How to Size an RO System for High-Purity Water

Why Feedwater Data Dictates RO System Sizing

Sizing an RO system for high-purity water is a mathematical exercise driven by the impurities you must remove. There is no universal membrane count that works for every application. A system designed for a clean municipal feed may fail on a well with high iron content. The first step is gathering accurate, recent water analysis.

You need a full feedwater profile. This includes total dissolved solids, specific cations, anions, silica, and pH. If your application demands ultra-high purity, you need data on trace metals and organic compounds. Without this baseline, any calculation is a guess. The numbers below in this guide assume you have a reliable water quality report.

Prerequisites: What You Must Have Before Calculating

Before you begin the calculation process, ensure you have the following documents and data ready.

1. Feedwater Analysis: A recent lab report with all major and minor ions listed.
2. Target Purity Specification: The exact conductivity or TDS value required for the downstream process.
3. Rejection Rate Data: The manufacturer’s specification for the specific membrane type you intend to use.
4. Operating Pressure: The maximum allowable pressure for the membrane.
5. Flow Requirement: The volume of product water needed per hour or per day.

If you are missing the target purity specification, you cannot size the system. Talk to your process engineer to define the acceptance criteria. Is the water for a semiconductor wafer? A boiler? A laboratory? The purity level determines the entire architecture.

Step 1: Determine the Total Dissolved Solids Load

Calculate the total dissolved solids (TDS) in the feedwater. This is the starting mass you need to move. For example, if your feedwater has 500 mg/L of TDS and you need 10,000 liters of product water, the total mass of dissolved solids entering the system is 5 kilograms.

This number sets the scale for the membrane load. Higher TDS means more stress on the membrane and potentially a higher rejection requirement if you are targeting very low product TDS.

Step 2: Define the Target Product Purity

State your target water quality clearly. For many industrial processes, a target product TDS of 10 mg/L or lower is common. For semiconductor applications, the target might be 1 part per billion.

This target dictates your required rejection rate. Rejection rate is a percentage. It tells you how much of the solute is stopped by the membrane. If you have 500 mg/L of feed and want 10 mg/L of product, you need a rejection rate of 98%.

Step 3: Calculate the Required Rejection Rate

Use this simple formula:

Required Rejection Rate = (1 - Product TDS / Feed TDS) x 100

Using the example from Step 1:
Required Rejection Rate = (1 - 10 / 500) x 100 = 98%

Now, check your membrane specifications. If the membrane has a guaranteed rejection rate of 99% at your operating pressure and flow, you are safe. If the guaranteed rate is 95%, you are not. You must use a membrane with a higher guaranteed rejection or design a multistage system.

Step 4: Select the Membrane Type and Count

Membrane selection is driven by rejection rate, flux rate, and feedwater chemistry. For high-purity water, polyamide membranes are the standard. They offer high rejection and high flux.

To calculate the number of membranes, divide the total required flow by the flow rate per membrane at your operating conditions.

• Total Required Flow: 10,000 L/h
• Product Flow Rate per Membrane: 50 L/h (This depends on the specific membrane and pressure)
• Number of Membranes: 10,000 / 50 = 200 membranes

This is a theoretical number. In practice, you must account for fouling and scaling. You will likely need a buffer. Add 10 to 20 percent to your membrane count to handle variations in feedwater quality.

Step 5: Design the Multistage Configuration

A single stage of 200 membranes is not practical. You need a multistage configuration. Common configurations include 2, 3, or 4 parallel stages.

A 4-stage system, for example, might look like this:

Stage	Membranes per Stage	Feed Flow (L/h)	Product Flow (L/h)
1	50	10,000	8,500
2	50	8,500	7,300
3	50	7,300	6,300
4	50	6,300	5,500

Note: The numbers above are illustrative. Actual recovery rates depend on the specific membrane and pressure.

In a high-purity application, you often add a final polishing stage. This might be an ion exchange system or a second RO stage. The RO system handles the bulk removal, and the polishing stage handles the final trace removal.

Step 6: Size the Feed Pump

The feed pump must overcome the pressure drop across the membrane housing and the piping. It is not just about moving water. It is about generating pressure.

Check the pressure drop per membrane housing. If you have 40 membranes in a housing, and each housing drops 1 bar, your pump needs to handle that pressure at the required flow.

Sizing the pump incorrectly is a common mistake. An undersized pump will cause the membrane to operate at low pressure, reducing rejection. An oversized pump will cause high pressure, accelerating membrane fouling and increasing energy costs.

Step 7: Account for Concentration and Scaling

As water is removed from the feed, the concentration of dissolved solids increases. This is called concentration. If you recover too much water, the concentration becomes too high, and salts will precipitate out of the solution.

This is scaling. Scaling clogs the membrane and reduces performance. You must limit the recovery rate to prevent this.

For high-purity water, the recovery rate is often lower than for standard industrial water. You might only recover 30 to 40 percent of the feedwater in the first stage. The rest is brine.

Common Mistakes in RO System Sizing

1. Using Average Feed Data: Feedwater quality changes. A sample taken in spring may not represent summer conditions. Use the worst-case data for sizing.
2. Ignoring Pressure Drop: Many designers size the pump for flow only. They forget the pressure loss in the piping and the membrane.
3. Underestimating Brine Volume: The brine stream is a waste product. It must be handled. If you do not plan for it, you will have a disposal problem.
4. Skipping the Pilot Test: A calculated system may not perform as expected in the field. A small pilot test with the actual feedwater is the best way to verify your design.

Final Verification: Pilot Test and Commissioning

Before you install the full-scale RO system, run a pilot. Use a small membrane module with a representative feedwater sample.

Measure the product water quality. Check the conductivity. Measure the flow rate. Compare the results to your design calculations.

If the pilot results match your design, you are ready to proceed. If they do not, adjust your design. This verification step saves money and time. It prevents the frustration of a system that cannot meet its specifications.

Sizing an RO system for high-purity water is a precise task. It requires accurate data, careful calculation, and a practical understanding of membrane behavior. Follow these steps, and you will have a system that performs reliably.