Softener 2
Softener 2
Softener 2
ABSTRACT
Sodium zeolite softeners are often seen as the most efficient method of scale control for reverse osmosis plants and are widely used for pretreatment. This paper looks at the impact on economics and system performance by replacing ion exchange softening with chemical antiscalant.
Saturation Index (SI). Precipitation is predicted to occur when the Saturation Index exceeds 1.0. The ion product, Ip, is defined for salt AmBn in Eq. (1). Table 1 lists the solubility constant, Ksp for common scales:1 Equation (1). Ip = [A] * [B] where [ ] denotes molar concentration
Table 1 Molar Solubility Product for Common Salts
m n
INTRODUCTION
Controlling the precipitation of sparingly soluble salts has long been an issue with making Reverse Osmosis (RO) technology viable. Scale precipitates on the membrane surface when a soluble salt exceeds its solubility limit in the brine stream of the RO. If some method for preventing scale build-up is not employed, permeate flow and quality are adversely affected. Excessive build-up on the membrane can permanently damage the element and accelerate the need for replacement. Although common treatment strategies for controlling scale in RO systems include reduced operating recovery and feedwater pH reduction, the two most commonly employed technologies in industrial installations are the use of ion exchange softeners and the use of chemical scale inhibitors. This paper reviews the four primary ways to control scale in RO systems, but focuses on the advantages and disadvantages of ion exchange softening versus chemical scale inhibitors. Two case studies are presented where ion exchange softening was replaced with chemical scale inhibitors as the method to control scale. The economics of this change are discussed with respect to RO system performance.
Ksp Molar Solubility Product 2.5 x 105 6.3 x 107 2.0 x 1010 5.0 x 1011
In the case of calcium carbonate (CaCO3), the Langelier Saturation Index (LSI) is used to determine if a water is scale forming for water with TDS levels less than 10,000 mg/L. LSI incorporates alkalinity concentration, calcium concentration, pH and temperature. A positive LSI indicates that the water is scale forming. Computer projection software is available that calculates saturation index and LSI based on feedwater characteristics and system operating parameters. In RO systems, soluble salts found in the feedwater are concentrated in the brine stream due to water being permeated through the membrane. In the boundary layer on the membrane surface, the concentration of salts exceeds that of the bulk water due to the concentration polarization effect. The concentration of ions is as much as 1020% higher along the membrane surface than in the bulk water. The actual concentration depends upon their rate of diffusion back into the bulk water, which is a function of turbulence in the bulk stream.2 Because of this phenomenon, some method of scale prevention should be considered when saturation indices (SI) exceed >0.8 or Langelier Saturation Index (LSI) exceeds 0.2.
SCALE
Precipitation of scale occurs when the solubility of a particular salt is exceeded. Scale can be predicted to form by comparing the ion product of the considered salt (Ip) to the solubility product for that salt (Ksp) under the same conditions. The ratio of Ip/Ksp is the
Presented at the 61st Annual International Water Conference, Pittsburgh, Pennsylvania, October 2226, 2000
Equation (4). CO2 + H2O H2CO3 H+ + HCO3 Carbon dioxide, a dissolved gas, passes through the membrane into the permeate stream. Depending upon the level of alkalinity in the feedwater, the impact on permeate quality can be significant, leading to the need for further treatment to improve its quality. The economics of feeding acid involve the cost of acid. Food grade acid must be used to prevent fouling of the membrane from contaminates found in lesser grades of acid. The dosage of acid is determined by the pH and alkalinity level of the brine stream. Significant levels of alkalinity may result in the need for additional treatment to remove carbon dioxide. System performance and permeate quality may be acceptable with relatively low pH and alkalinity feedwater sources. Higher pH and alkalinity feedwaters will require greater amounts of acid to prevent carbonate scale, increasing the cost of the treatment while decreasing the quality of the permeate produced.
REDUCED FEEDWATER pH
For most natural sources of water, calcium and carbonate are in equilibrium as described in Eq. (3). Equation (3). Ca++ + HCO3 H+ + CaCO3 Reducing pH, by adding a source of H+ to the feedwater (i.e., adding acid), shifts this equilibrium to the left keeping calcium ions in solution. Typically, sulfuric acid or hydrochloric acid is used. This simple treatment strategy is effective for carbonate scale only. When using acid, consideration must be given to the impact that reduced pH has on the solubility of other ions found in the feedwater such as silica. Use of sulfuric acid may also cause a potential problem with sulfate scale (e.g., calcium sulfate, barium sulfate, strontium sulfate) due to the additional sulfate ions contributed by the acid. Reducing feedwater pH also has an impact on permeate water quality that needs to be considered. As pH is reduced below 8.2, bicarbonate alkalinity (HCO3) is converted to carbonic acid (H2CO3) which dissociates into carbon dioxide (CO2). Eq. (4) shows this relationship.
2
This process is reversible and uses a synthetic resin as the medium to carry out the exchange. After the resin has become saturated with the unwanted ions, it is regenerated by washing it with a solution of the desired ion. The simplest form of ion exchange technology used is the sodium exchanger, which is regenerated with sodium chloride (NaCl). This process softens water by exchanging calcium and magnesium ions with sodium ions.
Table 2 Concentration Factor versus System Recovery for 100 gpm feedwater containing 100 mg/L Ca
Recovery R% 0 (Feedwater) 10 20 30 40 50 60 70 75 80 90 Conc. Permeate Brine [Ca++] in the Factor (Cf) Flow Stream Flow Stream 1 1.11 1.25 1.43 1.67 2.00 2.50 3.33 4.00 5.00 10.00 0 10 20 30 40 50 60 70 75 80 90 100 90 80 70 60 50 40 30 25 20 10 100 111 125 143 167 200 250 333 400 500 1,000
Conventional ion exchange softening systems are made up of a vessel containing a fixed bed of resin which is regenerated in a cocurrent basis, meaning the water to be treated and the regeneration solution are passed in the same direction, from the top to the bottom of the vessel. Feedwater enters the vessel, where calcium, magnesium and other cations for which the resin has a greater affinity are exchanged for sodium ions. When the effluent hardness concentration exceeds a predetermined level, the unit is taken off-line and regenerated. The complete cycle of exchange is comprised of 5 steps: Service: (production) The unit is producing water of the desired quality. Back-washing: (loosening) Water flows in the reverse direction of service to remove any particulate matter or debris that has entered the system. Regeneration: (chemical injection) Diluted regenerant is passed in the direction of service (cocurrent) through the bed to restore the capacity of sodium ions. Displacement rinse: (slow rinse) Water is used to push out the remaining regenerate solution from the vessel. Fast Rinse: Water is introduced at the service rate of flow until the desired quality of produced water is achieved. Although scale is the primary reason for using an ion exchanger softener, these units are often installed as sacrificial filters to trap particulate matter, including iron, to prevent it from entering the RO system. Additionally they are sometimes installed to remove residual cationic polymers, which may pass through a poorly maintained upstream media filter. Ion exchange resins are not intended to fill these roles and perform poorly in doing so. When used for purposes other than ion exchange, operation and maintenance costs of this unit process increase significantly. As with any unit process, there is a cost associated with purchasing, installing and operating an ion exchange unit. Most often, it is only the capitalized costs of the ion exchange unit that is reviewed when making the installation decision. Operating cost, usually not calculated and assumed to be negligible, is often ignored and can be significant. The total operating costs of a softener include: Salt Cost: A function of total feedwater hardness, desired hardness output and regeneration efficiency Water Costs: Includes the costs of treatment (pre and post) for water used to regenerate the unit which is wasted Electricity: Incremental pumping cost to overcome system pressure drop
3
Labor: Dedicated to testing, regenerating, receiving salt and performing system maintenance Maintenance: Includes resin replacement and general maintenance items The most significant cost for operating an ion exchange softener is that of salt, which is most directly impacted by, and is linear in relationship to the feedwater hardness. Disposal cost can be significant in some cases, particularly if the resin is regenerated in the hydrogen form. Figure 1 shows the total annual cost and cost per 1,000 gallons for operating a 200gpm softener at hardness levels from 25 to 300 ppm as CaCO3 regenerated with sodium chloride. The basis for this calculation appears in Table 3.
Table 3 Basis of calculations for Figure 1 Item Salt Cost Electricity Labor Labor Hours Resin Replacement Cost Water Cost Waste Water Resin Volume # NaCl / ft
3
Basis $65 / Ton $0.070 / kW Hour $50 hr 2 hr /wk $30/ft3 (10% annual replacement) $2.00 / kgal $1.00 / kgal 100 ft3 10.0 80% 95% 10 psi 190 (gal cycle) 750 (gal cycle) 5000 (gal cycle)
Pump Efficiency Motor Efficiency Pressure Drop Regeneration Volume Back Wash Volume Rinse Water
ering what type of RO pretreatment to use. The following two case studies show the impact on operating cost versus performance of two existing RO plants that converted from ion exchange softening to chemical scale inhibitors. CASE STUDY #1 A microelectronics plant in California commissioned an RO system in 1985. Key features of the water system are: Raw water source: Permeate use: Membranes: Output: Recovery: Array: Ion Exchange Softeners: Water analysis: Water System: City Rinse water for hard drive disk manufacture Polyamide, spiral wound 120 gpm 75% 4:2 2 x 40 ft3 resin regenerated with sodium chloride Variable. Table 4 shows the analysis used for design. Refer to Figure 2.
Previous Operation Plant used two 40 ft3 ion exchange softeners to control scale in the RO system. Figure 3 shows a computer projection for the calculated saturation index and LSI for the raw water based on the systems 75% recovery. By using ion exchange to soften the water, scale potential was minimized. After several years of operation, the softeners were in need of significant repair. Softener laterals were broken that regularly allowed resin beads to break through and foul down stream cartridge filters in front of the RO. Estimates for repair were $25,000. A resin analysis determined that it needed to be replaced at an estimated cost of $6,400. A thorough system survey determined the following additional information: Annual Salt Cost: Annual Water Cost: $24,000 per year for regeneration $11,600 per year for system regeneration $1,100 per year (additional) to overcome the vessel pressure drop.
CASE STUDIES
The economics of capital and operating costs versus system performance are important issues in consid-
Labor Cost:
$11,700 per year for maintaining softeners and handling salt delivery System Performance: Membranes cleaned on average every 10 months due to a decrease in permeate flow RO Operations: 1416 hours/day
Table 4 Case Study Water Analysis Ion Ca Mg Na K SO4 Cl HCO3 NO3 SiO2 pH LSI (brine) Case #1 (mg/L as ion) 150 57 98 3.4 118 251 375 24 25 7.9 2.48 Case #2 (mg/L as ion) 60 30 36 n/a 60 90 180 19 9 7.8 1.83
tor. A threshold scale inhibitor was used to control scale in the RO system. Because the system does not operate continuously, a feedwater flush system needed to be installed to remove concentrated salts from the RO prior to shutting it down. The cost of this installation was $500. Additional costs were: Annual Scale Inhibitor Cost: $8,000 per year Dosing Pump: $700 Labor: $1300 per year to take delivery and make up scale inhibitor System Performance: Membranes were cleaned after 11 months due to a decrease in permeate flow as routine Summary Performance of RO was unchanged. Salt rejection, permeate flow, feed/differential pressures remained consistent with previous performance. An annualized savings of $39,100 ($24,000 + 11,600 + 1,100 + 11,700 8,000 1,300) in operating costs were realized. Savings due to capital avoidance of $30,200 ($25,000 + 6,400 500 700) were realized. The Return on $1,200 Investment (ROI) exceeded 3150% [100% * ($39,100 1200)/1200].
Current Operation Due to the need for capital expenditure to re-build or replace the softeners, it was decided to replace them with a chemical scale inhibi5
CASE STUDY #2 A semi-conductor plant in the United Kingdom commissioned an RO system in 1990. Key factors for this water system are: Raw water source: City Permeate use: Rinse water for hard drive disk manufacture Membranes: Polyamide, spiral wound Output: 480m3/day (75 gpm) Recovery: 75% Array: 2:1:1 Ion Exchange Resin regenerated with sodium Softeners: chloride Water analysis : Variable. Table 4 shows the analysis used for design. Water System: Refer to Figure 4. Previous Operation The plant used an ion exchange softener with an estimated bed volume of 1.3 m3 (46 ft3) to control scale in its RO system. Figure 5 shows a computer projection for the calculated saturation index and LSI for the raw water based on 75% recovery. The ion exchange softener was being operated beyond its design capacity in order to provide sufficient feedwater for the RO. Installation of a replacement softener was estimated at $15,000 (10,000)8. Plant engineers were tasked by management to reduce costs in their area. A thorough system survey determined the following additional information: Annual Salt Cost: $13,800 (9,200) per year for regeneration Annual Water Costs: $1,342 (895) per year for system regeneration Electrical Cost: $337 (225) per year (additional) to overcome vessel pressure drop
6
Labor Costs:
$3,168 (2,112) per year for maintaining softeners and handling salt System Performance: Membrane cleaning not necessary due to operating performance RO Operations: Plant runs on average 1618 hours/day. RO had an auto-flush system. Due to the need to avoid capital expenditure to replace the softener and to reduce plant-operating costs, it was decided to replace the ion exchange softener with a chemical scale inhibitor. Current Operation A threshold scale inhibitor is used to control scale in the RO system. Because the RO had a pre-existing auto-flush system to prevent concentrate reject standing in the membranes during prolonged downtime, no investment was required. Key factors: Annual Scale Inhibitor Cost: $3,900 (2,600) per year Dosing Pump: $600 (400) includes a day tank Annual Labor Costs: $1500 (1000) per year to take delivery and make up scale inhibitor System Performance: After two years of operation, the membranes have not required cleaning. Summary Performance of the RO system has been unchanged since replacing the ion exchange system with a chemical scale inhibitor. Salt rejection, permeate flow and feed/differential pressures remained consistent with past performance. The system has gone over 2 years without requiring the membranes
to be cleaned. Savings in total annualized operating costs of $13,247 [$13,800 + 1,342 + 337 + 3,168 3,900 1,500] (8,832) are recognized. Additional savings of $14,400 [15,000 600] (9,600) in capital avoidance were recognized. The Return on the $600 (400) Investment (ROI) exceeded 2,100% [100% * ($13,247 600)/600].
CONCLUSION
Applying some method of scale control improves the economic viability of RO systems. Although there are four common ways to control scale in an RO, the most commonly employed technologies for industrial installations are the use of ion exchange softeners and chemical scale inhibitors. This paper discussed these technologies and presented two case studies that demonstrated capital costs avoidance as well as lower operating costs by replacing existing ion exchange softeners with chemical scale inhibitors. Savings were shown to be significant even for low flow rate systems. The amount of expected savings increases with feedwater hardness due to the direct, linear cost relationship of ion exchange softeners versus the substoichiometric dosage requirements of chemical scale inhibitors. The paper concludes that there can be considerable savings in operating costs by replacing softeners with scale inhibitor. Although rather inefficient and expensive, ion exchange resin will serve as a sacrificial filter helping
7
to remove particulate matter and residual coagulants. Lower overall cost and improved system performance would be achieved by installing a filter that uses a proper medium for reducing the fouling potential of the RO feedwater. A concern with using scale inhibitors that needs to be addressed is the setting up of safeguards, through either monitoring or automation, that insure the correct, prescribed amount of inhibitor to control scale is present in the feed/reject water at all times. Overcoming this issue will allow the confident use of chemical scale inhibitors to safely control scale and to improve the economic performance of RO systems.
REFERENCES
1. Byrne, W., Reverse Osmosis, A Practical Guide for Industrial Users, Tall Oaks Publishing, 1995, p. 118. 2. Ibid. p. 269 3. Degremont Water Treatment Handbook, Lavoisier Publishing, Volume 1, Sixth Edition, 1991, p. 230. 4. Kemmer, F. N., Nalco Water Handbook, McGraw Hill, 1988, p. 21.11. 5. Ibid. p. 21.11 6. RO11.2 Copyright PermaCare International 7. Ibid. 8. Exchange rate 1 = $1.50 US
NALCO
CHEMICAL
COMPANY