA Title 29 Code of Federal Regulations Section 1910.1000. This regulation controls an employee’s exposure to neutralizing amines and other air contaminants.

Food preparation – The listed neutralizing amines and octadecylamine are authorized for use in the production of steam in food process. This applies to plants and installations where steam contacts food products within the dosage limits set by the FDA and published in Food Additive Regulations, 21 CFR Subpart D Section 173.310. Authorized food industry applications include the manufacture of paper and paperboard used for food packaging.

The authorizations for use of filming and neutralizing amines in Federally-inspected meat, poultry and egg-processing establishments are obtained through the USDA. These authorizations require that all constituents must be included in Food Additives Regulations 21 CFR Subpart D Section 173.310 covering boiler water additives.

Fluid milk processing – Use of amine products in dairy and fluid milk processing operations is prohibited by FDA. This use is specifically excepted in Boiler Water Additives Section 21 CFR Subpart D Section 173.310..

Autoclaving – The use of steam treated with octadecylamine has been approved by FDA for autoclaving of surgical instruments and gauze in hospitals and drug establishments at limits specified in the regulations. Steam treated with neutralizing amines is also acceptable.

Calculating Boiler Chemical Requirements


A. Equivalents and equivalent weights

The concept of equivalent weights, equivalents, and equivalents per million (epm) is most useful in calculating basic chemical dosages. The equivalent weight of any ion, radical or compound is that weight which will combine with or replace a unit weight of hydrogen. All ions (and compounds) react together in the ratio of their respective equivalent weights. For example, one equivalent weight of sodium (23) reacts with one equivalent of chloride (35.5) to form one equivalent of sodium chloride (58.5).

The equivalent per million (epm) is determined by dividing the concentration in ppm of a substance by its equivalent weight. Once this is determined we can calculate the amount of any chemical required to react with that ion. This is done by multiplying the epm of the ion by the equivalent weight of the treatment chemical. The results provides the chemical dosages in ppm.

B. Purity or active ingredients of a treatment chemical

In making chemical calculations we must adjust equivalent weights of substances to reflect their active content or purity. For example, the equivalent weight of hydrated lime (calcium hydroxide) is 37.1, assuming the lime is 100% active. However, hydrated lime is only 90% active. The equivalent weight of hydrated lime must by adjusted by dividing the equivalent weight of pure lime (37.1) by the activity (0.90) which yields an equivalent weight for the product of 41.1.

C. Analyses in terms of calcium carbonate

In making field analyses, we report hardness and alkalinity in terms of calcium carbonate equivalents, instead of as calcium or magnesium ions. To determine the epm of hardness or alkalinity, we divide the concentration in ppm as calcium carbonate by 50, which is the equivalent weight of calcium carbonate.


1. Calcium as Ca ion = 100 ppm

Equivalent weight of calcium carbonate (CaCO3) = 50

Equivalent weight of calcium (Ca) = 20

Calcium as CaCO3 = 100 ppm X 50 = 250 ppm

20 ppm

The multiplying factor to convert from calcium as the ion to the calcium carbonate equivalent is 2.5 which is the ratio of the equivalent weights.

2. To go from ppm as CaCO3 to Ca, reverse the process.

250 ppm calcium as CaCO3 X 20 ppm = 100 ppm as Ca

50 ppm


A. Estimating the dissolved oxygen content of boiler feedwater

The dissolved oxygen content of feedwater can be estimated from the temperature of the water and the working pressure of the deaerator. The following table presents the dissolved oxygen concentration at a given feedwater temperature and pressure.

Maximum Expected Dissolved Oxygen Level

at Listed Deaerator Temperatures and Pressures











Deaerator Working Pressure, psig

Dissolved Oxygen







































































B. Sodium sulfite dosage

Once the oxygen content of the feedwater has been estimated, the sulfite dosages can be calculated. The sulfite dosage is the sum of the ppm needed to neutralize the dissolved oxygen, plus additional amounts needed to produce an acceptable boiler water residual. For low to moderate pressures, sulfite residuals range from 20 to 40 ppm . The excess required depends on the residual desired in the boiler, and the number of feedwater concentrations maintained in the boiler as controlled by blowdown.

The theoretical dosage of sodium sulfite (100% purity) is 8 ppm sulfite for every 1 ppm (0.7 cc per liter) of dissolved oxygen. However, correction must be made for the activity or purity of the commercial sulfite, which is about 90%, and for the efficiency of the scavenging reaction. From a practical viewpoint, the sulfite dosage is 10 ppm per ppm of dissolved oxygen in the feedwater. Additional sulfite must then be added to produce the required sulfite residual

C. Sample Calculation

Assume boiler is operating at 10 cycles of concentration with feedwater at 205 oF. This equates to a dissolved oxygen content of 0.7 ppm or 0.5 cc per liter. Assume a feedwater demand of 500,000 lbs per day. The desired sulfite residual is 30 ppm.

1. For fixation of dissolved oxygen: 0.7 ppm x 10 ppm = 7 ppm

7 ppm sulfite/120 = 0.0583 lbs per 1000 gallons

2. For sulfite residual at 10 cycles: 0.0250 lbs per 1000 gallons

3. Total sulfite requirement = 0.0833 lbs / 1000 gallons

4. Sulfite required per day is:

500,000 lbs/day X 0.0833 lbs/1000 gal = 5.0 lbs/day

8.34 lbs/gal

D. Hydrazine Calculations

The theoretical dosage of hydrazine (100% active) is 1 ppm per ppm of dissolved oxygen. Because of the explosive nature of pure hydrazine, it is available for industrial use in 35% active solutions. It, therefore, takes 3 ppm of 35% active hydrazine to neutralize 1 ppm of dissolved oxygen. In addition, a low residual of from 1 to 3 ppm is required in the boiler to accelerate the reduction of ferrous oxides.


A. Soda Ash Requirement

Soda ash (Na2CO3) may be used to increase the boiler water alkalinity. It partially decomposes at boiler temperatures and pressures to produce caustic soda (NaOH) and carbon dioxide (CO2) gas. Since the carbon dioxide adds to the neutralizing amine demand in the condensate, the use of caustic soda as the alkalinity builder is preferred.

The soda ash requirement for treatment of a given feedwater is determined from the non-carbonate hardness (H-M) plus the required excess over and above the non-carbonate hardness. “H” and “M” values are determined as calcium carbonate from feed water analysis.

Example: Field tests on a given feedwater show H = 84 ppm and M = 59 ppm. Non-carbonate hardness (H-M) = 25 ppm as CaCO3.

Assume boiler operates at 10 cycles of concentration, and the residual boiler water alkalinity, M, is 250 ppm. To develop this alkalinity, we need 25 ppm soda ash (as CaCO3) in the feedwater (250 ppm boiler alkalinity divided by 10 concentrations), over and above the amount required to neutralize non-carbonate hardness in the feedwater.

From the preceding data, soda ash required is non-carbonate hardness plus excess = 25 ppm + 25 ppm = 50 ppm as CaCO3. Dividing 50 ppm by 50 (equivalent weight of CaCO3) we find 1 epm of soda ash is required.

The equivalent weight of soda ash (Na2CO3) is 53.0. So, 1 epm required X 53 = 53 ppm soda ash needed. If we divide 53 ppm by 120 we determine that 0.44 lbs. per 1000 gallons of soda ash is required.

B. Caustic Soda

Caustic soda is commonly used for direct upward adjustment of the total alkalinity in the boiler. The equivalent weight of caustic soda (NaOH) is 40. It takes 0.8 ppm NaOH (as 100% active) to produce a 1 ppm increase in total alkalinity as CaCO3. Caustic soda is commonly purchased as 50% liquid solution. In this case, 1.6 ppm of 50% liquid caustic is required to produce a 1 ppm increase in total alkalinity as CaCO3.

Example: The boiler feedwater is zeolite softened (0 total hardness). Total alkalinity of the feedwater is determined from field tests to be 15 ppm. The desired alkalinity in the boiler is 400 ppm “M” alkalinity. The boiler operates at 20 cycles of concentration. What is the required dosage of caustic soda when purchased as a 50% liquid solution.

The total alkalinity required in the feedwater is 400 ppm divided by 20 cycles of concentration = 20 ppm. Natural alkalinity is 15 ppm. The supplemental alkalinity requirement is, therefore, 20 ppm minus 15 ppm = 5 ppm required as calcium carbonate.

Dividing 5 ppm by 50 (equivalent weight of calcium carbonate) gives 0.10 epm caustic soda (as 100% active), or 0.20 epm as 50% active. The equivalent weight of NaOH is 40, so 0.20 epm X 40 = 8 ppm caustic soda (50%) required. Dividing 8 ppm by 120 gives 0.067 lbs of 50% caustic soda per 1000 gallons of feedwater is required to produce 400 ppm total alkalinity as CaCO3 in the boiler.


A. Phosphate products

Many products are available, both in powder and liquid form, to provide the required phosphate residual in the boiler. In most cases, the available phosphate in a water treatment product will be given as percent P2O5, but the specific mixture of phosphates used will not be disclosed. The P2O5 content and equivalent weights for various phosphates are tabulated in the following table.

Equivalent Weights of Phosphate Products

Phosphate Compound Approx. %P2O5 Equiv. Wt.

Trisodium phosphate – 12 H2O 18.7% 126.7

Trisodium phosphate (anhydrous) 43.2% 54.7

Disodium phosphate – 12 H2O 19.8% 119.4

Disodium phosphate (anhydrous) 50.0% 47.3

Monosodium phosphate – 1 H2O 51.4% 46.0

Monosodium phosphate (anhydrous) 59.1% 40.0

Sodium hexametaphosphate (anhydrous) 67.5 to 69.0% 34.0

Sodium tripolyphosphate (anhydrous) 57.0 to 58.0% 40.9

P2O5 100% 23.7

From these figures we can determine the equivalent weight of any phosphate or phosphate mixture from which the P205 content is known. Divide the equivalent weight of P2O5 (23.7) by the percentage of P2O5 in the product.

B. Phosphate Calculation

Phosphate reacts with feedwater hardness and OH alkalinity to produce hydroxyapatite, an insoluble sludge. If we write the chemical equation for hydroxyapatite as 3Ca3(PO4)2 * Ca(OH)2 and calculate the molecular weight of the compound, we see that only a little over 90% of the calcium reacts with the phosphate. The remainder combines with OH alkalinity to form calcium hydroxide. Since only 90% of the calcium hardness reacts with phosphate, the calculated amount of phosphate can be reduced by 10%.

Example: Determine the amount of disodium phosphate required to precipitate 30 ppm calcium (as Ca) to form hydroxyapatite.

30 ppm calcium as calcium ion equals 75.o ppm calcium as calcium carbonate, or 1.5 epm. Anhydrous disodium phosphate has an equivalent weight of 47.3. Therefore, to produce hydroxyapatite, multiply the epm calcium times the equivalent weight of disodium phosphate time 0.90 to determine the parts per million disodium phosphate required. The 0.90 factor reflects that only 90% of the calcium reacts with the phosphate. Completing this equation we have:

1.5 X 47.3 X 0.9 = 64 ppm disodium phosphate (anhyd)

This is the amount of phosphate required to react with 30 ppm calcium ion. 64 ppm disodium phosphate divided by 120 equals 0.533 lbs disodium phosphate per 1000 gallons treated water.

If we compare the calcium content expressed as calcium carbonate (75 ppm) with the amount of disodium phosphate required (64 ppm) we see that 0.85 ppm disodium phosphate is required to react with each ppm calcium expressed as calcium carbonate. Or, if we express calcium as the ion, about 2.13 ppm phosphate is required to react with each ppm calcium.


A. Polymers

Natural and synthetic polymers are routinely used to condition the sludge produced by the hardness precipitation reactions with phosphate and alkalinity in the boiler. Natural tannins, lignins, and synthetic acrylates and polyacrylates are examples. Typical polymer dosages are between 5 and 25 ppm as 100% active polymer. Boiler polymers are marketed as dilute solutions of these active ingredients, however, so overall product dosages are between 100 and 500 ppm.

Since boiler polymers are non-volatile, they concentrate in the boiler. Chemical procedures are available for estimating the polymer residual, but the tests are difficult to perform, and the accuracy is not very good. For these reasons, the dosage of polymeric sludge conditioners are frequently determined by direct calculation from steam production and cycles of concentration data. The dosage of the polymer product is adjusted by regulating the output of the chemical pump, or boiler blow down.

B. Chelants

Chelants such as Na4EDTA and Na3NTA are used as internal treatments for scale control in boilers. Chelants must be used with oxygen scavengers, alkalinity builders, antifoam, and polymer additives for a complete boiler water treatment program. The major advantage of chelant treatment is that no insoluble deposits form. Calcium and magnesium are held in solution.

Four (4) parts of Na4EDTA (as the dry salt) are required to complex 1 part of calcium hardness. Liquid solutions of chelant require a proportionately higher dosage. Because of the expense of chelants compared to phosphate-based programs. Chelant treatment is only appropriate for high quality feedwaters averaging less than 2 ppm total hardness; preferably less than 1.0 ppm. EDTA must be added after feedwater has been deaerated and oxygen traces scavenged with catalyzed sulfite. Otherwise, EDTA reacts with dissolved oxygen. (1 ppm O2 destroys 100 ppm EDTA)

EDTA is used in boilers up to 1200 psig. The practical limit for NTA programs is about 850 psig. For practical purposes, NTA is no longer commonly used in boiler applications.

This material originally prepared by:

J. Fred Wilkes

Consulting Chemical Engineer

P.O. Box 2320

Titusville, FL 32781-2320

Revised and edited by:

William F. Harfst

Chemical Consultant

Harfst and Associates, Inc.

P.O. Box 276

Crystal Lake, Illinois 60039

Chemical Treatment Requirements for Steam and Hot Water Systems

Boilers, in one form or another, have been used for centuries to produce steam for heat and power. Today, boilers produce steam to drive electric turbines, heat buildings, and provide power for countless industrial and commercial applications. Although the steam uses may vary, the importance of maintaining the water quality in these systems remains the same. Without good water quality management, steam boilers and condensate systems are susceptible to costly operating problems and unscheduled outages.

This blog post discusses the problems water can cause in steam and hot water systems, and how these problems can be prevented or controlled by an effective water management program. We will cover the following topics:

  • Boilers and auxiliaries

  • Problems water can cause in steam systems

  • Removing dissolved oxygen

  • Deposit control methods

  • Corrosion and corrosion control

  • Steam purity and quality


Boiler design and operating characteristics play an important role in the selection of water treatment equipment, chemicals, and control ranges.

What is a boiler? A boiler is simply a heat exchanger that uses radiant heat and hot flue gases, liberated from the burning of fuel, to generate steam and hot water for heating and process loads, including power generation. In a steam plant, the boiler is best defined as a vessel for the generation and enclosing of a vapor under pressure.

A steam plant consists of three (3) essential elements — (1) a furnace to convert chemical energy to thermal energy through combustion, (2) a boiler to convert the thermal and radiant energy of combustion gases to heat content (enthalpy) of the steam through heat transfer, and (3) a prime mover to convert a portion of the heat content of the steam to mechanical energy through expansion.

A boiler system may be considered as three (3) separate sections — (1) a feed system that supplies water to the boiler, (2) the boiler itself, and (3) the steam and condensate system.

Most industrial boilers operate in the low to medium pressure range. These pressure ranges are identified arbitrarily as less than 300 psig (low) and 301 to 600 psig (medium). Many industrial boilers operate in the range of 601 to 900 psig or higher, but as working pressure increases, feedwater quality requirements become progressively more stringent, and the nature of the water treatment program must also vary. Certain industrial boilers are designed to operate in the high pressure range of 1000 to 2000 psig. Utility steam generators operate at very high pressures (above 2000 psig), or even above the critical pressure and temperature (3206.2 psig and 705.4 oF) at which steam and water weigh the same.

Industrial boilers are best categorized as either firetube or watertube.

Firetube boilers – In these units, the hot fireside gases are passed through the tubes with the water on the outside of the tubes, all of which is contained within a pressure vessel.

Wet back firetube boilers have a waterwall at the back of the boiler in the area where the combustion gases reverse direction to make a return pass through the boiler tubes. This design increases the overall heat transfer area of the boiler.

Dry back firetube boilers have refractory at the back end instead of a water wall. Internal maintenance is simplified, but refractory replacement is expensive and overheating, gouging and cracking of tube ends at the entrance to the return gas passages often cause problems.

Watertube boilers – In these boilers the water is passed through the tubes and the hot combustion gases are on the outside of the tubes.

In terms of number of units in service, firetube boilers are more common than watertube boilers. Inherent safety problems with firetube boilers sparked the further development of the watertube design. The small diameter tubes used to contain water in the watertube boiler are not stressed critically even when pressure is increased to thousands of pounds per square inch. This greatly reduced the explosion risk associated with the firetube boiler. Today, firetube boilers are limited to operating pressures of about 300 psig.

Several watertube boiler types are found in service today.

Longitudinal (Long) Drum, Straight Tube boiler generates 5,000 to 80,000 pounds of steam per hour with design pressures ranging from 160 to 325 psi. The maximum size of a longitudinal (long) drum boiler is limited because steam and water circulating in tubes from the front and back headers are connected to the steam drum in circumferential rows. The drum diameter determines the number of tubes that can be connected to the drum and limits the width and, therefore, the size of the boiler.

Cross Drum, Straight Tube boilers have the steam drum placed at right angles to the boiler tubes. This boiler could conceivably be built to any required capacity. Boilers of this type have a wide field application and have been designed from 5,000 to 525,000 pounds of steam per hour with design pressures from 160 to 1450 psi.

Bent Tube Design boilers are more commonly encountered today than the straight tube design. The bent tube design opened the door for practically unlimited size and pressure ratings for boilers. Several types of bent tube boilers have been designed for a wide variety of applications. One is the low-head bent tube multi-drum boiler that is similar to the straight inclined tube box-header boiler with a cross drum except that the tubes are directly connected to drums and better facilities are available for removal of sludge and separation of steam.

The integral furnace-type boiler uses heat-adsorbing water wall tubes instead of the firebrick previously used and these water wall tubes are incorporated in the complete boiler water circulation design.

The most popular and efficient boiler of the bent tube type has been the Stirling design. The original Stirling had two upper drums and a lower drum, but was changed to three upper drums and one lower drum later. The main advantage of the multi-drum bent tube design is its ability to produce steam of good quality where water conditions cannot be maintained within reasonable limits, and its ability to handle widely fluctuating loads.

With the development of better methods of externally treating water, it was possible to design a more efficient steam boiler design that was simpler than the four drum design. This is the two drum boiler with one drum directly over the other along with water cooled furnace walls. This is the basic design of the present high pressure, high temperature boilers used in industrial and utility power plants.

A-type boilers have two small lower drums and a larger upper steam drum. This permits efficient steam/water separation. Most steam production occurs in the center furnace wall tubes entering the steam drum.

D-type boilers are flexible units with a mud drum and steam drum stacked vertically with water wall combustion chamber plus multiple convective passes. The fire is either along the drum line or at right angles to the drum.

O-type boilers are also compact 2 drum design. Boiler tubes form an O-shape as they connect to the top and bottom drums. Because of height limitations in transportation, the units are frequently quite long. The fire is down the center of the boiler in line with the drums.

Packaged Steam Generators have become very common as the need for boilers that can be shipped and installed as a single unit has steadily grown. These are shop built, skid mounted units that contain all the accessories and appurtenances necessary for controlling the feedwater, steam and fuel. Packaged steam generators may be firetube or watertube types.

High Temperature Hot Water (HTHW) Generators for institutional, industrial and commercial heating installations have experienced lagging interest because of sophisticated system design requirements. Nevertheless, by increasing the temperature (351 to 450 o F) and pressure of hot water, and increasing the size of the generator, some advantages are gained over steam heating systems. HTHW properties are materially different from steam. It has high density, high specific heat, low viscosity and good thermal conductivity. Pressure changes have negligible effect on the density, specific heat and thermal conductivity of HTHW.

High Pressure, High Temperature Boilers for utility power applications have experienced steady increases in operating pressures from 1200 psi to 5500 psi. Superheated steam temperatures have increased from 700 o F to 1200 o F. Boiler capacities have increased from around 140,000 pounds of steam per hour to about 3,000,000 pounds of steam per hour. These increases have been necessitated by growing power requirements.

The increased costs of fuel, labor and materials have forced the introduction of more efficient boilers and turbine generator sets to supply more power at less cost. The greatest gain in thermal efficiency is obtained by superheating steam that is already at a high temperature corresponding to the saturation steam temperature for the higher pressures. Another gain in efficiency is obtained by reheating the steam.

Boiler Rating Factors

In the early days of steam power, boilers and engines were coordinated in size through knowledge that the typical steam engine required about 34 pounds of steam per horsepower-hour. A typical 100 horsepower steam engine required 3,450 pounds of steam per hour when it was operated at rated load. A terminology was created in the boiler industry to fit these facts. A 3,450 pounds per hour steam boiler was called a 100-horsepower boiler. It was sized appropriately for a 100 horsepower engine. Thirty-four and one half pounds of boiler steam per hour became know as developed boiler horsepower. The magnitude of the developed boiler horsepower was subject to the minor variations in steam and feedwater conditions common to the age.

Boiler Rating Factors

1 Boiler horsepower

Nominally 34.5 pounds of steam per hour

1 Rated boiler horsepower

10 square feet of effective heating surface (6, 8 and 12 square feet have also been used

1 Developed boiler horsepower

33, 475 Btu absorbed per hour

(34.5 lbs water X 970.3 Btu/lb

Percent rated capacity

Developed horsepower / rated horsepower

Factor of evaporation

Btu absorbed per pound of steam / 970.3

Actual evaporation

Pounds of steam generated per hour

Equivalent evaporation

Actual evaporation X factor of evaporation

Kilo Btu per hour, kB

Total Btu absorbed per hour / 1000

Meg Btu per hour, MB

Total Btu absorbed per hour / 1,000,000

Steam Drum Internals

One of the mains challenges in boiler design is to develop a unit that will produce the required amount of steam of a quality below guaranteed moisture content. A normal guarantee for steam quality is 0.5% moisture. However, this moisture is boiler water. If the boiler water contains 2,000 ppm of dissolved solids, 0.5% moisture or one half pound of boiler water per 100 pounds of steam would result in a concentration of 10 ppm solids in the steam. This means that every 1 million pounds of steam delivered by the boiler would carry along 10 pounds of solids.

Most modern generating plants demand very high steam purities with total solids entrainment well below 0.5 ppm even though boiler industry guarantees only go down to 1 ppm. To obtain this degree of purity, boiler steam drum internals must be highly efficient and maintained in proper working order. In addition, boiler water conditions must be maintained so that foaming caused by high boiler water solids concentration and alkalinity or by contaminants such as oil or grease is held to a minimum.

Two common steam separator designs are the baffle plate and corrugated separator. These are routinely used in modern steam boilers to achieve the guaranteed steam purity targets. For separation at high pressures, centrifugal units are required. Two common types of centrifugal separators are the cyclone and the turbo units.

Steam drum internals maintain steam quality and purity

Steam separators, even of the centrifugal type, are incapable of reducing silica concentration in the steam. As the silica is in the vapor form, it cannot be removed by simple steam separation. However, steam washers have been developed that show promise as a means of reducing silica in steam.

Heat Recovery Boiler Components

With firetube boiler designs, considerable heat value is lost in combustion gases discharged to the stack at elevated temperatures. For most firetube designs, superheaters and air preheaters are not practical. Economizers, used to raise boiler feedwater temperature, may be employed for some heat recovery.

Modern watertube boilers may include several components to recover added heat from combustion gases. These include superheaters, reheaters, economizers, stage heaters, and air preheaters. Substantial increases in steam generator efficiency are obtainable by the use of economizers (1% increase for each 10 o F increase in feedwater temperature) and air preheaters (2.5% efficiency increase for each 100 o F drop in exit gas temperatures). This means a 2% efficiency gain for each 100 o F increase in combustion air temperature.

Economizers usually employ cast iron or steel tube heat exchangers to preheat feedwater. Finned tubing is used to extend the heat absorbing surfaces. Cast iron is chosen where flue gas temperatures are low and acid condensation is expected. Economizers may be located integrally within the boiler setting or separately installed in the flue gas flow preceding the air preheater.

Air Heaters achieve final heat recovery from boiler flue gases. Two basic types are tubular and regenerative. Tubular air heaters consist of cast iron or steel tubes joined to tube sheets that are enclosed in a reinforced casing that have air and gas inlets/outlets. They are available in vertical and horizontal designs.

Regenerative air heaters contain rotating (2 – 3 rpm) heat storage elements that accumulate heat from flue gas and transfer it to incoming cold air. These offer large contact areas for heat transfer with little resistance to air and gas flow. Corrosion can be a problem in the low temperature gas zone with either type.


Regardless of the boiler type, water can cause problems in steam and condensate systems. These problems generally fall into one of three categories

  • Scale

  • Corrosion

  • Carryover

Scale forms in the boiler as a result of the precipitation of impurities in the boiler feedwater. The common scales found in modern steam boilers include calcium carbonate, calcium phosphate, calcium sulfate, silica and iron oxides. These scale deposits generally form in high heat transfer areas, but can also be found as loose material in the lower drum or in the steam drum below the water line.

Because most plants soften the boiler feedwater, calcium scales are less common than in the old days when feedwater contained appreciable hardness. As a result, most scale deposits contain a high percentage of iron oxides that have been returned to the boiler through the condensate system.

Corrosion is an ongoing process in steam and condensate systems. The type of corrosion, however, varies depending on the environment and the mechanism of attack.

Pitting-type attack is common on boiler metal surfaces. This is caused by dissolved oxygen in the boiler feedwater. This form of corrosion is readily identified by the numerous small pits that form on the boiler tubes and drums.

Caustic attack occurs wherever free “OH” alkalinity can concentrate in the boiler. Typically, this is at a crevice or leak where boiler water under pressure can flash to steam leaving concentrated boiler water salts behind. Caustic attack also occurs beneath accumulated corrosion products and deposits whenever “OH” alkalinity is present. In firetube boilers, caustic attack is frequently observed at the ends of the boiler tubes when refractory-backed boilers operate for extended periods at maximum output. The white hot refractory promotes steam blanketing or caustic gouging right at the rolled tube ends.

Carryover refers to the contamination of the steam with boiler water. It is caused by several conditions common to both firetube and watertube boilers.

Misting occurs when small water droplets are carried with the steam. These droplets form each time a steam bubble separates from the water surface much like the mist that forms when a carbonated soda is poured into a glass. The release of gas bubbles creates small droplets or a mist on the surface of the soda.

Foaming is a result of “bubbles” in the boiler water that cause an expansion of water volume in the boiler. Foaming is promoted by high total dissolved solids, high alkalinity, and contamination by organics like oil and grease.

Priming refers to “surging” of water in the boiler drum. This is usually related to the boiler design and operating conditions. A common cause is rapid increases in load that cause a rise in water level in the boiler.

Mechanical conditions may also promote carryover. These include maintaining too high of a water level, malfunction of steam separation equipment, and as mentioned previously, sudden load swings.

Volatile carryover is characterized by the vaporization of silica into the steam. These volatile solids can deposit downstream on steam turbine blades and other steam-using equipment.


The objective of a sound boiler water treatment program is to prevent or minimize the problems water can cause in modern steam boilers. Likewise, corrosion control in the steam condensate system is necessary to protect system piping and minimize the accumulation of iron oxides in the boiler.

Scale and deposit control begins with proper external treatment of the boiler feedwater. This includes hardness removal by softening the makeup and internal treatment of the boiler water to prevent the formation of scale on heat transfer surfaces. Various softening methods are utilized including ion exchange, reverse osmosis, and electrodialysis (EDR). High pressure boiler applications require further pretreatment of feedwater to remove hardness, alkalinity, and silica. In this case, complete demineralization of the makeup is required.

Internal treatment of the boiler water includes keeping the hardness in solution by the addition of chelants such as EDTA, or by the controlled precipitation of solids by the addition of phosphate. In addition, various polymers and sequestering agents are utilized to prevent the deposition of scale deposits on heat transfer surfaces.

Corrosion control begins by removing the dissolved oxygen from the boiler feedwater. This is best accomplished by mechanical deaeration of the feedwater in combination with the supplemental addition of a chemical oxygen scavenger such as sodium sulfite or hydrazine. Further improvements are realized by reducing the carbon dioxide potential of the steam by carbonate and bicarbonate alkalinity reduction in the boiler feedwater. Reducing the carbon dioxide (CO2) concentration in the steam decreases the tendency of carbonic acid to form in the condensate.

Carryover is primarily controlled mechanically by the use of steam separation equipment and by proper operating practice. Some forms of carryover, however, such as foaming can be minimized by controlling the free “OH” alkalinity of the boiler water and the addition of chemical antifoams.


Dissolved oxygen and carbon dioxide cause problems in boilers and condensate systems. Oxygen is the cause of pitting-type corrosion in boilers and feedwater systems. Carbon dioxide volatizes with the steam to produce carbonic acid in the condensate, which is corrosive to system metals. Removal of these gases from the boiler feedwater is necessary to control these corrosion reactions.

Oxygen enters the steam system in the makeup water. Fresh makeup typically contains 6 to 8 ppm dissolved oxygen. Additional oxygen may be drawn into the condensate from areas that are under vacuum. Oxygen present in the makeup and condensate combine to produce a corrosive mix in the boiler feedwater. Unless it is removed, it will promote oxygen pitting attack of the feedwater lines and boiler internals.

Carbon dioxide may be present in the fresh makeup, but it is primarily produced by the thermal breakdown of bicarbonate alkalinity in the boiler to produce free COgas. The CO2 is carried by the steam where it dissolves in the condensate to form carbonic acid.

Deaeration is the removal of corrosive gases, such as oxygen and carbon dioxide, from the boiler feedwater. The deaeration principle is based on two fundamental chemical laws. The first states that the solubility of a gas decreases with an increase in the temperature of the liquid. The second, Henry’s Law, says that the concentration of the dissolved gas is proportional to the partial pressure of the gas in the free space above the liquid. From this it is clear that removal of dissolved gases can be accomplished by heating the liquid and by reducing the system pressure.

Carbon dioxide can be removed from cold water by passing it through an aerator or degasifier. Degasifier columns are constructed with trays or packing to break the water into smaller droplets to enhance the removal efficiency of the dissolved gas. A flow of air is passed through the column, counter to the water flow, where it strips the CO2 from the water. The air stream is then exhausted to the atmosphere. This process can also be carried out under vacuum.


A typical deaerator consists of a deaerating section and a storage section. The deaerating section has three basic zones:

  • Inlet spray and vent condenser

  • Heating and distributing zone

  • Deaerating zone

Tray-type Deaerator

Water enters the deaerator through the upper dome and is sprayed or atomized by spray valves into an atmosphere of steam. Initial heating of the water takes place in this zone.

The water then passes onto the distribution trays for further heating. The distribution trays are designed so the water overflowing the top trays cascades to the next tray and then on to each successive tray. The water is heated to the saturation temperature of the steam in this zone. This accomplishes essentially 100% oxygen and CO2 removal.

Low pressure steam provides the “heating and scrubbing” action in the deaerator. The steam enters the deaerator through the shell side and provides a blanket of steam entirely surrounding the internal compartment. It then flows into the bottom of the deaerating compartment and upward through the deaerating tray zone. The hottest steam meets the hottest water in the lower tier of trays, which ensures optimum gas removal.

The final path of the steam passes by the vent condenser. Here the influent water is preheated and final condensation of the steam occurs. The remaining steam is exhausted, along with the non-condensable gases, to the atmosphere.

Spray-type Deaerator

The spray-type deaerator breaks the influent water flow into smaller droplets by spray nozzles instead of a series of trays. The principles of deaeration are the same, however.

Both spray and tray type deaerators utilize similar heating sections where initial feedwater heating takes place and close to 90% of the gas release occurs. This is done either by an inlet spray pipe or by spring-loaded water spray nozzles. The intimate mixing of the atomizing spray with the steam provides a high degree of scrubbing action to release the gases from the water surface.


Tray and spray type deaerators have the same basic requirements for complete removal of dissolved gases.

  • Heat the water to full saturation temperature and pressure

  • Agitate to expose maximum surface to scrubbing atmosphere

  • Vent non-condensable gases to the atmosphere

The various deaerator designs accomplish these objectives in different ways, but the basic requirements for oxygen removal remain the same. For spray deaerators, a minimum of a 50 o F increase in temperature is required to allow spray-type units to operate effectively at full load. This means that spray type deaerators operating in the range of 2 to 5 psig would require maximum inlet water temperatures of around 170 o F at all times. As a result, the effectiveness of deaeration in a spray type unit decreases as the load increases. Generally, deaerators perform poorly at loads below 25% of design rating because the heating steam requirements are not sufficient to maintain high steam flow and velocity.

Tray type deaerators use a different method for the release of the dissolved gases. Tray type deaerators are designed with as many as 24 tray tiers to permit adequate cascading and maximum surface exposure to the scrubbing action of the steam. Regardless of inlet water temperature, the same amount of water surface is exposed for gas release. As operating loads change, the ratio of surface area to water flow increases. This assures effective deaeration under all inlet water temperatures and flow conditions. For a deaerator used under varying load conditions or at high inlet water temperatures, the tray type deaerator gives the most satisfactory results over the entire operating range.


Sufficient steam flow is required to heat the water to saturation and transport the dissolved gases to the atmosphere. Excessive steam flow, however, wastes energy. How much steam is required to insure optimum dissolved gas removal at maximum efficiency?

As a general rule of thumb, the following formula is useful in calculating the vent rate of any deaerator.

Total Vent Rate, lbs/hr = Operating pressure (absolute)

Times the cubic centimeters O2 in influent

Times the capacity of the unit, lbs/hr

Divided by 200,000 (a constant)

The amount of steam required to heat the makeup and condensate to within 1 o F of the saturated steam temperature is equal to 15 to 16 percent of the total deaerator output in pounds per hour. This is normal for a unit having an average percentage of cold water makeup. The steam requirement is variable, especially if the makeup is first treated in a hot process softener or comes from an evaporator or passes through stage heaters. If superheat exists in the heating steam, less steam flow is required. The amount of steam required by the deaerator varies from 0.98 to 1.10 percent decrease in steam flow for every 20 o F rise in superheat temperature over the saturation temperature of the steam at design working pressure. This does not include the total steam requirement needed for venting, nor does it include the amount of heat that is lost from the deaerator through radiation losses.

Overall, corrosion damage caused by dissolved oxygen and carbon dioxide in boiler, turbines, process equipment and steam distribution systems is a costly problem. Complete removal of these gases by mechanical deaeration is the best method of prevention.


Steam and hot water generators are specialized heat exchangers designed to heat water to saturation temperature and pressure. Generally constructed from steel, modern boilers strive to accomplish this basic function with maximum heat transfer efficiency. This imposes increased heat flux across the boiler tube, which mandates ever increasing vigilance over the boiler water chemistry.

The quality of the water used in these units poses a threat to their continuous, reliable and safe operation. Mineral impurities may deposit on boiler surfaces, reducing the heat transfer efficiency. Dissolved gases cause corrosion of boiler metal. And impurities such as silica may adversely affect the steam purity. These potential problems, and others, must be taken into consideration when designing an effective boiler water treatment program.


Boiler deposits form as a result of the precipitation of mineral impurities present in the feedwater. These impurities include calcium and magnesium hardness, iron corrosion products and silica.

Scale deposits generally form in the areas of highest heat transfer because of the inverse solubility of many calcium salts with temperature. These deposits have an insulating property that reduce heat transfer efficiency and increase tube metal temperature. Severe deposition results in overheating type boiler failures. Typical boiler scales include calcium carbonate, calcium sulfate, calcium phosphate, iron oxide, magnesium silicate and magnesium hydroxide.

Scale deposits also contribute to boiler corrosion. Porous deposits allow the concentration of caustic soda under the deposits, resulting in caustic embrittlement of boiler metal.

Control of boiler deposits is best achieved by effective pretreatment methods for hardness and alkalinity reduction in combination with various internal treatment methods to produce boiler sludge that is non-adherent and easily removed by routine blowdown. Internal boiler treatment programs fall into two categories: either precipitating phosphate cycle or non-precipitating chelant programs.

Chelant Treatment Program

Chelants react with feedwater impurities to form soluble complexes with hardness, iron and copper. The two chemicals used for this purpose are ethylaminediaminetetraacetic acid (EDTA) and nitrilotriacetic acid (NTA). (Because of concerns over the potential carcinogenic properties of NTA it has fallen out of favor in boiler water treatment applications.) The chelant “holds on” to the complexed ion to prevent it from precipitating as a boiler deposit.

EDTA forms a more stable complex than NTA and finds application in boilers up to 1200 psig. NTA is limited to operating pressures of less than 900 psig.

Chelants must be applied to feedwaters with less than 2 ppm total hardness. The feedwater must be fully deaerated to prevent decomposition of the chelants. High hardness, aerated feedwaters substantially increase the chelant demand, and because of the higher cost of the chelant chemical, make the economics of the treatment program unfavorable.

Chelants are aggressive toward steel. The chelant solution must be fed to the boiler water line through a stainless steel injection quill into the center of the pipe ID. This gives the chemical time to complex with the feedwater impurities and become fully diluted prior to coming into contact with boiler metal. High levels of chelants will attack the feedwater piping and dissolve protective iron oxide from the boiler.

Chelant dosages are controlled either by calculating the total chelant demand of calcium, magnesium, iron, copper and aluminum in the feedwater and then injecting stoichiometric amounts to meet this demand, or by substoichiometric injection of only enough chelant to complex the calcium hardness. Excess chelant residuals in the boiler are kept to a minimum to prevent boiler corrosion. Typical free chelant residuals in the feedwater are maintained in the 1 to 2 ppm range.

Four (4) parts of Na4EDTA (as the dry salt) are required to complex 1 part of calcium hardness. Liquid solutions of chelant require a proportionately higher dosage. Because of the expense of chelants compared to phosphate-based programs, chelant treatment is only appropriate for high quality feedwaters averaging less than 2 ppm total hardness; preferably less than 1.0 ppm. EDTA must be added after the feedwater has been deaerated and any oxygen traces scavenged with catalyzed sulfite. Otherwise, EDTA reacts with dissolved oxygen. 1 ppm O2 destroys 100 ppm EDTA.

Phosphate Treatment Program

Conventional phosphate treatment utilizes inorganic phosphate, such as sodium phosphate, to react with calcium hardness to produce an insoluble salt of calcium phosphate. Excess hydroxide alkalinity (OH alkalinity) is required to promote the reaction of calcium and phosphate to form calcium hydroxyapatite Ca3(PO4)2(OH)2, a nonadherent form of calcium phosphate that is more readily removed in the boiler blowdown. Magnesium hardness also reacts with hydroxide alkalinity to form magnesium hydroxide, Mg(OH)2 , sludge. If silica is present, it is absorbed onto the magnesium hydroxide to form serpentine.

The conventional phosphate method is applicable to boilers operating up to 1000 psig. Pretreatment of the boiler feedwater for hardness reduction is recommended, but successful application of this approach has been achieved with feedwater hardness as high as 60 ppm. An excess of 20 to 40 ppm orthophosphate is maintained in the boiler water to drive the precipitation reaction to completion and guard against upsets in feedwater hardness. Hydroxide alkalinity (OH) is provided by the addition of caustic soda. Excess OH residuals of 50 to 250 ppm as OH are recommended for complete formation of hydroxyapatite and serpentine sludge.

Coordinated phosphate treatment limits the amount of free hydroxide alkalinity present in the boiler water to prevent caustic corrosion of boiler metal. The amount of OH alkalinity is controlled by the PO4 and H2O equilibrium:

PO4-3 + H2O —– HPO4-2 + OH

The sodium to phosphate ratio (Na:PO4) is maintained between 2.85:1 and 3.0:1 according to the pH versus PO4 equilibrium curve. Above the equilibrium line, free caustic soda exists with trisodium phosphate. Below the curve, no free OH alkalinity is present.

Coordinated phosphate programs are used in high pressure boilers operating in the 1000 to 1500 psig range. Demineralized feedwater is required to maintain an alkalinity-free makeup. A phosphate residual of 5 to 15 ppm PO4 is maintained in the boiler. Sodium hydroxide, disodium phosphate and trisodium phosphate are utilized to control and adjust the pH of the boiler water along the equilibrium curve. Disodium phosphate, being more acidic, depresses the boiler pH. Trisodium phosphate reacts in the boiler to produce caustic alkalinity and raise the pH. Careful monitoring and control of boiler water chemistry is required to maintain the equilibrium balance between pH and phosphate residual.

Coordinated phosphate treatment is very effective for boilers operating in the 1000 to 1500 psig range. Above 1500 psig, however, this program often results in caustic corrosion of boiler metal. This is an indirect result of phosphate hideout phenomena. If boiler water solids concentrate or dry out on the boiler surface, trisodium phosphate does not precipitate or dry out as trisodium phosphate (Na3PO4), but rather as a chemical composition of sodium hydrogen phosphate having the composition ratio of Na2.8H0.2PO4. The molar ratio of sodium to phosphate is known as the congruent ratio. As this reaction occurs, sodium hydroxide forms in the boiler water as follows:

Na3PO4 + 0.2H2O —– Na2.8H0.2PO4 + 0.2NaOH

The free caustic soda (NaOH) is then available to concentrate on the metal surfaces. This typically happens under corrosion products or within pockets of steam that prevent rinsing of concentrated solids from the tube wall.

Congruent phosphate treatment is a variation of the coordinated phosphate regime. This boiler water control method was developed to insure that hydroxide formation does not occur under hideout conditions. Here the boiler water chemistry is controlled to achieve “zero” free caustic alkalinity. The sodium to phosphate ratio is maintained between 2.3:1 and 2.6:1 via the pH versus PO4 control chart. These limits are set based on the premise that the Na to PO4 ratio in the boiler should be no greater than the congruent ratio. Under these conditions, no free sodium hydroxide should form anywhere in the boiler, thus preventing caustic attack of steel. A mixture of trisodium and disodium phophsate is used to maintain the boiler pH and phosphate levels below the 2.6:1 equilibrium curve, i.e. no free OH alkalinity.

Congruent phosphate programs find application in boilers operating in the 1500 to 2500 psig range. Demineralized feedwater is required. Typically, phosphate residuals are maintained between 2 and 5 ppm.

Equilibrium phosphate treatment allows some free hydroxide alkalinity, but the sodium to phosphate ratio is not determined by the equilibrium curve used in the coordinated and congruent phosphate methods. Here the phosphate is maintained at less than 2.4 ppm with hydroxide alkalinity kept at less than 1.0 ppm OH. In this way, the pH becomes a function of the amount of OH present, typically pH 9.3 to 9.6. The problem of phosphate hideout, common in the congruent and coordinated phosphate programs, is reported to be minimal under this control mechanism.

Summary of Phosphate Treatment Programs







20 to 40

50 to 250

Not Appl

11 to 12


5 to 25


2.85 to 3.1

9 to 10.5


2 to 5


2.3 to 2.6

8.8 to 9.4


< 2 to 4

< 1.0

Not Appl

9.3 to 9.6

Phosphate Products

Many products are available, both in powder and liquid form, to provide the required phosphate residual in the boiler. In most cases, the available phosphate in the product will be given as percent P2O5, but the specific mixture of phosphates used will not be disclosed. The P2O5 content and equivalent weights for various phosphates are presented in the following table.

Equivalent Weights of Phosphate Products

Phosphate Compound





Trisodium phosphate -12 H2O



Trisodium phosphate (anhydrous)



Disodium phosphate – 12 H2O



Disodium phosphate (anhydrous)



Monosodium phosphate – 1 H2O



Monosodium phosphate (anhydrous)



Sodium hexametaphosphate (anhydrous)

67.5 to 69.0


Sodium tripolyphosphate (anhydrous)

57.0 to 58.0





From these figures we can determine the equivalent weight of any phosphate or phosphate mixture from which the P2O5 content is known. Simply divide the equivalent weight of P2O5 (23.7) by the percentage of P2O5 in the product.

Phosphate reacts with feedwater hardness and OH alkalinity to produce hydroxyapatite, an insoluble sludge. If we write the chemical equation for hydroxyapatite as 3Ca3(PO4)2 * Ca(OH)2 and calculate the molecular weight of the compound, we see that only a little over 90% of the calcium reacts with the phosphate. The remainder combines with OH alkalinity to form calcium hydroxide. Since only 90% of the calcium hardness reacts with phosphate, the calculated amount of phosphate can be reduced by 10%.

Example: Determine the amount of disodium phosphate required to precipitate 30 ppm calcium (as Ca) to form hydroxyapatite.

Answer: 30 ppm calcium as calcium ion equals 75.0 ppm calcium as calcium carbonate, or 1.5 equivalent per million (epm). Anhydrous disodium phosphate has an equivalent weight of 47.3. Therefore, to produce hydroxyapatite, multiply the epm calcium times the equivalent weight of disodium phosphate required. The 0.90 factor reflects that only 90% of the calcium reacts with the phosphate. Completing this equation we have:

1.5 X 47.3 X 0.9 = 64 ppm disodium phosphate (anhydrous)

This is the amount of phosphate required to react with 30 ppm calcium ion. 64 ppm disodium phosphate divided by 120 equals 0.533 lbs disodium phosphate per 1000 gallons of treated water.

If we compare the calcium content expressed as calcium carbonate (75 ppm) with the amount of disodium phosphate required (64 ppm), we see that 0.85 ppm disodium phosphate is required to react with each ppm calcium expressed as calcium carbonate. Or, if we express calcium as the ion, about 2.13 ppm phosphate is required to react with each ppm of calcium.

Boiler Water Dispersants and Sludge Conditioners

Boiler phosphate and chelant treatment programs are frequently supplemented with polymer dispersants to help condition boiler sludge for removal by routine blowdown. These sludge conditioners are designed to keep the insoluble materials nonadherent, fluid, and free flowing so they do not “bake on” to heat transfer surfaces.

Many types of boiler sludge conditioners have found application in water treatment. Early products were based on naturally occurring compounds such as tannins, lignins, starch and carboxymethylcellulose (CMC). Tannins and lignins are extracts from bark and wood pulp that have demonstrated effectiveness as boiler sludge dispersants and chelants. These materials are no longer in common use having been replaced by synthetic boiler polymers that are chemically better defined.

Polyacrylate is the first synthetic polymer used in water treatment. It functions primarily as a dispersant for calcium and magnesium sludge, but is not as effective in dispersing iron as other synthetic polymers. Typical dosages are 5 to 20 ppm as active polymer.

Polymethacrylate is a close cousin to polyacrylate, but is better for iron sludge dispersion. Typical dosages are 5 to 20 ppm.

Sulfonated styrene/maleic copolymers are recent additions to the boiler polymer lineup. These products are better dispersants for iron than polyacrylate or polymethacrylate. Dosages fall into the 5 to 10 ppm range.

Sulfonated polymers are effective at lower dosages of 1 to 10 ppm. This dispersant is also effective in transporting iron through the boiler and may be used up to 1500 psig.

Other synthetic boiler polymers have been developed that claim superior performance under specific operating conditions. These include

  • phosphino/carboxylic copolymers

  • sulfonated styrene/acrylic/maleic terpolymers

  • acrylic acid/methylpropane/sulfonic acid (AMPS)

Boiler water dispersants that are used in the treatment of steam systems where the steam comes into contact with food or food products are regulated by the FDA. Boiler water dispersants approved for such use under FDA 21 CFR Sec 173.310 include:

  • Acrylic acid/2-acrylamido-2-methyl propane sulfonic acid copolymer (AMPS)

  • Poly (acrylic acid-co-hydrophosphite), sodium salt

  • Polymaleic acid

  • Sodium polyacrylate

  • Sodium polymethacrylate

  • Sodium carboxy-methylcellulose

  • Tannin

  • Sodium lignosulfonate

  • Lignosulfonic acid

  • Sodium alginate

  • Ammonium alginate

  • Sodium humate

  • Sodium acetate

  • Acrylamide-sodium acrylate resin

Typical polymer dosages are between 5 and 25 ppm as 100% active polymer. Boiler polymers are marketed as dilute solutions of these active ingredients, however, so typical product dosages are between 100 and 500 ppm.

Since boiler polymers are non-volatile, they concentrate in the boiler. Chemical tests are available for estimating the polymer residual, but the tests are difficult to perform and the accuracy is not very good. For these reasons, the dosage of polymeric sludge conditioners is frequently determined by direct calculation from steam production rates and cycles of concentration. The dosage of the polymer product is adjusted by regulating the output of the chemical pump or by boiler blowdown.


Corrosion is a reaction between a metal and its environment. In an operating boiler, several corrosion mechanisms exist that can result in rapid attack of boiler steel and significantly reduce the useful life of steam generating equipment. These include oxygen-pitting, caustic embrittlement, and hydrogen embrittlement.

Oxygen pitting attack is caused by dissolved oxygen in the boiler feedwater. This problem is readily identified by the numerous small pits that form on the metal surface. Prevention of oxygen pitting is accomplished by complete removal of the dissolved oxygen by mechanical deaeration and chemical oxygen scavengers. Common oxygen scavengers include reducing agents such as sodium sulfite and hydrazine.

The dissolved oxygen content of feedwater can be estimated from the temperature of the water and the working pressure of the deaerator. The following table presents the dissolved oxygen concentration at a given feedwater temperature and pressure.

Maximum Expected Dissolved Oxygen Level

At Listed Deaerator Temperatures and Pressures

Deaerator Working Pressure, psig

Dissolved Oxygen







































































Caustic embrittlement is caused by the concentration of caustic soda at a crevice or leak where boiler water can flash to steam leaving concentrated boiler water salts behind. If the metal is stressed, the caustic soda will cause cracking of the steel. In a metallurgical exam, the cracking is seen as between the metal grains. Caustic embrittlement is actually a misnomer. It is more accurately characterized as caustic induced stress corrosion cracking of steel.

Caustic embrittlement can be effectively controlled by sealing leaks and welding tube joints, eliminating free OH alkalinity such as in congruent phosphate programs, and in some cases adding sodium nitrate as an inhibitor.

Hydrogen embrittlement is caused by the presence of hydrogen that results from the corrosion of iron.

Fe + 2H+ —– Fe+2 + H2

The hydrogen from this reaction may penetrate steel and react with the carbon to produce methane (CH4).

Although the hydrogen can diffuse through the steel microstructure, the methane cannot. The trapped methane exerts pressure on the grain boundaries and weakens the metal. These are embrittlement failures commonly referred to as hydrogen damage or hydrogen embrittlement.

Oxygen Scavenger Dosage Calculations

Once the oxygen content of the feedwater is determined, it is possible to calculate the scavenger dosage. The sulfite dosage is the sum of the ppm needed to neutralize the dissolved oxygen plus additional amounts needed to produce acceptable boiler water residuals. For low to moderate pressures, sulfite residuals range from 20 to 50 ppm. The excess required depends on the residual desired in the boiler and the number of feedwater concentrations maintained in the boiler.

The theoretical dosage of sodium sulfite (100% purity) is 8 ppm sulfite for every 1 ppm (0.7 cc per liter) of dissolved oxygen. However, correction must be made for the activity or purity of the commercial sulfite, which is about 90%, and for the efficiency of the scavenging reaction. From a practical viewpoint, the sulfite dosage is 10 ppm per ppm of dissolved oxygen in the feedwater. Additional sulfite must then be added to produce the required sulfite residual.

Sodium Sulfite Required to Produce Residuals

of 30, 20, 10, or 5 ppm in the Boiler

(dosages in lbs per 1000 gallons)

Boiler Cycles

30 ppm

20 ppm

10 ppm

5 ppm



















































From Water Treatment for Industrial and Other Uses by Eskel Nordel, 2nd Ed. Page 266

The theoretical dosage of hydrazine (100% active) is 1 ppm per ppm of dissolved oxygen. Because of the explosive nature of pure hydrazine, it is available for industrial use in 35% active solutions. It, therefore, takes 3 ppm of 35% active hydrazine to neutralize 1 ppm of dissolved oxygen. In addition, a low residual of from 1 to 3 ppm is required in the boiler to accelerate the reduction of ferrous oxides.


High temperature hot water systems (HTHW) differ from steam boilers in that these are essentially closed systems. Hot water is circulated from the generator through the areas of heat exchange and then back to the generator. Water losses are minimal unless through unintentional leaks. Nevertheless, high temperature hot water generators are subject to corrosion damage and deposition unless steps are taken to implement an effective water treatment program.

Makeup Water Quality

The quality of the water used to fill a hot water system has an affect on the performance of the corrosion control program. Hard water poses a scaling problem in these systems. For hot water systems, the water should be softened if it exceeds 300 ppm total hardness. For high temperature hot water (HTHW) loops, consider softening the makeup water if it exceeds 10 ppm.

The natural alkalinity in raw water presents further problems in hot water systems. The bicarbonates decompose to hydroxide, which in turn elevate the pH. This promotes caustic embrittlement of steel. Use demineralized or dealkalized water if the total alkalinity exceeds 400 ppm in hot water systems, or if it is greater than 15 ppm in HTHW systems.

Overall, the water used in a hot water loop should be of the best quality available. As a general rule, demineralized is better than soft water which is better than hard water. Some engineers resist the use of demineralized water because of the notion that it is more “aggressive” or corrosive than soft water. Although it is true that untreated, aerated demineralized water is very corrosive, particularly at hot water temperatures, corrosion inhibitors such as sodium nitrite and sodium sulfite passivate metal surfaces and remove dissolved oxygen resulting in a final product water that is noncorrosive.

If the closed system experiences significant water losses, the water should be deaerated to prevent oxygen ingress into the system. Fresh makeup water contains up to 8 ppm dissolved oxygen. If oxygenated water is continuously added to a closed system, particularly at elevated temperatures, the result will be oxygen-pitting attack on mild steel. In this situation, maintenance of sufficient sulfite or nitrite residuals are required to protect the system from corrosion damage.

Chemical Treatment Options

Several chemical treatment methods have been developed for closed loop systems. The selection of one treatment method over another is determined by the water quality, system metallurgy, and any environmental or safety issues at hand.

If the heating system is new, it should be chemically cleaned prior to the start of the treatment program. Chemical cleaning removes oil, mill scale, dirt, welding fluxes and other contaminates that can interfere with the performance of the treatment program. Chemical cleaning is also recommended for older systems that have suffered from corrosion. After the system is clean, apply one of the following treatment programs.

Sodium chromate has been used successfully for years in hot water systems. Alkaline sodium chromate is an oxidizing agent and functions by forming a dense gamma oxide film on mild steel. A minimum of 300 ppm as sodium chromate is required for system protection. High temperature hot water systems and diesel engine cooling water loops require higher dosages of from 2000 to 2500 ppm.

Chromates are toxic substances and a suspected carcinogen. Because of this toxicity, the discharge of chromate has been restricted by the EPA. Chromate is also known to cause dermatitis in workers who come into prolonged contact with this chemical.

Borate-nitrite formulations provide equivalent corrosion protection to chromate. The sodium tetraborate creates a buffer in the system that stabilizes the pH between 9.0 and 9.5. A minimum of 500 ppm of sodum nitrite is required for corrosion protection of mild steel. A 1000 ppm residual as sodium nitrite is recommended in hot water systems. And for those waters that are high in chlorides and sulfates, 1500 ppm of sodium nitrite is required. A general recommendation for inhibitor levels is 800 to 1200 ppm as sodium nitrite.

Like chromates, high concentrations of nitrites are thought to attack some pump seals. The exact mechanism for this attack is unclear. The same seal may function very well in one system and poortly in another, leading one to conclude that other factors like suspended solids levels may play a greater role than inhibitor concentrations.

Borate-nitrite powders have a low solubility in water (about 3% by weight). Pre-mixed liquid formulations, although more costly, are often used for this reason.

Sodium sulfite – caustic soda programs are used in many hot water systems. The sulfite residual should be maintained between 30 and 60 ppm with sufficient caustic soda added to adjust the pH to within 9.3 to 9.5. This is an effective approach when properly applied. It is less expensive than other options and presents few disposal problems.

If the system suffers from air in leakage, the sulfite will be consumed at a rapid rate. Continued addition of more sulfite will cause the dissolved solids in the water to increase significantly.

The use of caustic soda for pH adjustment causes the water to be poorly buffered. Overfeed of caustic increases the pH above the desired 9.0 to 9.5 range. Draining the system or treatment with sulfuric acid is then required to bring the pH within range.

Hydrazine – morpholine is an all-volatile treatment approach that is very effective in high temperature hot water systems. This is particularly true where increased levels of dissolved solids pose a potential deposition problem.

Hydrazine reacts with dissolved oxygen to promote the formation of a dense, corrosion resistant magnetic iron oxide (magnetite) film on steel surfaces. Sufficient morpholine is added to adjust and maintain the pH between 9.0 and 9.5. Generally, a 50 to 200 ppm residual of hydrazine is maintained in the system to guard against oxygen ingress and to promote the maintenance of the magnetite film.

On the negative side, the pH of the water is poorly buffered by the morpholine, so overfeed situations can lead to pH’s above 9.5. Hydrazine also partially decomposes to form ammonia, which can cause accelerated corrosion of copper and other yellow metals.

Hydrazine has recently come under scrutiny as a possible carcinogen. Although it is not banned from use, many plants are seeking safer alternatives to the use of this oxygen scavenger.

Molybdates are used alone or in combination with other inhibitors in hot water systems. Molybdates are often referred to as “chromate substitutes”, since they function in a manner similar to this classic anodic inhibitor. In truth, molybdates are much weaker oxidants than chromate.

A minimum of 100 to 200 ppm of molybdate as MoO4 is required for corrosion protection. Higher dosages are required in more aggressive waters. The pH of the system should be maintained above 7.5. Enhanced protection of yellow metals can be obtained by blending molybdate with tolytriazole. Often molybdate is used in combination with nitrite to afford better protection at lower molybdate concentrations.

Molybdates are generally accepted as being less toxic than chromate. However, the EPA is looking more closely at the environmental impact of molybdates. This may eventually lead to more stringent limitations on the use and discharge of molybdate inhibitors.

Overall, proper water treatment practice is required to prevent water-related problems in steam and hot water systems. These problems include scale deposition and corrosion of boiler metal. In addition to the selection and maintenance of makeup water pretreatment schemes, the application and control of an effective chemical treatment program is required to insure the continuous, reliable and safe operation of the plant equipment.

Reverse Osmosis

In 1959 Reid and Breton first demonstrated the ability of cellulose acetate membranes to separate dissolved salts from solution. Ever since, membrane separation technology has been expanding by leaps and bounds. Today, numerous commercial applications for membrane separations exist including high-purity water production, boiler feedwater, food and beverage processing, drinking water, waste removal and seawater desalination.

In this discussion on the fundamentals of reverse osmosis we will highlight the following topics:

  • Deionization process

  • Reverse osmosis – how does it work?

  • RO system design and operation

  • Pretreatment requirements

  • RO cleaning methods

  • Troubleshooting hints


As the universal solvent, water contains many impurities. These include substances that are dissolved or suspended, plus various dissolved gases.

Dissolved solids, as the name implies, are those impurities that are soluble in water. As water percolates through the earth’s crust, it dissolves the minerals from the local geology. These include calcium and magnesium (hardness), carbonate and bicarbonate alkalinity, and iron and manganese. Other impurities often found dissolved in water are sodium, potassium and barium salts of chloride and sulfate. Depending on the minerals found in the area, silica may also be present in significant quantity.

Suspended solids are those impurities that are not dissolved, but are carried in the water as filterable impurities. Some of these are visible to the eye, but others are very small particulate matter that can only be seen under magnification. These include sand and sediment, clays and colloidal material, oils and greases, microorganisms, and process contaminants.

Dissolved gases are also present in natural water supplies. These impurities are not visible, but are often detected by smell or taste. Common dissolved gases include carbon dioxide, oxygen and nitrogen. Carbon dioxide is common in well waters that are high in bicarbonate alkalinity. Depending on the surrounding conditions, other gases such as hydrogen sulfide, methane and ammonia may be present. Hydrogen sulfide is produced by bacteria. It imparts a rotten-egg smell even at low concentration. Methane may be present as the gas seeps into the aquifer from underground pockets of natural gas.

The goal of water treatment is to improve the quality of water by removing some or all of these dissolved and suspended impurities. Most of the suspended solids, for example, can be removed by filtration. Softening by ion exchange removes calcium and magnesium hardness along with most of the iron and manganese. Dealkalization decreases the carbonate and bicarbonate alkalinity. And deaeration is effective in reducing the concentration of dissolved gases.

Any water treatment process that removes essentially all of the dissolved and suspended solids is called deionization or demineralization. The goal of this process is to produce a water supply of exceptional purity. Two common methods for accomplishing this goal are ion exchange and reverse osmosis. In this discussion, we will review how reverse osmosis is used to demineralize or deionize water.


A membrane functions much like a filter, allowing water to pass through the membrane pores while preventing the passage of dissolved and suspended solids. Unlike traditional depth filtration where 100% of the flow is passed through the filter media to strain out impurities, membrane separation utilizes the principle of crossflow filtration. In this case, as feedwater flows over the membrane surface, a portion of the water permeates through the membrane with the remainder of the flow carried to waste. This separates the feedwater into two streams: a purified product water and a concentrated waste or brine. The water flow “sweeps” the membrane surface clean to prevent the accumulation of suspended and dissolved solids that would eventually block the flow of water through the membrane. The driving force for reverse osmosis is the pressure differential across the membrane. In this way, reverse osmosis can be thought of as the ultimate pressure filter as it removes chemical species as small as an ion.

If a concentrated salt solution is separated from a dilute salt solution by a semipermeable membrane (a membrane that allows water to pass through, but not dissolved salts), water will pass from the dilute solution through the membrane into the concentrated solution. This is a natural process called osmosis. As osmosis continues, the water flow will cause the level of the concentrated solution to rise. The height differential between the two columns represents the osmotic pressure. Flow will continue until the pressure exerted by the height of the concentrated solution equals the osmotic pressure. At this point the system is at equilibrium.

Osmosis can be reversed by applying pressure to the concentrated solution forcing water to flow from the concentrated solution through the membrane into the dilute solution; hence the name reverse osmosis. Since the salts in the concentrated solution can not pass through the membrane, this process results in the production of a purified water stream.

Membranes are classified based on the physical size or molecular weight of the substances that are filtered out by the membrane. Four basic membrane filtration systems exist — microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse osmosis (RO).





Revese Osmosis

Particle size, nm

100 to 1000

1 to 100




Suspended solids and large colloids

Proteins and large organics

Organics and dissolved solids

Dissolved salts and organics

Molecular weight cut off

> 100,000

1,000 to 100,000

200 to 400


Operating pressure

10 psig

10 to 100 psig

50 to 225 psig

200 to 800 psig


A typical reverse osmosis membrane consists of a dense surface skin and a porous substructure. Salt rejection occurs at the surface skin layer with the permeate passing into the porous sublayer.

Two basic types of membranes are in commercial use – asymmetric and thin-film composit. Asymmetric membranes are formed by using the same polymer for the dense surface skin and the porous sublayer. Cellulose acetate, cellulose triacetate, and polyamide are common polymers used in asymmetric membrane manufacture.

In thin-film composite membranes the surface skin and microporous sublayer are formed from two different polymers. Commonly, aromatic polyamide is used for the surface skin with a graded polysulfone resin used in the sublayer.

Each membrane type offers certain advantages and disadvantages for the end user. Cellulose acetate membranes are lower in cost and resistant to chlorine attack. However, the membrane tends to chemically degrade (hydrolyze) outside a pH range of 5 to 8. They are also susceptible to biological degradation and, therefore, require chlorine addition to the feedwater to control bacteria growth.

Aromatic polyamide membranes offer hydrolytic stability, better salt and organic rejection and are not biodegradable. But they are higher in cost than cellulose acetate membranes and have no tolerance for chlorine.

Other membrane polymers are available, each with their own set of advantages and disadvantages. In general, thin-film polyamide composite membranes (TFC) offer certain performance advantages over cellulose acetate membranes (CA). TFC membranes have half the salt passage, higher flux rates and require half the net driving pressure of CA membranes, but they come at a higher cost.

TFC versus CA Membranes

Cellulose Acetate

Thin Film Composite

410 to 600 psi

150 to 500 psi

0 to 30 oC

0 to 45 oC

pH 4 to 6.5

pH 2 to 11

Flux 5 to 18 gfd

Flux 10 to 205 gfd

70 to 95% rejection

97 to 99.5% rejection

Chlorine stable

Chlorine intolerant

Lower cost

Higher cost

There are seven US manufacturers of reverse osmosis membranes and one manufacturer of electrodialysis membranes. These companies have engineered the various membrane materials into unique reverse osmosis element designs. The four basic types of RO element design are (1) tubular, (2) plate-and-frame, (3) spiral wound, and (4) hollow fiber. Tubular and plate-and-frame designs represent higher initial cost and lower membrane surface area per unit. As a consequence, spiral wound and hollow fiber elements dominate the water treatment marketplace.

In a spiral wound membrane design, two layers of membrane material are glued to a permeate collector fabric. Plastic mesh is used to form a feedwater channel between the membrane layers. These layers of membrane, permeate collector and feedwater spacer are rolled around a hollow, perforated center tube that collects the product water.

The spiral wound membrane module is inserted into a pressure vessel housing. Several membrane elements can be linked together inside a single pressure vessel. High pressure feedwater is directed into the end of the element. The permeate is collected in the permeate channel and flows toward the center tube. The concentrated brine exits at the other end of the element.

The hollow fiber membrane module (also called a permeator) incorporates a bundle of hollow fiber membranes in a single element. Feedwater enter the inside diameter of the hollow fiber with the permeate collected in the perforated center tube. Alternatively, hollow fiber modules are designed with ouside-in permeate flow. Feedwater that is not recovered as produce exits as concentrate.

Hollow fiber elements offer an advantage over the spiral wound design in that they pack a large membrane surface area in a single element. The spiral wound membrane type is used more often than the hollow fiber design because of lower operating pressures and energy requirements. Because of this, hollow fiber membranes have been used sparingly since the 1980’s except for seawater desalination projects.


The operating characteristics of the various membrane materials are best defined by the percent rejection of salts at the membrane surface, the flux rate, percent recovery, differential pressure, net driving pressure, and normalized permeate flow. These parameters are also useful for monitoring the performance of RO systems in service and for scheduling routine cleaning.

Percent rejection defines the ability of the membrane to remove dissolved salt from solution. An ideal membrane would allow only water to pass through it, rejecting 100% of the dissolved salts and organic molecules at the membrane surface. In actual practice, RO membranes are not perfect and allow some passage of salts and low molecular weight molecules through the membrane. Typically, 1 to 2% of the salts in the feedwater will pass through. This is equivalent to 98 to 99% salt rejection. A 1.5% rejection rate is considered very good. The passage of low molecular weight organics can be significantly higher, up to 75% in some cases. On average, expect a 95 to 99.5% rejection rate for most inorganic salts.

Flux rate refers to the production capability of the membrane; that is, the amount of product water produced per square foot of membrane surface area per day (gfd). The membrane flux rate increases with increasing driving pressure and feedwater temperature. Great advances have been made in increasing the flux rate. In 1970, a typical production rate from a 4-inch membrane module was 375 gallons per day (gpd) at 600 psig with a 97% rejection rate. Twenty years later, production rates have increased to 1800 gpd at 225 psig with a 97% rejection rate. Typical flux rates for cellulose acetate membranes are 5 to 18 gfd and 10 to 205 gfd for thin film composite membranes.

Membrane flux rates are measured at a standard temperature of

77 oF. Lower feed water temperatures reduce the flux rate. Warmer water increases the flux rate. For every 1 oF change in temperature, the flux rate increases or decreases by 1.5% of rated capacity.

Percent recovery indicates the amount of feedwater that is recovered as treated product water(permeate) With cross-flow filtration some of the feedwater is lost as concentrated brine. Typically, 75 to 80% of the feedwater is recovered as permeate with 20 to 25% being rejected to waste as brine.

Differential pressure measures the pressure drop between the feedwater and the concentrate (brine). Fouling or scale formation in the feedwater flow channels will cause an increase in differential pressure across the RO system. Routine monitoring of the pressure differential helps determine the degree of buildup.

Net driving pressure indicates the pressure drop between the feedwater pressure and the permeate pressure. This is the pressure differential across the membrane. An increase in net driving pressure indicates trans-membrane fouling.

Normalized permeate flow corrects the measured permeate flow back to standard operating conditions at 77 oF. Warmer feedwater temperatures produce a higher flux rate, and colder feedwater reduces the flux rate. Normalized permeate flow calculations compensate for these differences in feedwater temperature by applying a temperature correction factor (supplied by the membrane manufacturer) and changes in net driving pressure to the measured permeate flow. This permits an apples-to-apples comparison between permeate flow rates under varying conditions.

NPF = (NDPnew/NDPnow) x TCF77 X Fp

NPF = normalized permeate flow

NDPnew = net driving pressure when new

NDPnow = net driving pressure when tested

TCF77 = temperature correction factor

Fp = permeate flow in gpm

A reduction in normalized permeate flow indicates fouling of the membrane surfaces.


Both spiral wound and hollow fiber membranes are susceptible to fouling. Feedwater impurities and over concentration of salts in the brine stream promote the accumulation of scale and foulants on the membrane surface. This causes reduced flux and higher operating pressures. For these reasons, pretreatment of the RO feedwater is required to maintain the operating efficiency of the system and prolong the useful life of the membranes.

A water analysis, Langelier Saturation Index (LSI), and Silt Density Index (SDI) are used to determine the precise pretreatment requirements for a particular RO system. Since water supplies vary considerably from one location to another, each pretreatment requirement will be different. The following pretreatment methods are commonly used in RO systems.

Carbon filtration is used to remove chlorine and organic molecules from the RO feedwater. RO systems that are made with cellulose acetate membranes are susceptible to attack by bacteria that feed on the cellulose acetate material. To minimize this problem, CA membranes require 0.3 to 1.0 ppm chlorine residual in the feedwater for bacteria control. Thin-film composite membranes, however, are intolerant of chlorine. Systems that use TFC membranes are designed with pretreatment provisions that remove chlorine and other oxidizing agents like ozone or permanganate. Carbon filtration accomplishes this task easily.

Carbon filters have an infinite capacity for chlorine and produce a final effluent of less than 5 ppb total chlorine. Activated carbon is also capable of removing total organic carbon (TOC) and silt, which cause fouling problems on the membrane surface.

Several grades of activated carbon are available. The carbon media should be made from higher rank coals like sub bituminous or bituminous instead of lignite-based coal. Lignite coals do not offer the same abrasion resistance as the bituminous coals and are more likely to produce carbon particles in the filter effluent. These carbon fines, if they get into the RO module, will cause fouling and a loss of flux. When specifying activated carbon, ask for acid washed carbon media made from bituminous coal having an Iodine Number of at least 800.

Multimedia filtration is used to reduce the suspended solids and colloidal material in the feedwater. These filters contain at least four types of media. The light coarse material is on top of the filter bed followed by progressively less coarse material in the lower layers. This design permits the unit to function as a depth filter by removing the filterable solids throughout the depth of the bed rather than just on the surface of the media. A properly designed multimedia filter will have a flow rate of not more than 7 gpm per square foot of surface area and will have a particle size cutoff of 10 microns.

At times it is necessary to inject a polymer coagulant upstream of the multimedia filter to enhance the removal of the colloidal material. These polymers are high molecular weight materials, normally cationic (positively charged), which coagulate the fine colloids and suspended solids into larger particles that are readily removed in the filter.

Softening the RO feedwater by ion exchange is a popular method for reducing mineral scale formation on the membrane surface. Sodium softening exchanges sodium for scale-forming ions such as calcium, magnesium, barium, strontium, iron and aluminum. Sodium forms very soluble salts that do not form mineral scales on the membrane surface or feedwater flow channels.

A sodium softener is regenerated with sodium chloride brine. The spent regenerant along with the softener rinse water is discharged to waste. This, along with the concentrated brine stream from the RO, can add significantly to the total waste flow from the RO system. For this reason, ion exchange softening should be considered only on very high hardness feedwater or those waters containing appreciable amounts of barium or strontium.

Cartridge filters are installed ahead of the RO high pressure pumps to remove any last traces of suspended solids or biomass that pass through the multimedia or carbon filters. A 5 micron absolute-type filter consisting of a spun filament depth filter with a pore size gradient is generally recommended for this type of application.

Dechlorination, using a chemical reducing agent, is sometimes required to remove the last traces of chlorine or other oxidizing agents prior to the RO membranes. Sodium bisulfite or sodium metabisulfite is used for this purpose. Sodium bisulfite reacts with chlorine to produce sodium sulfate, which is rejected by the RO membrane into the concentrated waste stream.

Acid injection may be required to control the pH and minimize the scale-forming tendency of the feedwater. Acid injection is indicated if the Langelier Saturation Index (LSI) of the brine stream is above +3.0. Either sulfuric or hydrochloric acid is used for this purpose. Sulfuric acid is less costly, however, and therefore more commonly used.

Antiscalants are sometimes effective in extending the intervals between chemical cleanings of the RO membranes. These products are formulated to include inorganic phosphates, organophosphonates, and dispersants. Use antiscalant products that are approved by the membrane manufacturer and follow all directions in applying and controlling the product dosage.

Some antiscalants contain negatively-charged polymers and dispersants that react with cationic polymers that might be injected upstream prior to the multimedia filters. The antiscalant must be compatible with these polymers to avoid any adverse reactions that would foul the membrane.


As a membrane system continues to operate, the dissolved and suspended solids in the feedwater accumulate along the membrane surface or in the feedwater flow channels. As these solids build up, they restrict the passage of water through the membrane. This causes an increase in driving pressure or a loss of flux.

In the early days of RO operation, little was known about the impurities that cause membrane fouling. Today, many of these impurities have been identified. Autopsies of failed membrane modules have revealed a build up of mineral scales like calcium carbonate, colloidal materials like clays and silica, dead and living microorganisms, carbon particles, and chemical attack by oxidizing agents such as chlorine, ozone or permanganate. Likewise, dissolved metals like iron and aluminum, whether naturally occurring or added as a coagulant, will cause premature fouling and failure of the membrane.

Despite all efforts to protect the RO system from fouling and loss of flux, eventually the membranes will require chemical cleaning. Continuous monitoring of the RO operating parameters is required to pinpoint when clean is required. The following guidelines are useful in determining the best time to clean

  • When the pressure differential increases by 15%

  • When normalized permeate flow decreases by 15%

  • When the salt rejection rate decreases by 15%

Under ideal conditions, assuming the RO pretreatment system is properly designed and operated, the frequency between membrane cleanings should be 6 months or longer. Cleaning every 1 to 3 months is considered a fair performance and suggests that some improvements in the pretreatment systems are required. Cleaning frequencies every month or more indicate a change in raw water quality, a problem with the pretreatment system or in the operation of the RO unit.

A well-designed RO system will include provisions for a cleaning skid. The skid should include a chemical tank, solution heater, recirculating pumps, drains, hoses and all other connections and fittings required to accomplish a complete chemical cleaning. Specifications and drawings for a cleaning skid are available from the equipment manufacturer or membrane supplier.

Fouling impurities fall into one of two categories: inorganic deposits and organic foulants. Inorganic deposits include calcium salts, metal oxides, colloids and silica. Organic foulants include biofilms and organic molecules. The chemical nature of the foulant determines the best cleaning chemicals for the job. Acid cleaners are used to remove inorganic deposits and alkaline cleaners are required for organic deposits.

The cleaning method is best determined by an assessment of the chemical nature of the deposits. An autopsy of a sacrificial membrane is required the first time around to verify the foulant. The most effective cleaning chemicals are selected based on the results of the autopsy analysis.

Membrane Cleaning Chemicals

Cleaning Chemical


0.1% (W) NaOH

0.1% (W) Na-EDTA

pH 12

30 oC max

Best for biofilms

OK for silica and organics

0.1% (W) NaOH

0.1% (W) Na-DSS

pH 12

30 oC max

Good on biofilms, organics and inorganic colloids

1.0% (W) STP

1.0% TSP

1.0% (W) NaEDTA

Good for biofilms and organics

0.5% (V) HCl

Best for inorganic salts

0.5% (V) H3PO4

Good for metal oxides

OK for inorganic salts

2.0% (W) Citric acid

OK for inorganic salts

0.2% (W) NH2SO3H

OKk for inorganic salts and inorganic colloids

2 to 4% (W) Na2S2O4

Good for metal oxides

2.4% (W) Citric acid

2.4% (W) NH4F-HF

pH 12

30 oC max

Best for silica

OK for inorganic salts and colloids

(W) indicates percent by weight (V) indicates percent by volume

NaOH sodium hydroxide

NaEDTA sodium ethylene diamine tetraacetic acid

NaDSS sodium dodecylsulfate

STP sodium triphosphate

TSP trisodium phosphate (Na3PO4– 12H2O)

HCl hydrochloric acid

H3PO4 phosphoric acid

C3H3(OH)(CO2H)3 citric acid

NH2SO3H sulfamic acid

Na2S2O4 sodium hydrosulfite

Each cleaning procedure is unique depending on the nature of the foulant and the system operation. In general, it is best to remove organic foulants first with an alkalinity cleaning (high pH) followed by an acid cleaning (low pH) for inorganic scale and metals removal.

Begin the alkaline cleaning by filling the CIP (clean in place) tank with permeate. Circulate the permeate through the RO reject line and back to the tank at low pressure (40 to 60 psig). Heat the water to approximately 80 to 90 oF and slowly mix the cleaner to the desired strength. Alternate circulating and “soak” cycles at 5 to 30 minute intervals for up to 4 hours, if necessary. Repeat with a fresh solution of cleaner if the membranes are severely fouled.

After the alkaline cleaning is done, flush the system with permeate and neutralize any residual alkaline cleaner with mild acid, if necessary. The pH should be neutral (pH 6 to 8). Refill the CIP tank with fresh permeate and heat to temperature. Mix the acid cleaner to the required concentration and alternate circulating and “soak” cycles at 5 to 30 minutes intervals as described for the alkaline cleaner. Monitor the cleaner pH and add more acid to “sweeten” the solution if the pH drops by 0.5 pH unit. Once the acid cleaning step is complete, flush the system with fresh water and neutralize to pH 6 to 8.

Start the RO and operate normally with the permeate discharged to drain until the conductivity of the permeate returns to normal operating levels. Place the system back into service after the proper water quality is achieved.


Reverse osmosis produces a high purity water stream and a concentrated brine or reject stream. Approximately 75 to 80% of the RO feedwater is recovered as useful permeate, but 20 to 25% of the feedwater ends up as waste. For a 75 gpm system operating 24 hours per day, this is equivalent to 250,000 gallons of brine per week.

Finding a suitable method for disposal of the brine stream is one of the challenges of RO design. The salt content of the brine stream is approximately 4 times that of the feedwater. For most installations, however, the reject is still of sufficient quality to permit discharge directly to the environment by one of the following options.

  • Surface water discharge

  • Deep well injection

  • Spray irrigation

  • Municipal waste water treatment plant

  • Evaporation

  • Drain field and bore holes

In many cases, the water is of sufficient purity that it can be recycled and reused prior to discharge. Typical designs for recovering RO reject include capturing the brine stream in a storage tank for use in washing and rinsing floors and equipment or as backwash water for filters. Reuse is the best option as it conserves water resources and improves the overall efficiency of RO operation.


Reverse osmosis is a useful technology for producing high purity water for many industrial, commercial and residential applications. Proper design, operation and maintenance of the RO system is required to insure minimal problems with membrane fouling, shortened membrane life, and increased operating and maintenance costs. Proper pretreatment is required to prolong the useful life of the membranes. Eventually, however, cleaning is necessary to restore the design flow rates and pressures. If this is done properly, a reverse osmosis system will provide a continuous source of high purity water for 5 to 7 years or more between membrane replacements.

Ion Exchange Softeners and Dealkalizers

Ion exchange resins are used extensively in commercial and industrial water treatment to improve the quality of water prior to its intended use. Water softeners are used in homes, laundries, hospitals and manufacturing plants to remove hardness from water. In washing operations soft water produces a pleasing soap lather, rinses cleanly and saves on detergent. Soft water is routinely used as boiler makeup to minimize scale build up on heat transfer surfaces and in other systems, such as cooling towers, where water hardness may interfere with the operating efficiency of heat transfer equipment.

Ion exchange systems are used in the production of dealkalized or demineralized water. These waters are of higher purity than soft water. They are required in applications such as high pressure boilers, semiconductor and electronics manufacturing, pharmaceuticals and metal finishing.

Ion exchange applications are also found in the treatment of waste waters. In particular, ion exchange columns are used to recover metals from plating operations and in the treatment of cooling tower blowdown prior to discharge or recycle.

This paper discusses the role ion exchange plays in the treatment of water for industrial use. In this discussion we will focus on the following topics

  • Water quality – hardness, alkalinity and pH

  • A brief history of ion exchange

  • Types of ion exchange resins

  • How does the ion exchange process work

  • Regeneration procedures

  • Ion exchange systems

  • Troubleshooting

  • Safety issues


Water impurities exist as dissolved or suspended solids. As water flows over the ground or percolates through the earth’s crust, it picks up impurities. Inorganic minerals like limestone and quartz dissolve into water adding to its dissolved solids content. These are primarily calcium, magnesium and sodium compounds. Other common impurities include iron, manganese and silica. In this way, the quality of a particular water supply is strongly influenced by local mineral deposits.

Water also contains suspended solids. Suspended solids, as the name implies, are not dissolved in water, but exist as insoluble particles and colloids. These include sand, clay, and silt. Surface waters contain higher levels of suspended solids than ground waters, but wells located in sandy areas can contain higher levels of insoluble material.

Ion exchange is used to remove dissolved impurities from water supplies. Two common applications are softening and dealkalization. The softening process removes hardness. Dealkalization is used to remove or reduce alkalinity.

Total Hardness refers to the calcium and magnesium content of the water. By definition, the calcium and magnesium concentration determines the water hardness. The water may also contain sodium, iron and silica, but these substances are not defined as hardness, just the calcium and magnesium.

Total Alkalinity is defined as the amount of bicarbonate (HCO3), carbonate (CO3) and hydroxide (OH) alkalinity in the water. Most ground and surface waters contain bicarbonate alkalinity in equilibrium with carbon dioxide (CO2). Carbon dioxide is a soluble gas. A few waters may have some naturally occurring carbonate alkalinity, in which case, there will be no free carbon dioxide. A water sample can not contain all three forms of alkalinity (bicarbonate, carbonate and hydroxide) at the same time. Other dissolved ions like chloride, sulfate and phosphate do not contribute to total alkalinity.

pH is the measurement of the hydrogen ion concentration of the water. The pH scale runs from 0 to 14 with pH 7 being neutral. pH values below 7 are termed acidic. pH’s above 7 are basic. pH is related to the total alkalinity of the water, but pH and alkalinity are not the same. Generally, the higher the total alkalinity, the higher the pH.

All substances dissolve in water to form ions. The ions have a positive or negative charge. Positively charged ions are called cations. Calcium (Ca ++) and magnesium (Mg ++) hardness are positively charged ions and are, therefore, cations.

Negatively charged ions are called anions. Bicarbonate (HCO3 ), carbonate (CO3 -2), and hydroxide (OH ) alkalinities are all negatively charged, and are therefore, anions.

The sum total of all cations must equal the sum total of all anions to maintain electrical neutrality in the water.

















As early as 1845, H. S. Thompson noted that garden soils had the ability to exchange calcium for ammonia when a solution of liquid manure was poured through them. Later, in the 1850’s, this “base exchange” property of soil was attributed to the presence of zeolites, a class of naturally occurring minerals consisting primarily of silica and alumina oxides.

In 1905, Robert Gans discovered that zeolites could be used to remove calcium and magnesium from water. Several natural and synthetic zeolites were identified and used in commercial water softening equipment. Shortly thereafter, stabilized greensand, another naturally occurring mineral, was shown to be effective in removing calcium, magnesium, iron and manganese from water. Greensand, although not as efficient as zeolite, was more durable and, thus, was the mainstay of ion exchange technology for over 20 years.

In 1934 and 1935, several new ion exchange materials were developed that offered significant performance improvements over greensand. This work culminated in the development of a synthetic ion exchange material made by the sulfonation of a resin produced by the copolymerization of styrene and divinylbenzene. This material rapidly replaced greensand because of its higher exchange capacity, more efficient regenerant consumption and improved hydraulic characteristics.

Today, synthetic ion exchange materials are used almost exclusively for water softening, dealkalization and demineralization. Because of their past history, however, modern synthetic ion exchange resins are still commonly called zeolites or greensand.


Ion exchange resins are small, bead-like particles manufactured with a styrene-divinylbenzene copolymer (DVB) backbone. These gel-type resins are made with an 8% or 10% crosslinking of the copolymer structure. The degree of crosslinking determines the “strength” or durability of the resin.

Four fundamental types of ion exchange resins are used in water treatment. All of the resins use the same DVB chemical backbone. The primary difference between the resin types is in the functionality of the ion exchange sites located on the DVB backbone.

Strong acid cation resins contain a sulfonic acid group on the exchange site. They are capable of removing all cations (positively charged ions) associated with strong and weak acid salts. These resins are used in a wide variety of applications, but are commonly found in sodium ion exchangers (water softeners) used for routine hardness removal, hydrogen dealkalizers, and for cation removal in demineralization systems.

Weak acid cation resins are capable of removing calcium and magnesium hardness associated with alkalinity. Non-carbonate hardness is not removed by weak acid cation resin. The primary advantage of weak acid resins is their higher regeneration efficiency as compared to strong acid resins. Weak acid resins are frequently used ahead of strong acid resins to reduce the cost of producing demineralized water.

Strong base anion resins are capable of removing the anions (negatively charged ions) of strongly and weakly dissociated salts. Used in either the chloride form (regenerated with salt brine) or the hydroxide form (regenerated with caustic soda), strong base resins are used in chloride dealkalizer systems and demineralizer trains to remove or reduce alkalinity and silica.

Anion resins are further classified as either Type I or Type II. Type I and Type II resins differ in the functional groups located at the exchange site. Type II resins have a higher exchange capacity and regeneration efficiency than Type I resins. But Type II resins are not as durable as Type I resins. The regeneration efficiency of Type II resins tends to degrade rather rapidly. Eventually, the operating performance of Type II resins degrade to the point where they match the exchange characteristics of Type I resins.

Weak base anion resins are used to exchange all anions except the weakly dissociated silica anion. Weak base resins regenerate more efficiently than strong base resins. As a result, weak base resins are commonly used where complete silica removal is not required. They are also used in combination with a strong base exchanger to improve the efficiency of multi-bed demineralizers.

Ion Exchange Resin Suppliers

Four manufacturers of ion exchange resins exist in the United States. Each offers an extensive product line consisting of several types and grades of ion exchange materials. Other companies market imported resins, or serve as a distributor of resins that are marketed under their own trade names.



Trade Name

Dow Chemical


Miles, Inc.

Wofatit and Lewatit

Purolite Co.


Resin Tech


Rohm and Haas Co.


Sybron Chemicals, Inc.




Mitsubishi Kasei


Although some physical and chemical differences exist between the various brands of ion exchange resin, these differences are often minor and do not affect the overall performance of the resin in most applications. The following table offers a comparison between the product equivalents marketed by the four major U.S. resin manufacturers.


Cation Resins





IR-120 Plus












Anion Resins


















Ion exchange resins selectively remove cations (positively charged ions) and anions (negatively charged ions) from water by replacing the ion located at the exchange site on the resin for the ion dissolved in the water. A cation resin exchanges cations and an anion resin exchanges anions.

The exchange sites are active molecular sites located along the styrene divinyl benzene backbone of the resin. The functional units for cation resins are sulfonic acid and carboxylic acid. For anion resins, it’s a quaternary amine group. In both cases, the functional sites serve as the source of the exchangeable ions.

For example, a strong acid cation resin in the sodium form (common water softening resin), has a sodium ion (a cation) located at the exchange site. As water flows past the exchange site, the sodium is exchanged for other cations that are dissolved in the water such as calcium and magnesium hardness. Iron, another cation, is also exchanged for sodium at the exchange site. This produces a treated effluent containing no hardness, i.e. soft water.

If the cation resin is in the hydrogen form, meaning that hydrogen (a cation) is located on the exchange site; cations will be exchanged for hydrogen. This is the scenario for a hydrogen dealkalizer system. Calcium, magnesium, iron and sodium are exchanged at this site for hydrogen. The treated effluent has a high concentration of hydrogen ions, but no hardness. Recall from the discussion on pH, water that has a high hydrogen ion content has a low pH. The pH from a hydrogen dealkalizer is below 3.0. Water with a pH below 4.3 contains no carbonate or hydroxide alkalinity. All of the carbonates have reacted with hydrogen to produce carbonic acid and free mineral acidity (FMA). Treated water from a hydrogen dealkalizer is effectively softened and dealkalized, but the water cannot be used without subsequent neutralization of the FMA and corresponding upward pH adjustment as low pH water is corrosive to most metals.

Anion resins are used in either the chloride form or the hydroxide form. Here the dissolved anions are exchanged for either chloride or hydroxide. Strong base anion resins in the hydroxide form exchange hydroxide for strongly dissociated (ionized) anions like sulfate and chloride. Other weakly ionized anions like carbonate, bicarbonate and silica are also exchanged by strong base anion resins. Weak base resins also exchange anions, but are unable to remove weakly ionized anions like silica.



Removes the following ions…..

Strong acid

Removes all cations (positive ions)

Weak acid

Calcium, magnesium and sodium associated with carbonate alkalinity

Strong base

Removes all anions (negative ions) including silica and carbon dioxide

Weak base

Removes chloride and sulfate


Ion exchange resins do not have an infinite capacity for ion exchange. As the ion exchange process continues, all of the exchange sites are used up or exhausted by the dissolved ions in the water. At this point the resin is no longer capable of exchanging ions and must be restored to its original ionic form by regenerating with a strong solution of salt brine, acid or caustic soda.

The ion exchange capacity of the resin is a measure of the amount of dissolved ions that can be exchanged by the resin between regenerations. The exchange capacity is a function of the resin type, amount of regenerant used per cubic foot of resin, and regenerant flow rate. The resultant exchange capacity is expressed in Kilograins per cubic foot of resin (Kgr/ft3), or milliequivalents per gram (meq/gr).

One grain is equal to 1/7000th of a pound. Most water analyses report hardness and alkalinity values in units of parts per million (ppm) or milligrams per liter (mg/l). In this case, concentrations reported in ppm are equal to mg/l, that is, ppm and mg/l can be used interchangeably. Ion exchange manufacturers still work with water analyses expressed in grains per gallon, however, instead of ppm or mg/l. To convert water quality data from ppm or mg/l to grains per gallon, simply divide ppm by 17.1. For example, water having a total hardness of 171 ppm contains 10 grains per gallon of hardness.

Resin manufacturers report the exchange capacity of the ion exchange media in Kilograins per cubic foot of resin. (1 Kilograin is equal to 1000 grains). This provides an estimate of the amount ofcations or anions that can be removed by the resin between regenerations.



Regeneration Level

Exchange Capacity

Strong acid cation

5 lbs Salt per ft3

17.8 Kgr/ft3

Strong acid cation

5 lbs Acid per ft3

12.5 Kgr/ft3

Strong base anion Type I

4 lbs Caustic per ft3

11 Kgr/ft3

Strong base anion Type II

4 lbs Caustic per ft3

21 Kgr/ft3

The suggested regenerant strength, regeneration level and resultant exchange capacity are available from the resin supplier, equipment manufacturer, or water consultant.

Increasing the regeneration level will provide a higher exchange capacity. This is not a linear relationship, however. Doubling the salt dosage does not double the exchange capacity. The most efficient salt dosage for industrial softeners is between 6 and 8 pounds of salt per cubic foot of resin.


Once the exchange capacity of the softener has been exhausted, it is necessary to regenerate the unit to remove the accumulated ions and restore the ion exchange resin back to its original chemical form. Various types of regeneration chemicals are used for this purpose depending on the water treatment requirements. Cation resins are regenerated with sodium chloride (salt), sulfuric acid or hydrochloric acid. Anion resins are regenerated with sodium chloride, caustic soda or ammonium hydroxide.

The regeneration level is a measure of the amount of regenerant required per cubic foot of resin. Typically, strong acid cation units are regenerated with 6 pounds of sodium chloride per cubic foot of resin, or 6 to 8 pounds of sulfuric acid (as 100% acid) per cubic foot.

The regeneration level for anion resin is typically 4 pounds of caustic soda (as 100% caustic) per cubic foot of resin.

These regeneration levels may vary from one installation to another. Since the exchange capacity of the resin increases with increasing regeneration levels, some plants use more acid and caustic per cubic foot to extend the run times of the ion exchange equipment. This is at the expense of efficiency, since the Kilograins of dissolved solids removed per pound of regenerant decreases at higher regeneration levels. In other cases, the resin may be old and require a higher regeneration level to meet the water quality specifications.

The chemical regenerants are mixed with water prior to entering the ion exchange equipment. For softening equipment, the salt is added to a brine tank to produce a saturated brine solution. Saturated brine contains 2.5 pounds of salt per gallon of brine. This is equivalent to approximately 25% to 26% salt. The saturated brine is then pumped or educted into a dilution water flow to produce a minimum 8% brine solution. Under ideal conditions the resin bed should be regenerated with a minimum 8% brine solution for at least 20 minutes. Enough brine is introduced into the ion exchange bed to achieve the required regeneration levels.

Sulfuric acid is used to regenerate strong acid cation units in the hydrogen form. Bulk sulfuric acid (93%) is pre-diluted in a day tank to yield a 20% acid solution. This is then educted or pumped into a dilution water line where it is diluted to the proper concentration prior to entering the cation unit. Normally the regeneration is carried out in a stepwise fashion. First a 2% acid regenerant is introduced to remove most of the calcium from the bed. Then a 4% acid is used to complete the regeneration. This prevents the formation of unwanted calcium sulfate precipitants that can foul the resin. In cases where calcium sulfate precipitation is of particular concern, the acid is introduced in 2%, 4%, and 6% stages. If hydrochloric acid is used instead of sulfuric acid, the stepwise regeneration procedure can be eliminated. Hydrochloric acid is more expensive than sulfuric acid, however. In either case, enough acid must be dosed per cubic foot of resin to achieve the desired regeneration level of 6 to 8 pounds per cubic foot.

Anion resins in the chloride form are regenerated with sodium chloride (salt). Caustic soda is used for anion resins in the hydroxide form. Here bulk liquid caustic soda (50%) is mixed in a day tank to achieve a 20% working solution. The caustic is educted or metered into a dilution water line where it is diluted to a 4% solution prior to entering the anion vessel. Sufficient caustic is used to achieve a regeneration level of about 4 pounds per cubic foot of resin.


Ion exchange equipment is operated in one of three modes

  • Service

  • Regeneration

  • Standby

Service runs vary depending on the feedwater quality, bed volume, and regeneration level. For example, a sodium softener containing 86 cubic feet of ion exchange resin is operated at a regeneration level of 10 pounds of salt per cubic foot of resin. This produces an exchange capacity of 25 Kilograins per cubic foot. The total exchange capacity of the softener is (25 Kgr/ft3 X 86 cubic feet) or 2,150 Kilograins softening capacity.

If the feedwater to this softener has a total hardness of 300 ppm (300 ppm / 17.1) or 17.5 grains per gallon, the total service capacity of the softener is (2,150,000 grains / 17.5 grains per gallon ) = 122,857 gallons between regenerations. For practical purposes, most softeners are not kept in service until completely exhausted. The softener is frequently removed from service at 80 to 90% of full capacity. In this example, 85% of full capacity would be 104,500 gallons.

A similar calculation can be performed on a hydrogen dealkalizer to determine the estimated service run. The total exchange capacity of the bed is calculated based on the acid regeneration level and resultant total exchange capacity of the bed just as with the sodium softener.

Regeneration of ion exchange equipment consists of:

  • Backwashing

  • Chemical injection

  • Slow rinse

  • Fast rinse

Backwashing: Ion exchange resin is an excellent filter media. Suspended solids in the feedwater are readily trapped in the resin bed where they can cause fouling if not removed. The purpose of the backwash step is to remove these unwanted foulants. Backwashing also lifts and expands the resin prior to regeneration to insure optimum contact between the resin and the regenerant chemicals.

The backwash step is carried out from the bottom up, counter to the direction of the service flow. The flow of the backwash water lifts, suspends and expands the resin bed. This backwash flow carries the suspended solids, resin fines and other debris down the drain.

The backwash flow rate is regulated to expand the bed by 50 to 100% of it service volume. If the backwash rate is too high, however, some of the resin will be washed out of the vessel along with the dirt and debris. If the backwash water temperature changes, the flow rate must be adjusted to prevent this from occurring. As the water becomes colder (more dense) the flow rate must be reduced. Warmer water requires higher flow rates. The technical literature supplied with the resin contains detailed information on the required backwash flow rates for that particular type and grade of resin.

Chemical Injection: The regeneration of ion exchange resin is accomplished by the introduction of chemicals that remove the adsorbed ions from the exchange sites, and restores the resin to its original chemical form. Cation resins are restored to the sodium (salt regeneration) form or hydrogen (acid regeneration) form. Anion resins are restored to the chloride or hydroxide form.

The concentrated regenerants in the storage tank are diluted to the proper concentration prior to entering the ion exchange vessel. This must be precisely controlled or the resin will not be regenerated properly. Too low of a regenerant concentration will reduce the exchange capacity of the bed. Too much regenerant wastes chemical and decreases the regeneration efficiency.

During the regeneration step, samples of dilute regenerant are collected to determine the percent concentration. This is easily accomplished by using a hydrometer which measures percent brine or specific gravity. The results of the hydrometer measurements are then used to make adjustments to the acid or brine strength.

Slow rinse: At the conclusion of the chemical injection step, the regenerant flow is stopped and the dilution water flow is used to “push” the remainder of the chemical regenerant through the resin bed. This completes the regeneration and makes certain that all of the regenerant is utilized efficiently.

Fast rinse: After all the regenerant has passed through the resin bed, the unit is rinsed under full flow conditions. This removes any last traces of regenerant and makes sure the water quality meets the finished product specifications. In high-purity applications, like demineralized water systems, the fast rinse continues until the water meets a minimum conductivity standard. In other cases, the fast rinse continues for a preset time period.

Standby Mode is utilized if the softener or dealkalizer is not to be placed into service immediately at the conclusion of the regeneration. If an ion exchange unit sits in standby for a prolonged period (more than a day), the exchange equilibrium tends to reverse at the exchange sites. This can sometimes adversely affect the quality of the product water. For this reason, ion exchange units are placed into a service rinse mode for 5 to 15 minutes, or until the product water meets a minimum standard before it is placed into service. This rinse cycle flushes any unwanted dissolved solids from the resin bed and guarantees that the product water is acceptable for use. If not, the unit is placed into standby mode again until the problem can be corrected.

After a successful service rinse, the softener or dealkalizer is placed back into service.


Design engineers use various combinations of ion exchange units to produce a final product water of any desired quality. Some of the systems are very elaborate and incorporate a series of weak acid, strong acid, weak base, or strong base exchange beds. Most, however, are very simple in concept and design. The more common system designs are for water softening, dealkalization, and demineralization.

Water softeners are used in a wide variety of industrial applications. Here a strong acid cation resin in the sodium form is used to remove calcium and magnesium. The calcium and magnesium hardness is exchanged for sodium, which does not contribute to the water hardness. In this case, the total dissolved solids of the water remains the same, since softening does not reduce the amount of mineral solids in the water.

Softening systems are comprised of an ion exchange vessel ( two or more vessels, if an uninterrupted supply of soft water is required). A salt brine tank is provided, and a control system to regulate the regeneration process. The regeneration can be controlled by a timer that initiates the regeneration at a preset time and day, or by a water meter that starts the regeneration after a preset number of gallons have passed through the softener.

Hydrogen dealkalizers are used to remove hardness and alkalinity from the raw feedwater. Here the hydrogen form of the cation resin is used to convert natural bicarbonate (HCO3) alkalinity to carbonic acid and free mineral acidity (FMA). Once formed, the carbonic acid (H2CO3) readily breaks down to release free carbon dioxide (CO2) and water (H2O), which can be easily removed by simple aeration. Since the strong acid resin replaces all the cations with hydrogen, the effluent from the dealkalizer is both softened and acidic. Generally, the pH of the effluent is less than 3.0.

Water from the hydrogen dealkalizer can be neutralized by blending with effluent from the sodium softener. Any desired alkalinity and pH can be achieved by regulating the percent blend of hydrogen dealkalized and sodium softened water. After blending, the water is passed over an aerating tower (or decarbonator) to remove the free carbon dioxide released in the neutralization step.

Chloride anion dealkalizers are used as an alternative to hydrogen dealkalizers. These units are regenerated with salt brine just like sodium cation softeners. The carbonate and bicarbonate alkalinity is removed by the anion resin and replaced with chloride. Frequently, a sodium cation softener is combined with a chloride-anion dealkalizer to produce a final product water that is softened and dealkalized. The total dissolved solids have not been reduced by this method, but the softened, dealkalized water is suitable for use as makeup in many boiler feedwater applications.

The primary advantage of dealkalization, whether by hydrogen or chloride anion methods, is the reduction of carbonate alkalinity in the boiler feedwater. Under boiler temperatures and pressures, carbonate alkalinity breaks down to release carbon dioxide into the steam. This depresses the pH by forming carbonic acid in the steam condensate. Eliminating the carbonate alkalinity from the boiler feed reduces the CO2 in the steam and helps protect the condensate system from corrosion attack.

Demineralization or deionization achieves complete removal of all ions. The result is high –purity water with extremely low conductivity (high resistivity). This is accomplished by used a strong acid cation unit in the hydrogen form in series with a strong base anion unit in the hydroxide form. The cation unit exchanges hydrogen (H+) for all the cations, producing an effluent that has free mineral acidity (FMA) and a pH of around 3. The effluent from the cation unit is passed through the anion exchanger where the anions are exchanged for hydroxide (OH). The hydroxide (OH) in the anion effluent reacts with the hydrogen (H+) from the cation unit to produce water (H+ + OH = H2O). In this way, all the mineral salts are removed from the feedwater producing a final product water of exceptional purity.

The effluent from the cation unit will contain free carbon dioxide (CO2). Although carbon dioxide can be removed by the strong base anion exchanger, it is more economical to remove this gas in a degasifier. This also extends the service runs on the anion unit. For these reason, a degasifier is often installed between the cation and anion units on waters that are high in carbonate alkalinity.


Ion exchange design engineers take every precaution to guard against operating problems. Nevertheless, problems can and do occur. The more common problems have the following symptoms:

  • Loss of thruput capacity

  • Reduction in water quality

  • Increased pressure drop across the bed

Loss of Thruput Capacity

A loss of capacity is indicated when the amount of water produced between regenerations declines. These shortened service runs are caused by a variety of factors. Here are some things to check when this problem occurs.

  • Did the feedwater quality change? Higher dissolved solids in the feedwater will give shorter runs.

  • Has the bed level decreased? A loss of resin during the backwash cycle or because of a broken underdrain lateral will reduce the exchange capacity of the bed.

  • Was the last regeneration sequence completed properly? Check the regenerant levels in the storage tanks, the regenerant concentrations, flow rates and times.

  • What is the condition of the resin? Resin that has degraded chemically or physically will not produce acceptable water quality. Collect a sample for laboratory analysis.

Reduction in Water Quality

The ion exchange system is unable to produce water of acceptable quality. The reasons for this often baffle the experts, but here are some things to check.

  • Raw water leaking past a valve. Valves that do not seat properly, or leak because of wear and tear will allow raw, untreated water to contaminate the treated effluent. Check each valve for proper operation.

  • Depending on the type of unit, old resin can suffer from ion “leakage”. Check the chemical and physical condition of the resin. Calcium precipitation in the cation unit can also be a problem.

  • Fouling of the resin bed by organics, iron, oil, microbiological growths or dirt. These contaminants will adversely affect the regeneration of the resin and alter the flow through the bed.

  • Poor quality or off-spec regeneration chemicals. Verify that the chemical regenerants meet your specifications.

Increased Pressure Drop Across the Bed

An ion exchange bed exerts a resistance to flow that is measured as a pressure decrease across the vessel. Some pressure drop is a normal condition, but excessive loss of head and reduced flow is not. Here are some things to check.

  • Is there a blockage in the inlet to the vessel?

  • Has the resin degraded? Ion exchange resin can be degraded by high temperatures and chemical oxidation. Chlorine, a strong oxidizing agent, can cause decrosslinking of the resin resulting in swelling and clumping of the resin. These conditions increase the resistance to water flow and hence a loss of head.

  • Fouling of the resin with dirt, iron, calcium deposits, or corrosion by-products will restrict water flow and increase the pressure drop across the unit.

  • Is there a problem with the underdrain system? Normally this is difficult to check without removing the resin and any support media. Check this last.

Many of the problems associates with ion exchange equipment are difficult to diagnose. If the cause of the problem is not readily apparent, consult an expert in the field for further advice and recommendations.


Working with ion exchange systems presents hazards associated with the handling of strong chemicals, ion exchange resins, electrical equipment, confined spaces, and related equipment such as fans, blowers and motors. The best safety practice is to read and become familiar with the instruction manuals that are provided with the ion exchange system. Read and follow all instructions as outlined by the manufacturer. If you have any questions, contact the equipment manufacturer or knowledgeable expert before proceeding.


Ion exchange is an effective method for improving water quality by removing troublesome impurities like hardness and alkalinity. These impurities, if not removed, will cause problems in boilers and heat exchangers such as scale deposition and fouling.

Since the acceptance of synthetic ion exchange resin as a replacement for zeolites and greensand, many types and grades of ion exchange material are used in the design of water treatment plants. The more popular systems include softeners for removing calcium and magnesium hardness’ and dealkalizers for removing alkalinity. These systems exhibit improved properties such as high exchange capacity, efficient regenerant utilization, and long life.

It is clear that ion exchange serves a useful and necessary role in the improvement in water quality for industrial, commercial and residential use.