Reverse Osmosis System.pdf

MRO-4H REVERSE OSMOSIS SYSTEM

INSTALLATION, OPERATION, AND

MAINTENANCE MANUAL

JOB NAME:

MODEL:

P.O. NUMBER:

JOB NUMBER:

Marlo Incorporated: 2227 South Street • Racine, Wisconsin 53404 • 262.681.1300 • FAX 262.681.1318 • www.Marlo-Inc.com • info@ Marlo-Inc.com

Standard Photograph The appearance of your equipment may differ.

MRO-4 REVERSE OSMOSIS SYSTEM

INSTALLATION, OPERATION, AND MAINTENANCE MANUAL

TABLE OF CONTENTS

Section 1 Operation Parameters & Projected Performance Pretreatment Considerations Basics of Chemical Pretreatment Basics of Reverse Osmosis

Section 2 RO Major Component Identification RO Installation

Section 3 Pre-Startup Checklist Start-Up Procedures Control Panel Operation

Section 4 Routine Maintenance Record Keeping Requirements

Section 5 Troubleshooting Membrane Cleaning Procedures

Section 6 Vendor Literature Major Component I/O Manuals

Section 7 System General Arrangement Drawings Electrical Enclosure Drawings Wiring Diagrams

Section 8 Pretreatment

SECTION 1

Operation Parameters & Projected Performance

Pretreatment Considerations

Basics of Chemical Pretreatment

Basics of Reverse Osmosis

DESIGN PARAMETERS

Model: MRO-10800-4H

Membrane Array: 2:2:2:1

Housings:

Size: 4” Diameter x 1 Long Material: FRP Design Pressure: 300 PSIG

Membranes:

Type: TFC Brackish Toray TM710 Size: 4” x 40” Quantity: (1) per Housing

Power Supply Required 460 volts/ 60 hertz/ 3 phase (pump power)

120 volts/ 60 hertz/ 1 phase (control power)

Pump Motor HP 5

RO SYSTEM FEEDWATER CHARACTERISTICS AND QUALITY REQUIREMENTS

Inlet feedwater temperature range: 55-80° F Feedwater free chlorine content: < 0.1 ppm Maximum feedwater turbidity: 1 NTU Maximum Silt Density Index (SDI-15 minute): 4.0 Feedwater pH range: 4 - 11 Maximum RO concentrate LSI: 1.07

RO SYSTEM OPERATIONAL PARAMETERS

Minimum inlet feedwater pressure (at RO inlet valve): 30 PSIG Maximum prefilter housing pressure drop: 20 PSIG Low inlet pressure cut-out switch (factory set): 10 PSIG

Designed inlet feedwater flowrate: 11.5 GPM Designed RO product flowrate: 7.5 GPM Designed RO recirculation flowrate: 4.0 GPM Designed RO reject flowrate to drain: 4.0 GPM

Membrane feed pressure: 200-300 PSIG (Monitored only) Membrane outlet pressure (concentrate): 200-300 PSIG (Monitored only)

Membrane cleaning frequency: At 10% system performance loss, As determined by “normalized” data

RO SYSTEM PROJECTED PERFORMANCE:

Overall system TDS reduction (0 years): Normalized 97% TDS reduction Minimum TDS reduction (0 years): Min. 95% TDS reduction

The RO system has been designed to operate on a carbon filtered, softened water supply that has been sufficiently filtered to reduce the Total Dissolved Solids and particles to achieve the Turbidity and SDI levels recommended above. The feedwater must be free of free chlorine and other oxidizing agents to prevent irreversible damage to the RO membranes.

DESIGN PARAMETERS

Model: MRO-22K-4H

Membrane Array: 3:2

Housings:

Size: 4” Diameter x 3 Long Material: FRP Design Pressure: 300 PSIG

Membranes:

Type: TFC Brackish Toray TM710 Size: 4” x 40” Quantity: (3) per Housing

Power Supply Required 460 volts/ 60 hertz/ 3 phase (pump power)

120 volts/ 60 hertz/ 1 phase (control power)

Pump Motor HP 7.5

RO SYSTEM FEEDWATER CHARACTERISTICS AND QUALITY REQUIREMENTS

Inlet feedwater temperature range: 55-80° F Feedwater free chlorine content: < 0.1 ppm Maximum feedwater turbidity: 1 NTU Maximum Silt Density Index (SDI-15 minute): 4.0 Feedwater pH range: 4 - 11 Maximum RO concentrate LSI: 1.07

RO SYSTEM OPERATIONAL PARAMETERS

Minimum inlet feedwater pressure (at RO inlet valve): 30 PSIG Maximum prefilter housing pressure drop: 20 PSIG Low inlet pressure cut-out switch (factory set): 10 PSIG

Designed inlet feedwater flowrate: 23 GPM Designed RO product flowrate: 15 GPM Designed RO recirculation flowrate: 5 GPM Designed RO reject flowrate to drain: 8 GPM

Membrane feed pressure: 200-300 PSIG (Monitored only) Membrane outlet pressure (concentrate): 200-300 PSIG (Monitored only)

Membrane cleaning frequency: At 10% system performance loss, As determined by “normalized” data

RO SYSTEM PROJECTED PERFORMANCE:

Overall system TDS reduction (0 years): Normalized 97% TDS reduction Minimum TDS reduction (0 years): Min. 95% TDS reduction

The RO system has been designed to operate on a carbon filtered, softened water supply that has been sufficiently filtered to reduce the Total Dissolved Solids and particles to achieve the Turbidity and SDI levels recommended above. The feedwater must be free of free chlorine and other oxidizing agents to prevent irreversible damage to the RO membranes.

REVERSE OSMOSIS SYSTEM

REVERSE OSMOSIS UNIT CHEMICAL FEED RECOMMENDED DOSAGES:

(Dosages may need adjustment under actual operating conditions.)

Pretreatment (if needed)

Sodium Bisulfite – Fed before the RO to remove chlorine from the water because those substances are harmful to the RO elements. For every part of chlorine in the raw water, feed 3 parts of sodium bisulfite Recommended Starting Dosage - 3-5 ppm

Antiscalant – Fed before the RO to reduce the scaling potential of the water on the RO elements.

Recommended Starting Dosage ï 2-3 ppm (this dosage will vary depending on which anti-scalant is used)

Post treatment (if needed)

Caustic – Fed after the RO to achieve the target pH level of 7

Recommended Starting Dosage - 1 ppm

Chlorine– Fed after the RO to achieve the target chlorine residual

Recommended Starting Dosage - 0.5 ppm

NNoottee:: TThheessee ddoossaaggeess aarree ggeenneerraall gguuiiddeelliinneess oonnllyy.. SSppeecciiffiicc ddoossaaggeess aarree ddeeppeennddeenntt oonn rraaww wwaatteerr qquuaalliittyy..

RReeffeerr ttoo tthhee aaccttuuaall cchheemmiiccaall mmaannuuffaaccttuurreerr’’ss rreeccoommmmeennddaattiioonnss ffoorr yyoouurr ssppeecciiffiicc iinnssttaallllaattiioonn..

INTRODUCTION

This manual contains important information about your unit, including information needed for installation, operating, and maintenance procedures. These reverse osmosis systems have been designed and engineered to provide high quality water when operated according to specifications. A thorough understanding of the system and its operation is required for proper application. Review this manual in detail before making any attempt to install or start the system. Abuse or tampering of equipment by inexperienced operators can cause irreversible damage.

Process Description

Reference Drawings are located in section 4. Use the Process and Instrumentation Diagram as guides to your understanding of the processes involved.

Pretreatment Depth Filters (If needed) – Raw water first flows through depth filters for high quality filtration. The depth filters use a multilayer design where the filter bed uses successively smaller media to provide efficient filtration at higher flow rates than single media filters. To maintain clarity, the depth filter requires periodic backwashing to remove the filtered media from the filter bed.

Pretreatment Carbon Filters (If needed) – Raw water flows through carbon filters for chlorine and organics removal. To maintain efficient absorbtion , the carbon filter requires periodic backwashing to unpack the filter bed.

Pretreatment Water Softeners (If needed) – Chlorine free water flows through the water softener to remove hardness which will foul the membranes. The softener is periodically regenerated with salt to replenish the resin’s softening capacity.

Pretreatment Chemical Feed (If needed) – Raw water is chemically treated before it enters the

RO unit. Sodium bisulfite, if required, is fed to remove chlorine from the water because those substances are harmful to the RO elements. Next, antiscalant chemical is fed to reduce the potential the water has to scale the RO elements.

Reverse Osmosis System - The mixture of water and chemicals flows to the reverse osmosis system. The RO system splits the water into two streams. The product stream (permeate) flows to points of use, and the waste stream (concentrate) flows to drain. Control of the RO unit is accomplished with the RO control panel.

Post-treatment Chemical Feed (If needed) – RO permeate water is chemically treated after the

RO unit. Chlorine is fed to prevent bacteria growth and caustic is fed to adjust the product water pH, if required.

PRETREATMENT CONSIDERATIONS FOR REVERSE OSMOSIS

It is imperative that adequate consideration be given to the pretreatment needs of reverse osmosis systems. Some pretreatment schemes are based on membrane considerations, such as dechlorination for PA and TFC membranes or chlorination and pH adjustment for CA membranes. The major pretreatment requirement for any RO is the control of membrane “fouling”. Fouling involves the entrapment of material in the feed/brine path or deposition on the surface of the membrane. There are four categories of fouling. Figure 1 depicts these foulants, possible sources, and possible pretreatment techniques for these foulants.

Scaling by Salts:

Salts in the feedwater are usually concentrated by a factor of two to four times (50 – 75% conversion) in the RO process. Solubility limits of some salts can be exceeded and they precipitate. The scaling components of concern are calcium carbonate, calcium sulfate, silica, strontium sulfate, barium sulfate, and calcium fluoride. Four basic pretreatment design approaches are used to control scale:

1. Reduce RO recovery to where salt concentrations are below solubility limits.

2. Alter the feed temperature to increase solubility limits.

3. Remove one of the ions responsible for the scale-forming compounds.

4. Add an antiscalant which inhibits precipitation of the scaling compound. The proper antiscalant can increase the solubility by a factor of 2 to 5.

Calcium carbonate (line) scaling is a problem with most natural feedwater and is controlled by:

1. Softening which removes the calcium ion.

2. Acid feed which reduces the biocarbonate / carbonate ion level by converting it into carbon dioxide.

3. The use of an antiscalant

The LSI (Langlier Saturation Index) is used to determine the lime-scaling tendency of a water and is dependent on temperature, pH, alkalinity and calcium levels.

Calcium sulfate solubility limits are not usually exceeded when feedwater TDS is under 1,500 ppm and recovery is 80% or less.

The solubility of reactive silica in waters is variable but is about 120 – 150 ppm at 20° C or greater, and 50

– 60 ppm at 10° C. When saturated, the silica can polymerize to form colloidal silica. Higher silica solubility limits have been observed in RO systems but it is difficult to predict due to other factors, such as pH, and the effects of other ions in solution. Evaluation on a case-by-case basis would be required to determine silica scaling potential.

Metal Oxide Fouling:

Iron and manganese, when oxidized, can form insoluble precipitates that can foul a membrane. Iron fouling is the most prevalent. Iron fouling can be avoided by:

1. Removal of the iron from the feedwater before it gets to the RO.

2. Preventing the oxidation of soluble ferrous iron into the insoluble ferric iron.

3. Operating at a lower pH while the iron is still soluble.

4. Adding a sequestering agent to keep the iron in solution.

Iron can be removed by manganese zeolite filtration, aeration followed by filtration, or by sodium zeolite exchange of dissolved iron and manganese. Generally, 4 ppm of soluble iron can be tolerated in the reject water if dissolved oxygen is 5 ppm or less.

Colloidal Fouling:

The coagulation of colloids, as they concentrate in an RO unit, can result in colloid fouling. Typically, colloids range in the 0.2 to 1.0 micron size range and are: clays (aluminated silicates) found in raw waters, iron colloids from the corrosion of carbon steel piping, and aluminum hydroxide resulting from improper alum clarification. The SDI test method determines the fouling tendency of water due to colloids. Normally, well waters have an SDI of less than three and present no colloids fouling problem.

Surface waters normally have SDI’s of greater than 5 and pretreatment must be considered. The SDI can be reduced by such methods as:

1. Multimedia filtration – with or without upstream addition of polyelectrolyes

2. Ultrafiltration

3. Microfiltration

The characterization of colloids in the water, and the effectiveness of pretreatment, is difficult to predict and may require on-site pilot work.

Biological Fouling:

Biological fouling is caused by the growth of microorganisms in the RO. CA membranes are susceptible to biological degradation, so a residual chlorine level of 0.3 to 1.0 ppm is required. PA and TFC membranes are resistant to biological attack. The formation of a bio-film on any membrane can severely retard flux. It is controlled in CA membranes by using residual chlorine and, in PA or TFC membranes, by periodic sodium bisulfite treatments. Preferably, all membranes should utilize adequate pretreatment and correct operations for the RO that inhibit propagation of microorganisms.

Table 1:

RO FOULING

CATEGORY EXAMPLES SOURCES PRETREATMENT

Scaling Salts Calcium Carbonate Feedwater pH Adjustment (Improve Solubility) Calcium Sulfate Antiscalant (Improve Solubility) Barium Sulfate Softening (Remove Calcium Ion) Strontium Sulfate Precipitation (Remove Silica) Calcium Flouride Heat Exchangers Silica

Metal Oxides Iron Feedwater, Oxidation followed by filtration Manganese Corrosion, Use corrosion-proof materials Aluminum Clarifiers

Coloids (SDI) Clay Surface Waters Coagulation / Flocculation Silt (SDI 10 – 175), Clarifiers Rust Corroding Pipes Multimedia Filters Microfiltration Ultrafiltration

Biological Bacteria Non-Chlorinated Feed, Chlorinate (Disinfect) Off-Line RO’s Proper Off-Line Storage Surface Waters

BASICS OF CHEMICAL PRETREATMENT

There are a number of chemicals that can be introduced into the RO feed to enhance the operation of the RO system.

Acids Caustic

Dechlorination Chemicals

Antiscalants and Dispersants

Acids

Acids, typically hydrochloric [HCl] or sulfuric [H2SO4], are fed to lower the feed pH.

Sulfuric acid is used more often than hydrochloric acid, based on a lower operating cost, reduced fuming to the atmosphere which can corrode surrounding metal components, and a better membrane rejection of the sulfate ion than the chloride ion. Technical grade sulfuric acid, with no other additives, is suitable for use with a RO system. Sulfuric acid is commercially available as a 20% and 93% solution.

The 93% solution is also referred to as 66° Baume solution. Caution is required in diluting 93% sulfuric acid, since the maximum heat of dilution of approximately 280°F occurs around 60%. It is critical that the concentrated acid is added slowly to the top of dilution water that is being agitated to minimize the buildup of heat and boiling of the makeup solution.

Hydrochloric acid is preferred when there is a concern for scaling by calcium sulfate, barium sulfate or strontium sulfate. Sulfuric acid increases the sulfate ion level in the RO feed, which directly increases the potential for sulfate-based scaling. Technical grade hydrochloric acid (with no additives) typically is suitable for use with a RO system. Hydrochloric acid is typically available as a 30% to 37% solution. It is sometimes referred to as muriatic acid.

The primary purpose for reducing the feed pH is to reduce the potential of calcium carbonate scaling in the RO concentrate as measured by using the Langlier Saturation Index [LSI]. LSI is a method of reporting the scaling or corrosive potential of low TDS brackish water based on the level of saturation of calcium carbonate [CaCO3]. LSI is important in RO water chemistry in determining whether a w ater will or will not form calcium carbonate scale. Water with a negat ive LSI is considered corrosive to metal piping and will not form calcium carbonate scale. Water with a positive LSI is not corrosive, but it will tend to form calcium carbonate scale.

The LSI value is calculated by subtracting the calculated pH of saturation of calcium carbonate from the actual feed pH. Calcium carbonate solubility decreases with increasing temperature (as evidenced by the liming of a tea kettle), higher pH, higher calcium concentration, and higher alkalinity levels. The LSI value can be lowered by reducing pH by the injection of an acid (typically sulfuric or hydrochloric) into the RO feed water. A recommended target LSI in the RO concentrate is negative 0.2 (which indicates that the concentrate is 0.2 pH units below the point of calcium carbonate saturation). A negative 0.2 LSI allows for pH excursions in actual plant operation. A polymer-based antiscalant can also be used to inhibit the precipitation of calcium carbonate. Some antiscalant suppliers have reported the efficacy of their product up to a positive LSI value of 2.5 in the RO concentrate (though a more conservative design LSI level is +1.8).

Caustics

Caustics, for a few process applications, can be injected to increase the RO feed pH. Typically, the only caustic injected is sodium hydroxide [NaOH] based on cost, availability and its solubility in water.

Most of the time the quality of the NaOH can be technical grade (and free of any additives). If the technical grade NaOH is not clear in color (e.g. has a blackish tint), then a higher grade of NaOH may be desirable such as rayon or nylon grade. NaOH is commercially available as 100% solid flake or as 20% or 50% solution. Caution is required in storing 50% caustic since it can freeze at temperatures just below 60°F (15.5°C), with a special concern for cold air drafts.

Caution is required in raising the feed pH as it can decrease the solubility of CaCO3, as measured by an increasing LSI, iron and manganese.

The most common process application that uses a caustic feed involve 2nd pass RO systems. In a 2- Pass RO, the permeate from a 1st pass RO becomes the feed to the 2nd pass. The 2nd pass RO “polishes” the 1st pass permeate to quality levels as good as 4 megohm-cm. Caustic is introduced into the 2nd pass feed for 4 reasons.

1. At a pH of 8.2 and higher, all carbon dioxide gas is converted into the bicarbonate ion. The bicarbonate ion will be rejected by the RO. Carbon dioxide, being a gas, would pass through the RO into the permeate stream unimpeded. The carbon dioxide would become an undesirable load to downstream polishing mixed beds.

2. Certain TOC constituents are better rejected at a higher pH.

3. Silica rejection and solubility are higher at higher pH levels (particularly above a pH of 9).

4. Boron rejection is higher at higher pH levels (particularly above a pH of 9).

Dechlorination Chemicals

At no time should there be a Free Chlorine residual in the feed water. Even very low levels of chlorine in the feed stream will result in irreparable oxidation damage of the membrane. Therefore, operators should ensure that oxidant does not enter the RO system. The two most common pretreatment methods for reducing chlorine levels are by absorption onto granular activated carbon filter media or by the use of a chemical reducing agent such as sodium bisulfite.

Pressurized carbon filters are typically used on small systems (50 gpm to 100 gpm or less) due to capital cost considerations. It is recommended that a premium grade of carbon is used that has been acid-washed for the removal of heavy metals and hardness and has very low levels of carbon fines that can foul the RO. Anytime new carbon media is installed, the carbon filter must be thoroughly rinsed until all carbon fines are absent from the effluent (this can take a few hours to days). One cannot rely on a 5-micron cartridge filter to protect the RO element from fouling due to carbon fines. The advantages of a carbon filter are their ability to remove organics from the feedwater that could foul the RO and that they are more reliable in treating all the feed water than a chemical feed system. The disadvantage is that carbon filters are notorious for breeding bacteria that can result in a bi ological fouling of the RO. Carbon filters have been known to see a doubling of viable bacteria counts in 24 hours, with capital intensive and operator intensive periodic steam sanitization the only answer for controlling microbial populations. Steam sanitization for critical applications (like USP Water production for the pharmaceutical industry) can be as frequent as 1 to 7 days.

Sodium Bisulfite [SBS] is the typical chlorine reducing agent of choice for larger RO systems. Note:

SBS should not be cobalt catalyzed. A SBS solution is made by dissolving solid sodium metabisulfite into water and has a pH of 4.6 at 1.0% (by weight) solution strength. A 10% (by weight) SBS solution will require the addition of 0.51 pounds of solid sodium metabisulfite into one gallon of water. The sodium metabisulfite is commercially available at 97.5% to 99% purity and can be stored safely up to six months in a dr y storage area. The SBS solution is not stable to air and reacts with oxygen as well as chlorine, therefore it is recommended that batches less than 2% by weight be used within 3 to 7 days and batch solutions less than 10 % be used within 7 to 14 days. Theoretical ly,

1.47 ppm of SBS (or 0.70 ppm of sodium metabisulfi te) wil l stoichiometrically neutralize 1.0 ppm of chlorine. Designers have been known to use a dosing rate of 1.8 ppm to 3.0 ppm of SBS per

1.0 ppm chlorine so as to include an industrial safety factor for brackish water RO systems. SBS needs to be added far enough upstream of the RO elements to insure at least 20 seconds of reaction time. Proper in-line mixing is required which preferably includes a static mixer.

Na2S2O5 (sodium metabisulfite) + H2O —> 2 NaHSO3 (sodium bisulfite)

NaHSO3 + HOCl —> NaHSO4 (sodium bisulfate) + HCl (hydrochloric acid)

NaHSO3 + Cl2 + H2O —> NaHSO4 + 2 HCl

The advantages of SBS dechlorination are that it is less capital intensive than carbon filters for large systems, the reaction by-products are r eadily removed by t he RO, and re sidual SBS i s readily removed by the RO.

The disadvantages of SBS c an be the need to handle and mix a small volume of chemical and an increased risk in chlorine making it to the membranes if sufficient monitoring and c ontrols are not designed into the dechlorination system. In a few cases where Sulfur Reducing Bacteria [SBRs] are present in the feed supply, the bisulfite functions as a nutrient and enhances their growth. SBRs are usually found in anaerobic (low oxygen content) shallow wells in regions such as Florida. Frequently hydrogen sulfide (H2S) is also present as a by-product of the SBR metabolic process. Monitoring for dechlorination can be performed by the use of an Available Free Chlorine monitor, monitoring for a residual bisulfite concentration, or by an ORP meter. We recommend that the feed to the RO/NF system is equipped with a O RP (Oxidation-Reduction Potential) meter. The feedwater can then be c ontinuously monitored for the presence of oxidant. Except in wastewater applications where chloramines are used, the ORP meter reading should always be below 300 mV. If it exceeds 300 mV, the plant operator should receive a warning that a dangerous level of oxidant is getting to the membrane and should take action, such as adding or increasing the dose of SBS, to reduce the oxidant concentration. If the ORP value reaches 350 mV, the plant should be shut down until the oxidant concentration can be reduced to a safe value (ORP < 300 mV).

Antiscalants and Dispersants

There are a large number of chemical suppliers who supply a variety of proprietary antiscalants and dispersants to improve the operation of RO and NF systems.

Antiscalants are a f amily of chemicals designed to inhibit the formation and pr ecipitation of crystallized mineral salts that form scale. Most antiscalants are proprietary organic man-made polymers (e.g. polyacrylic acids, carboxylic acids, polymaleic acids, organo-phosphates, polyphosphates, phosphonates, anionic polymers, etc.). Molecular weight of these polymers can range from 2,000 Dalton to 10,000 Dalton.

Antiscalant technology for RO systems initially was derived from chemistries used in cooling water and boiler water applications. It should be dul y noted that the large number of different antiscalant chemistries that have evolved have produced a wide variety of results and efficacy depending on the application and organic polymer used.

Caution should be used in the use of antiscalants made of polyacrylic acid. They are susceptible to the formation of a f oulant that settles on the membrane surface if there are high levels of iron. This foulant will increase feed pressure requirements, but typically can be cleaned using a low pH cleaning.

Caution should also be used in the use of antiscalants that are anionic in nature (e.g. polyacrylic acids) when a c ationic-based coagulant or filtering aid i s used in t he pretreatment. A very viscous, sticky foulant can be produced that will increase feed pressure requirements, and it can be very difficult to clean off.

An antiscalant that was popular in the early days of RO was sodium hexametaphosphate (SHMP), but its use has been greatly reduced with the advent of proprietary antiscalants. SHMP has a number of limitations.

Batches of diluted SHMP has to made every 2 to 3 days due to hydrolysis by exposure to air, which would dilute its efficacy and create a potential calcium phosphate scaling risk. SHMP has a reduced protection to calcium carbonate scaling with a maximum LSI rating in the concentrate of +1.0.

Antiscalants retard the growth of crystalline salt structures in the RO feed and concentrate streams, thereby allowing a c oncentration of sparingly soluble salts in excess of the normal solubility limits. Antiscalants can be used to replace, or can be usedin conjunction with, acid feed to control calcium carbonate scaling.

A number of factors can affect the rate of mineral scale formation. Lower temperatures reduce the solubility of mineral scales (the exception here is calcium carbonate scale formation is enhanced when temperature increases). The solubility of sparingly soluble salts increase with higher TDS levels (this is due to the increased interference by all the ions in the seeding process of scale formation).

Most common mineral scalants of concern:

Calcium carbonate [CaCO3]

Calcium sulfate [CaSO4]

Strontium sulfate [SrSO4]

Barium sulfate [BaSO4]

Less common mineral scalants are:

Calcium phosphate [CaPO4]

Calcium fluoride [CaF2]

Dispersants are a family of organic man-made polymers designed to inhibit the agglomeration and deposition of foulants onto the membrane surface. Dispersants are sometimes referred to as anti-foulants. Foulants tend to be a softer, non-crystalline deposit. Dispersant chemicals frequently have antiscalant properties. The efficacy of differing dispersants can vary for different foulants, so one needs to know what foulant they are treating for.

Foulants treated by dispersants:

Mineral Scales Metal Oxides and Hydroxides [iron, manganese, aluminum]

Polymerized Silica Colloidal material [defined as very small particles that stays in an infinite suspension and can be composed of clay, iron, aluminum, silica, sulfur and/or organic matter]

Biological matter

Predicting the maximum solubility of super-saturated silica can be dif ficult. In particular, iron present in the feed water can readily produce iron silicates and dramatically reduce the allowable concentration of silica in the RO concentrate stream. Other major factors are pH and temperature. Predicting the maximum levels of metals (e.g. iron, manganese, aluminum) can also be difficult. The soluble forms of the metal ions allow for higher levels of saturation. The insoluble forms tend to act more like particles and colloids.

Optimal dosing and al lowable maximum saturation levels of scalants and f oulants for antiscalants/dispersants are best determined by the chemical supplier who utilizes a proprietary software package.

Maintaining proper dosing levels of an antiscalant/dispersant is important. Under-dosing can cause scal ing or foul ing. Ov er-dosing can cause a deposi t ion of the antiscalant/dispersant onto the membrane, creating a fouling problem.

It is important that the antiscalant/dispersant be thoroughly flushed from the RO elements at shutdown, as it can settle onto the membrane and c ause a f ouling problem. The injection of antiscalant/dispersant should cease during low pressure flushes when the RO feed water is the source of flush water.

The design of the antiscalant/dispersant injection system into the RO feed stream needs to make sure that it is properly mixed before it enters the RO elements. The use of a s tatic mixer is the most effective mixing method, but adds cost to a system. Most systems have the injection point just before the RO feed cartridge filter(s) and rely on the cartridge filter residence time and RO feed pump agitation to affect the mixing.

It is recommended that if there is an acid pH adjustment of the RO feed water, the acid should be injected upst ream and t horoughly mixed before i t reaches the antiscalant/dispersant injection point. Concentrated pockets of low pH ac id may destroy the efficacy of the antiscalant/dispersant.

The chemical metering pump used for the injection of antiscalant/dispersant should be adjusted to maximize the frequency of injection. A suggested minimum stroking frequency is once every 5 seconds. Typical dosing rates of antiscalants/dispersants are 2 to 5 ppm. To achieve a reasonable stroking frequency of the metering pump, a dilution of the antiscalant/dispersant may be required.

Antiscalants/dispersants are supplied as either a concentrated liquid or in a powder form. Dilution water should be hardness-free and preferably be of RO permeate quality. Diluted antiscalant/dispersants can become biologically fouled in the day tank depending on temperatures and how much it has been diluted. A suggested residence time for a diluted solution is 7-10 days.

Normally, undiluted antiscalant/dispersant solutions do not suffer from biological fouling.

Another major issue in selecting an antiscalant/dispersant is the assurance that it is compatible with the RO membrane being used. There are incompatible chemicals that can irreversibly foul or damage a RO membrane. We rely on t he supplier of the chemical to have performed all RO membrane compatibility and process efficacy tests.

Some recommended questions to ask chemical suppliers of antiscalants and dispersants are:

Is it compatible with the RO membrane type in question?

Is there an installation list of end-users with over 1000 hours of successful operation?

Does it react adversely with any other constituents in the RO feed water (e.g. iron, heavy metals, cationic polyelectrolytes, etc.)? What are the recommended dosing rates and maximum dosing rates?

What are the projected limits of solubility for individual scaling and fouling components?

Are there any special discharge concerns?

Is it suitable for potable water applications (if required)?

Does the supplier offer other RO chemical feeds like antiscalant-compatible coagulants, biocides and cleaning chemicals?

Does the supplier offer off-site technical support services like membrane autopsies or cleaning of elements?

Glossary

Alkalinity: Alkalinity is comprised primarily of carbon dioxide, bicarbonate, carbonate and hydroxides.

Naturally occurring alkalinity is the earth’s natural buffering system in that small doses of strong acids (e.g. acid rain) react with alkalinity and result in relatively small changes in pH. Carbon dioxide and bicarbonate are in a balance between the pH range of 4.4 and 8.2. At a pH of 4.4 or lower, all alkalinity is in the form of carbon dioxide. At a pH o f 8.2, there is no c arbon dioxide and all alkalinity is bicarbonate. Bicarbonate and carbonate are in a balance between the pH range of 8.2 and 9.6. At a pH of 9.6, there is no c arbon dioxide or bicarbonate and all alkalinity is carbonate. As the pH increases above 9.6, hydroxyl alkalinity due to the presence of the hydroxide ion starts to occur. Most naturally occurring water sources have a pH between 6 and 8.4, so the presence of hydroxides is the result of man-made activity. Alkalinity, especially by boiler water chemists, can be r eported as M- Alkalinity and P -Alkalinity. M-Alkalinity measures the Total Alkalinity in a water in terms of “ppm as calcium carbonate” based on an acid titration to a pH of 4.2 using a Methyl orange indicator endpoint.

P-Alkalinity measures the amount of bicarbonate, carbonate and hydroxyl alkalinity based on an acid titration to a pH of 8.2 using a Phenolphthalein pink indicator endpoint.

Barium (Ba): A divalent cation. The solubility of barium sulfate (BaSO4) is low and can cause a RO scaling problem in the back-end of a RO. Barium sulfate solubility is lower with increasing sulfate levels and decreasing temperatures. Typically, barium can be found in some well waters, with typical concentrations less than 0.05 ppm to 0.2 ppm. It is important that barium be measured with instruments capable of 0.01 ppm (10 ppb) minimum detection levels. With saturation at 100%, super-saturation up to 6000% is typical with an antiscalant.

Bicarbonate (HCO3): A monovalent anion. The solubility of calcium bicarbonate is low and can cause a RO scaling problem in the back-end of a RO. Calcium bicarbonate solubility is measured using LSI (Langlier Saturation Index) for brackish waters or the Stiff-Davis Index for seawaters and is lower with increasing temperature and i ncreasing pH. Bicarbonate is one component of alkalinity and its concentration is in a balance with carbon dioxide between the pH range of 4.4 and 8.2 and in a balance with carbonate between the pH range of 8.2 and 9.6.

Calcium (Ca): A divalent cation. Calcium, along with magnesium, is a major component of hardness in brackish water. The solubility of calcium sulfate (CaSO4)(gypsum) is typically limited to 230% with the use of an antiscalant. The solubility of calcium carbonate is typically limited to a LS I (Langlier Saturation Index) value of positive 1.8 to 2.5.

Carbon Dioxide (CO2): Carbon dioxide is a gas that when dissolved in water reacts with the water to form weak carbonic acid (H2CO3). If a pure water was completely saturated with carbon dioxide, its concentration would be about 1600 ppm and the pH would be about 4.0. A typical source for carbon dioxide in natural waters is the result of a balance with bicarbonate alkalinity based on the pH of the water. The concentration of carbon dioxide in water is typically indirectly determined by graphical comparison to the bicarbonate concentration and pH. Carbon dioxide and the bicarbonate ion are in a balance between the pH range of 4.4 and 8.2. The alkalinity is all carbon dioxide at pH 4.4 and is all bicarbonate at pH 8.4. The RO design program calculates the carbon dioxide level based on the bicarbonate level and pH of the water. Carbon dioxide, being a gas, is not rejected or concentrated by a RO membrane, therefore its concentration will be the same in the feed, permeate and concentrate.

Acidifying the RO feed water will lower pH by converting bicarbonate to carbon dioxide.

Carbonate (CO3): A divalent anion. The solubility of calcium carbonate is low and can cause a RO scaling problem in the back-end of a RO. Calcium carbonate solubility is measured using LSI (Langlier Saturation Index) for brackish waters or SDSI (Stiff-Davis Index) for seawaters and is lower with increasing temperature and increasing pH. Carbonate is one component of alkalinity and its concentration is in a balance with bicarbonate between the pH range of 8.2 and 9.6. At a pH of 9.6 and higher, there is no carbon dioxide or bicarbonate, with all alkalinity being in the carbonate form.

Ionic Strength: The solubility of sparingly soluble salts increases with increasing feedTDS. To account for this effect in calculating the solubility of a salt (e.g. calcium sulfate, barium sulfate, strontium sulfate or SDSI), the Ionic Strength of a water is calculated. The Ionic Strength of each ion is derived by taking the ppm concentration of each ion (as calcium carbonate) and multiplying each monovalent ion by 1 x 10-5 and each divalent ion by 2 x 10-5. Summing the Ionic Strength of each ion then derives the total Ionic Strength of the water.

Iron (Fe): Iron is a water contaminant that takes two major forms. The water-soluble form is known as the ferrous state and has a + 2 valence state. In non-aerated well waters ferrous iron behaves much like calcium or magnesium hardness in that it can be removed by softeners or its precipitation in the back end of the RO system can be controlled by the use of a dispersant chemical in an RO feed water. The water-insoluble form is known as the ferric state and has a + 3 v alence state.

Typical ly, RO manufacturers will recommend that combined iron levels be less than 0.05 ppm in the RO feed. If all iron is in the soluble ferrous form, iron levels up to 0.5 ppm in the feed can be tolerated if the pH is less than 7.0 (though an iron dispersant is recommended). The introduction of air into water with soluble ferrous iron will result in the oxidation to insoluble ferric iron. Soluble iron can be found in deep wells, but can be converted into the more troublesome insoluble iron by the introduction of air by being placed in tanks or by leaky pump seals. Soluble iron can be treated with dispersants or can be removed by iron filters, softeners or lime softening. Insoluble ferric iron oxides or ferric hydroxides, being colloidal in nature, will foul the front end of the RO system. Sources of insoluble iron are aerated well waters, surface sources, and i ron scale from unlined pipe and t anks.

Insoluble iron can be removed by iron filters, lime softening, softeners (with limits), ultrafiltration (with limits) and multimedia filtration with polyelectrolyte feed (with limits). Precautions are required with the use of potassium permanganate in manganese greensand iron filters in that potassium permanganate is an ox idant that could damage any polyamide membrane. Precautions are also required with a c ationic polyelectrolyte in that they can irreversibly foul a negat ively charged polyamide membrane. Corrosion proof vessels and piping (e.g. FRP, PVC or stainless steels) are recommended for all RO systems, RO pretreatment, and distribution piping coming to the RO system.

Iron as foulant will quickly increase RO feed pressure requirements and i ncrease permeate TDS. In some cases, the presence of iron can create a bio-fouling problem by being the energy source for iron-reducing bacteria. Iron-reducing bacteria can cause the formation of a slimy biofilm that can plug the RO feed path.

LSI (Langlier Saturation Index): LSI is a method of reporting the scaling or corrosive potential of low TDS brackish water based on the level of saturation of calcium carbonate. LSI is important in RO water chemistry in determining whether a water will or will not formcalcium carbonate scale. Water with a negative LSI is considered corrosive to metal piping and will not form calcium carbonate scale.

Water with a positive LSI is not corrosive, but it will tend to form calcium carbonate scale. LSI is important to RO chemists as a measurement of the scaling potential for calcium carbonate. The LSI value is calculated by subtracting the calculated pH of saturation of calcium carbonate from the actual feed pH. Calcium carbonate solubility decreases with increasing temperature (as evidenced by the liming of a teakettle), higher pH, higher calcium concentration, and higher alkalinity levels. The LSI value can be lowered by reducing pH by the injection of an acid (typically sulfuric or hydrochloric) into the RO feed water. A recommended target LSI in the RO concentrate is negative 0.2 (which indicates that the concentrate is 0.2 pH units below the point of calcium carbonate saturation). A negative 0.2 LSI allows for pH excursions in actual plant operation. A polymer-based antiscalant can also be used to inhibit the precipitation of calcium carbonate. Some antiscalant suppliers have claimed the efficacy of their product up t o a positive LSI value of 2.5 in the RO concentrate (though a more conservative design LSI level is +1.8). Sodium h exam etaphosphate, an inorganic antiscalant, was used in the early days of RO but the maximum concentrate LSI was + 0.5 and it had to be made in short-lived batches as the air easily oxidized it.

Magnesium (Mg): A divalent cation. Magnesium can account for about a third of the hardness in a brackish water, but can have a concentration five times higher than calcium in sea water. The solubility of magnesium salts is high and typically does not cause a scaling problem in RO systems.

Manganese (Mn): Manganese is a water contaminant present in both well and s urface waters, with levels up to 3 ppm. Manganese, like iron, can be found in organic complexes in surface waters. In oxygen-free water, it is soluble. In the oxidized state, it is insoluble and usually in the form of black manganese dioxide (MnO2) precipitate. An alert level for potential manganese fouling in a RO aerated RO feed waters is 0.05 ppm. Drinking water regulations limit manganese to 0.05 ppm due to its ability to cause black stains. Dispersants used to control iron fouling can be us ed to help control manganese fouling.

pH: The pH of the feed water measures the acidity or basicity. A pH of 7.0 is considered neutral. A pH between 0.0 and 7.0 is acidic. A pH between 7.0 and 14.0 is basic. To the analytical chemist, pH is a method of expressing hydrogen ion concentration in terms of the power of 10 with the pH value being the negative logarithm of the hydrogen ion concentration. To the water chemist, pH is important in defining the alkalinity equilibrium levels of carbon dioxide, bicarbonate, carbonate and hydroxide ions.

The concentrate pH is typically higher than the feed due to the higher concentration of bicarbonate/carbonate ions relative to the concentration of carbon dioxide. The RODESIGN program allows theuser to adjust the pH of the feed water using hydrochloric and sulfuric acid. Lowering the feed pH with acid results in a lower LSI (Langlier Saturation Index) value, which reduces the scaling potential for calcium carbonate. Feed and concentrate (reject) pH can also affect the solubility and fouling potential of silica, aluminum, organics and oil. Variations in feed pH c an also affect the rejection of ions. For example, fluoride, boron and s ilica rejection are lower when the pH bec omes more acidic.

SDSI (Stiff Davis Saturation Index): SDSI, in similar fashion as LSI, is a method of reporting the scaling or corrosion potential of high TDS seawater based on the level of saturation of calcium carbonate. The primary difference between SDSI for high TDS seawater and LSI for low TDS brackish water is the effect that increasing ionic strength has on increasing solubility. The solubility of sparingly soluble salts increase with higher TDS and ionic strength, based on the theory that a denser ion population interferes in the formation and/or precipitation of the sparingly soluble salt.

Silica (SiO2): Silica (silicon dioxide), in some cases, is an anion. The chemistry of silica is a complex and somewhat unpredictable subject. In similar fashion as TOC reports the total concentration of organics (as carbon) without detailing what the organic compounds are, silica reports the total concentration of silicon (as silica) without detailing what the silicon compounds are. The “Total Silica” content of water is composed of “Reactive Silica” and “Unreactive Silica”. Reactive silica (e.g.

silicates SiO4) is dissolved silica that is slightly ionized and has not been pol ymerized into a long chain. Reactive silica is the form that RO and ion exchange chemists hope for. Reactive silica is the form of silica to be us ed in RO projection programs. Reactive silica, though it has anionic characteristics, is not counted as an anion in terms of balancing a water analysis but it is counted as a part of total TDS. Unreactive silica is polymerized or colloidal silica, acting more like a solid than a dissolved ion. Silica, in the colloidal form, can be removed by a RO but it can cause colloidal fouling of the front-end of a RO . Colloidal silica, with sizes as small as 0.008 micron can be m easured empirically by the SDI (Silt Density Index) test, but only that portion that is larger than 0.45 micron or larger. Particulate silica compounds (e.g. clays, silts and sand) are usually 1 micron or larger and can be measured using the SDI test. Polymerized silica, which uses silicon dioxide as the building block, exists in nature (e.g. quartzes and agates). Silica, in the polymerized form, also results from exceeding the reactive silica saturation level. The solubility of reactive silica is typically limited to 200-300% with the use of a silica dispersant. Reactive silica solubility increases with increasing temperature, increases at a pH less than 7.0 or more than 7.8, and decreases in the presence of iron which acts as a catalyst in the polymerization of silica. Silica rejection is pH sensitive, with increasing rejection at a more basic pH as the reactive silica exists more in the salt form than in the acidic form.

Strontium (Sr): A divalent cation. The solubility of strontium sulfate is low and can cause a RO scaling problem in the back-end of a RO. Strontium sulfate solubility is lower with increasing sulfate levels and decreasing temperatures. Typically, strontium can be found in some well waters where lead ores are also present, with typical concentrations less than 15 ppm. With saturation at 100%, super-saturation up to 800% is typical with an antiscalant.

Sulfate (SO4): A divalent anion. The solubility of calcium, barium and strontium sulfate is low and can cause a RO scaling problem at the concentrate end of a R O. The solubility of these sparingly soluble salts is lower with decreasing temperature. The recommended upper limit for sulfate in potable water is 250 ppm based on taste issues.

Basics of RO and NF: Principle of Reverse Osmosis and Nanofiltration

The phenomenon of osmosis occurs when pure water flows from a dilute saline solution through a membrane into a higher concentrated saline solution.

How Reverse Osmosis Works

The phenomenon of osmosis is illustrated in Figure 1.4. A semi-permeable membrane is placed between two compartments. “Semi-permeable” means that the membrane is permeable to some species, and not permeable to others. Assume that this membrane is permeable to water, but not to salt. Then, place a salt solution in one compartment and pure water in the other compartment. The membrane will allow water to permeate through it to either side. But salt cannot pass through the membrane.

Figure 1.4 Overview of osmosis

Osmosis

Water diffuses through a semi-permeable membrane toward region of higher concentration to equalize solution strength. Ultimate height difference between columns is “osmotic” pressure.

Reverse Osmosis

Applied pressure in excess of osmotic pressure reverses water flow direction. Hence the term “reverse osmosis“.

As a fundamental rule of nature, this system will try to reach equilibrium. That is, it will try to reach the same concentration on both sides of the membrane. The only possible way to reach equilibrium is for water to pass from the pure water compartment to the salt-containing compartment, to dilute the salt solution.

Figure 1.4 also shows that osmosis can cause a rise in the height of the salt solution. This height will increase until the pressure of the column of water (salt solution) is so high that the force of this water column stops the water flow. The equilibrium point of this water column height in terms of water pressure against the membrane is called osmotic pressure.

If a force is applied to this column of water, the direction of water flow through the membrane can be reversed. This is the basis of the term reverse osmosis. Note that this reversed flow produces a pure water from the salt solution, since the membrane is not permeable to salt.

How Nanofiltration Works

The nanofiltration membrane is not a complete barrier to dissolved salts. Depending on the type of salt and the type of membrane, the salt permeability may be low or high. If the salt permeability is low, the osmotic pressure difference between the two compartments may become almost as high as in reverse osmosis. On the other hand, a high salt permeability of t membrane would not allow the salt concentrations in the two compartments to remain very different. Therefore the osmotic pressure plays a minor role if the salt permeability is high.

In practice, reverse osmosis and nanofiltration are applied as a crossflow filtration process.

The simplified process is shown in Figure 1.5.

How to Use Reverse Osmosis and Nanofiltration in Practice Figure 1.5 Reverse osmosis process

With a high pressure pump, feed water is continuously pumped at elevated pressure to the membrane system. Within the membrane system, the feed water will be split into a low-saline and/or purified product, called permeate, and a high saline or concentrated brine, called concentrate or reject. A flow regulating valve, called a concentrate valve, controls the percentage of feedwater that is going to the concentrate stream and the permeate which will be obtained from the feed.

The key terms used in the reverse osmosis / nanofiltration process are defined as follows.

Recovery - the percentage of membrane system feedwater that emerges from the system as product water or “permeate”. Membrane system design is based on expected feedwater quality and recovery is defined through initial adjustment of valves on the concentrate stream. Recovery is often fixed at the highest level that maximizes permeate flow while preventing precipitation of super-saturated salts within the membrane system.

Rejection - the percentage of solute concentration removed from system feedwater by the membrane. In reverse osmosis, a high rejection of total dissolved solids (TDS) is important, while in nanofiltration the solutes of…