Introduction to Basic Parameters for the Selection and Design of Reverse Osmosis and Nanofiltration Equipment

Abstract

When the actual operating performance of reverse osmosis (RO) and nanofiltration (NF) systems fails to meet the user's expected design results, it often leads to increased operating costs, disappointment, and dissatisfaction with the supplier for the end user. These unforeseen results are partly due to some end users' lack of understanding or insufficient attention to the necessity of proper management of system operating conditions, resulting in the system failing to achieve the expected operating results. More often, however, it is due to an incomplete understanding of RO technology and equipment, leading to inadequate selection during the project preparation and planning stages. This article aims to explain some important design parameters for RO and NF systems to enable end users to select equipment based on the characteristics of RO and NF and their own needs, ensuring optimal operation of RO and NF systems.

Introduction

RO and NF application technologies are advancing rapidly. The continuous improvement in the performance of many reverse osmosis and nanofiltration membranes and the emergence of new types of reverse osmosis and nanofiltration membranes are exciting, allowing users to make appropriate choices for different applications. More comprehensive and rational RO and NF system design makes it possible to achieve better system operation and lower operating costs. However, situations often arise where the system fails to meet user requirements, leading to user disappointment with the system supplier. This can be avoided by increasing users' basic understanding of the key technical details of today's RO and NF systems. A better understanding allows for more rational purchasing planning and engineering company selection, ensuring that the required equipment, controls, and operation management training meet the requirements.

Every necessary investment decision in system expansion or new construction should undergo a comprehensive economic analysis, ensuring technical reliability and economic rationality. Understanding similar successful systems in the vicinity and consulting system suppliers will help users refine their system selection, reducing risk and cost.

Almost all industrial-scale RO and NF systems installed to date have used spiral wound membrane elements. This article will not discuss the detailed structure of various membrane elements themselves; its focus is on introducing potential RO and NF users to the considerations when purchasing RO and NF equipment.

RO and NF Basics

To optimize system design and operation, users must consider some concepts in RO and NF design. RO and NF are cross-flow filtration technologies that remove impurities from water, achieving separation capabilities up to the level of ion removal. Undoubtedly, it can remove larger substances of various kinds. However, fouling caused by colloids, scale, and microorganisms (bacteria, viruses, and algae) is the most significant problem facing the operation of RO and NF systems. To avoid these problems and unnecessary costs, adequate preventative measures should be taken during the system design phase to address this potential fouling issue. Understanding how bacteria multiply over time explains why biofouling is a critical factor for RO and NF systems; surface water, wastewater, and seawater, rich in microbial activity, are highly susceptible to fouling membrane systems. Table 1 shows that even a single surviving bacterium after pretreatment can trigger severe microbial fouling of the membrane system within a short period.

Let's analyze the changing trends of product water and concentrate flow rates when water flows through a series of membrane elements: When all the feed water is pumped into the membrane element by a high-pressure pump, it becomes product water after filtration. The remaining water becomes the concentrate for that element, which then flows into subsequent elements, becoming their feed water. Due to the reduced water volume, the flow rate decreases, while the concentration of impurities in the water continuously increases. This situation changes continuously along the flow direction in all pressure vessels until the flow rate slows down to just maintain a vortex state across the membrane surface. When performing computer-aided design calculations, it should be ensured that the flow rate and concentration in the last membrane element are kept within the minimum limits to prevent impurities from precipitating and fouling the membrane surface due to excessively low flow rates, which reduce the element's cross-flow self-cleaning ability. This is why RO and NF systems must operate within their design limits. Any changes, such as variations in feed water temperature, feed water chemistry, or feed water flow rate, or uneven and unbalanced flow distribution, will cause the RO and NF systems to operate beyond the set parameter range, leading to membrane fouling.

Feed Water

The next consideration is the feed water, including the expected volume, quality, source, and potential problems. Sometimes the current raw water flow rate is sufficient, but ensure that future increases in water usage are included in the estimated flow rate, and that changes in current water requirements are also factored into the design.

The quality requirements for the product water are determined by the final water treatment process. The cost of producing product water increases with the purity of the final product water. If the water requirements do not necessitate the high quality of RO effluent, alternative treatment methods should be carefully considered, and this factor should be fully considered in the design.

The chemical composition and characteristics of the water source must be thoroughly understood. Any system design must be able to handle best- or worst-case conditions. Well water temperature and chemical composition are generally relatively stable, with low suspended solids (SS), but each well should be considered a different source. Surface water varies greatly. While generally having low dissolved solids (TDS), it often has high levels of suspended solids (SS), organic matter, and microorganisms. Municipal water supplies typically originate from surface water, and the system's operating conditions frequently change, leading to significant variations in water quality.

Preserving raw source water quality data is crucial. Only data obtained through proper sampling and correct analytical methods can be reliable. Sample bottles should be acid-washed, and caps should be fitted with plastic gaskets. Sampling points should be in free-flowing water areas, not dead zones. A single sampling point is insufficient; a series of samples must be obtained, covering all seasons and conditions, such as after rain or during droughts. Any potential future water source must undergo thorough water quality analysis.

Pretreatment

Now that we fully understand the characteristics of the water source, proper pretreatment of the raw water becomes critical to the successful operation of the system. Typical pretreatment processes include: clarification or lime softening, multi-stage filtration such as multi-media filters, softeners, activated carbon filters, security filters, and microfiltration. An ultraviolet (UV) sterilizer is often installed after the security filter to eliminate bacterial growth. Proper analysis and rigorous pilot-scale testing can prevent many problems caused by inadequate pretreatment. All filter or softener containers in the pretreatment stage must be rubber-lined or made of corrosion-resistant materials to reduce the iron ion content in the RO and NF feed water.

Experience has shown that a more conservative design generally results in better system operation and enhanced adaptability to water quality fluctuations. Although a conservative design leads to higher initial investment costs, it reduces total operating costs over the long term. Successful experience indicates that investment and operating costs should be considered holistically, and the higher investment cost resulting from a reasonable conservative design is worthwhile. Table 2 lists reasonable design data for common equipment in RO and NF pretreatment processes.

Table 2: Pretreatment Equipment Design Parameters

A reasonable pretreatment design should fully consider the membrane cleaning frequency. Table 3 provides guidelines for evaluating the effectiveness of RO and NF pretreatment.

Table 3 Pretreatment Evaluation Criteria

The addition of chemical agents also affects pretreatment. The amount of cationic coagulants and flocculants added during clarification and filtration must be strictly controlled to prevent overdosing. If the amount of coagulants and flocculants added is appropriate, they will be discharged with the sludge during clarification or filtration. However, if excessive, residual dissolved coagulants and flocculants will adhere to the membrane surface, causing membrane fouling. Another issue is that when cationic coagulants encounter anionic scale inhibitors, a reaction often occurs, producing precipitation and fouling the membrane element. If NaHSO3 is used for dechlorination of raw water, its addition point should be as far back as possible in the entire pretreatment process, usually before the security filter. The pH value during pretreatment should also be strictly controlled, as it affects the flocculation and chlorination sterilization effects.

RO and NF Units

Now, let's consider the RO and NF units themselves. In fact, the design of RO and NF depends on the type of membrane used. There are many membrane element suppliers on the market, and membrane selection is crucial and not easy. Currently, the choice of membrane material is limited to composite membranes and cellulose acetate membranes (CA or CA-CTA). Industrial applications mostly use composite membranes, which come in many models, such as high-flux, high-desalination-rate, high-surface-area, ultra-low-pressure, extremely low-pressure, low-pollution, and high-temperature types. These membranes have different chemical compositions and constituent materials. With proper design, different membrane elements can be used simultaneously in the same system to meet different treatment requirements. When selecting the optimal membrane element for your application, you should consult with membrane element manufacturers and engineering companies to make a joint decision, as different membrane elements have optimal usage methods and applicable ranges; otherwise, various models and specifications of membrane elements would not be produced.

When designing RO and NF systems, the operating parameters of each membrane element should be considered, including the following three main parameters:

• The permeate flux of the first membrane element in the system

• The concentrate flow rate of the last membrane element in the system

• The recovery rate of each membrane element

Fluorescence is defined as the amount of water permeated per unit area per unit time. Commonly used units are gallons/square feet/day (gpd) or liters/square meter/hour (l/m²h).

Membrane element manufacturers can provide computer-aided design (CAD) programs for membrane systems, and users should retain a printed copy of the system design. These programs can also estimate the potential performance of the system under new conditions after parameter changes. Careful observation of the system's water flow distribution and operating pressure balance is essential, especially when using high-flux membrane elements. The design programs can help compare the results of using different membrane elements or combinations of membrane elements under various given test conditions.

The installation space for the equipment and system is also a crucial factor. Is there sufficient space at the end of the RO and NF unit for future membrane element replacement and installation? Is cleaning convenient? Have dead zones been eliminated? Can the system automatically flush (displace or vent) during startup or automatically perform low-pressure flushing during shutdown? In standby mode, can the system guarantee against water loss, and can soft start be achieved to avoid shock damage to the membrane elements? These are all issues that should be considered when selecting RO and NF equipment. High-pressure piping for RO and NF systems should be made of stainless steel or other corrosion-resistant materials.

Waste

When designing RO and NF systems, the treatment of waste (concentrate) is crucial, especially for large membrane systems exceeding 500 m³/h. Optimistically, the wastewater from RO and NF is simply concentrated feed water and therefore not a major concern. However, wastewater typically accounts for about 25% of the influent volume, which is a significant amount. A well-considered design should address various concentrate treatment possibilities and maximize the utilization of concentrate before discharge.

Cleaning

After a period of operation, the membranes must be cleaned. For operators, this task should be as simple as possible. Generally, membrane water treatment systems should have an in-situ cleaning system connected to the RO and NF units via rigid pipes or quick-connect hoses. One in-situ cleaning system can serve multiple RO and NF units. The diagram below shows a standard in-situ cleaning system. The selection of the cleaning pump should ensure the required influent flow rate for each pressure vessel in the first stage. The more turbulent the cleaning solution, the better the cleaning effect. To improve cleaning efficiency, multi-stage reverse osmosis systems should be cleaned in stages, targeting different fouling conditions. Large-scale water treatment systems should consider setting up an independent unit cleaning method evaluation and testing platform. When only one or two membrane elements at the front end are fouled, cleaning all membranes in the entire system will only cause the contaminants from the front-end cleaning to flow into subsequent, less fouled membrane elements, reducing the system's cleaning efficiency. A separate element cleaning testing platform can avoid this disadvantage and can also be used to determine the operational performance of each individual element.

Conclusion

When you want to use RO and NF systems in your plant, there are many issues to consider. It is recommended to organize a project team to spend sufficient time gaining a comprehensive understanding of the entire process and examining all treatment technologies that can meet the water quantity and quality requirements. If you choose RO and NF technology, you can arrange for the operations supervisor to receive necessary on-site operational training at other companies that already have RO and NF systems. This will allow them to understand the actual operation of this new technology and avoid many problems and economic losses that may occur after the system is built and put into operation.

Problems during operation are always inevitable, but costly mistakes can be avoided as much as possible. By continuously increasing understanding of this technology, you will gain a thorough understanding of its advantages and disadvantages, ensuring the long-term safe, economical, stable, and reliable operation of RO and NF equipment.