Enhancing RO system performance utilizing antiscalants. International Environmental Technology Co

These concentrations can be calculated from the concentrations in the bulk concentrate using the concentration polarization factor.

As a result of this work, the following correlations for the maximum permissible recovery, Y_i as constrained by each specific constituent in the RO feed, the feed ionic strength, I and the characteristic performance index, α_i of the applied antiscalant were obtained:

151

where Y₁–Y₅ is maximum feasible recovery (%) due to calcium sulphate, barium sulphate, strontium sulphate, calcium fluoride and silica, respectively; α₁–α₄ — characteristic performance index (…* K_sp) of the applied antiscalant with respect to calcium sulphate, barium sulphate, strontium sulphate and calcium fluoride respectively; β — concentration polarization in final elements of the last array; I — ionic strength of the RO feed water; [   ] — concentration of the specific ionic species in the RO feed water, in mg/L ion; [SiO₂] — concentration of silica in the RO feed water, in mg/L; t — RO feed water temperature, °C.

Based on pure chemistry considerations, the maximum allowable system recovery, Y_m, is given by the following relation:

Y_m = minimum [Y₁ , Y₂ , Y₃ , Y₄ , Y₅]                 (6)

Moreover, the system process requirements and the associated engineering and economic

152

considerations and constraints may dictate another limiting system recovery. There is a design choice here and the designer will have to arrive at the most appropriate selection for the recovery of the system under consideration. Once the system recovery, Y_s, is defined, the corresponding pH of the RO feed water may be estimated from the following relation:

pH ≤11.65 + LSI^ı − log₁₀ {(CaH)⋅(ALK)

⋅(CF)2.9 ⋅(TDS)0.1}+13.12 log10 §¨t + 273·¸ (7)

© 300 ¹

where LSI^σ is maximum allowable LSI for the applied antiscalant; (CaH) — calcium hardness of the RO feed water, mg/L as CaCO₃; (ALK) — M alkalinity of the RO feed water, mg/L as CaCO₃; CF — concentration factor = β · (1 – SP · Y_s)/(1 – Y_s); SP — salt passage, decimal; TDS — minimum [RO feed water TDS, (5000/CF)].

It is important to note that the equilibrium constants for the sparingly soluble salts and the values for the different components of LSI were obtained from references [2,3].

5.  Antiscalant dosage

As an effective scale control is essential in RO systems, the system designer must consider his design decision. Typically the use of an antiscalant will provide a higher recovery than acid alone. The antiscalant complexes the cations species to prevent precipitation, whereas the acid reduces the concentrations of the anions species such as bicarbonate to prevent calcium carbonate precipitation. Both mechanisms are different and their relative share can be optimized to maximize system recovery and reduce cost. The amount of antiscalant addition is determined by the limiting salt, pH of the acidified RO feed water and system recovery. The limiting salt can be determined from the feed water chemical analysis and the solubility products of potential limiting salts. The factors, which must be considered when making design decision concerning acid and/or antiscalant addition, are:

•  Module configuration (spiral wound, hollow fine fiber, tubular).

•  Material of construction for the membrane elements.

•  Feed water quality.

•  System recovery.

•  Required output quality.

Until early eighties, the antiscalant manufacturers referred to the recommended antiscalant dosage as ppm dosed to the feed stream. This approach seemed to face problems in several applications. In 1984 some of the manufacturers adopted the approach of reporting the recommended dosage as ppm antiscalant concentration in the brine stream. The application of the second approach meant on many occasions lower antiscalant dosages in the feed stream at high recoveries and vice versa. This contradicts the actual process requirement for increased antiscalant consumption at higher recoveries.

At present some manufacturer-provided computer projection programs are available for estimating the required antiscalant dose for a given recovery, water quality, and membrane. Although technically sophisticated, the models built in these projection programs are largely dependent on the abovementioned approaches with some improvement. The author feels that these models still need further development as to serve the industry requirement and desire for optimal antiscalant dosage.

6.  Conclusions

The primary objective of the study was to find suitable means of calculating the maximum feasible recovery for the sparingly soluble constituents taking into consideration the continued improvements in antiscalants performance. This objective was realized as per the relations (1)–(5). Moreover, the recommended value for the pH of the acidified RO feed water may be estimated using the relation (7). All of these relations were derived during the course of this work.

The existing problems associated with the estimation