System Design

What Are Important Considerations In Water System Design?

Compiled by Mike Henley



(Editor’s note: This column is based on recent discussions in the LinkedIn Ultrapure Water Group. This column seeks to accurately reflect comments from each contributor. On occasion, there may be the need to edit contributor comments for clarity or length. Readers are invited to join the Ultrapure Water Group and to participate in discussions. An important purpose of the group is to provide a forum for practical examination of issues facing endusers of high-purity water.)


Water System Design Considerations

Mike: “Recently, a professional connection emailed a question about a water treatment project he is involved with. The description below is followed by some practical questions. We welcome your input.

Case description: "We have a new water treatment project and would be interested in recommendations from members of the Ultrapure Water Interest Group. Here are some details in Table A:


Water Quality Overview



Conductivity (µS/cm)




Permeate flowrate

24.9 m3/h



Turbidity (NTU)


TDS (ppm)


Carbonate (mg/L)


Bicarbonate (mg/L)


Calcium (mg/L)


Magnesium (mg/L)


Sodium (mg/L)


Potassium (mg/)


Chloride (mg/L)


Sulfate (mg/L)


Nitrate (mg/L) 27.8


Nitrite (mg/L)


Phosphate (mg/L)  


Fluoride (mg/L)  


Silica (mg/L) 19


Boron (mg/L)  


Barium (mg/L)


Strontium (mg/L)  


Ammonia (mg/L)  


In this system, the deionized water has a conductivity of less than 0.2 µS/cm.

Site staff has requested that the water treatment plant should include chlorination, filtration (sand and activated carbon filter), reverse osmosis (RO), and deionization. Also, the RO should be 2-pass (with or without decarbonation) and the reject of the 2-pass RO would be recycled to the feed of pass. The demineralizer is included only in the mixed bed.

The individual asking for advice had three questions:

1. What is the maximum recovery that can be considered? Could a system recovery of 65% in design be considered? Is this realistic.

2. What is optimum recovery for the first and second pass?

3. What pH management is better: high acid injection and use of decarbonator and caustic injection in the second pass, or low acid injection without decarbonator?”

Fritz: “Great topic. Highly interesting! I have some more questions:

1. What is the application? How will the water be used? (Some applications have additional requirements [e.g., validation, etc.].)

2. What about storage and distribution? I always try to design the sizes of tanks and generation unit as small as possible to keep up the microbial quality of the water (if there is a requirement).

3. Do you mean chlorination or dechlorination? For operation of RO, chlorine must be removed.”

Jean-Francois: “The permeate flowrate indication is 24.9 m3/hour. Is this intermittent or 24 hours per day and 7 days per week? Besides the 0.2 µS/cm maximum conductivity, are there some other specific requirements for the produced water? TOC? Specific elements or molecules? Bacteria levels? And, as Fritz points out, what is the application?”

(Note: We have contacted the individual seeking advice for answers to these questions by Jean-Francois and Fritz, and will post them when available.)

Nissan: “A radical pretreatment might be needed instead of traditional pretreatment as the values cited can easily cause havoc on the first-pass RO membranes and pretreatment modules. I would suggest looking at Biopuremax at, which would help the RO maintain better membranes without scale, improve rejection, and stabilize the permeate flow. The cost would be fairly expensive for CAPEX but the OPEX would be extremely low versus the maintenance of a traditional pretreatment and RO system, which would be very high and expensive over the life of the system.”

Sulfate Removal in Brine Water

Paul: “Dear Ultrapure Water Group Members,

I have a customer who produces sodium hypochlorite and as part of the production process they get brine water at a concentration of 20% to 22%. They would like to reuse this water, but it has a high sulfate content (8 to 10 grains per liter). So, they have turned to us to recommend a solution to remove the sulfate so they can reuse the brine. The client reported similar applications where nanofiltration (NF) membranes were being used. However, for us this is the first time we were asked about this kind of application. We would be interested in any suggestions.”

David: “Sulfate-selective membranes have been used for quite some time on off-shore oil platforms. The application there is to use seawater to flood the wells (lift the oil). If the local geology has high levels of barium or strontium, the sulfate from the seawater would create a barium sulfate or strontium sulfate scale and plug the well within 3 to 5 years. Re-drilling is too expensive. A sulfate-selective membrane rejects sulfate, and passes the remaining salts, thus reducing the energy requirement for production. Essentially, for rejection of high levels of divalent salts and low levels of monovalent salts, I would ask your membrane supplier(s) for guidance.”

Paul:Thanks David, good comment. I´ll do that.”

Eric:Does AWC antiscalant have phosphate in it?”

David:Eric, Thank you for your question. Although, we do not discuss our formulas in detail, we use a number of different blended products. Some are phosphate-free and some contain phosphonate.”

Do Regulations Help or Hinder Water Treatment?

Question: “When making water treatment decisions, do government regulations impact your choices, or do you have the freedom to go with the best available technology (BAT)? And, a related question: Do environmental regulations help or hinder the overall treatment system design and chosen equipment technologies? For instance, the U.S. Environmental Protection Agency may call for the BAT, but their definition of BAT may preclude the use of the newest cutting-edge technology. In our discussion, government regulators should be based on the area of the world where group members are located. Please share your thoughts and observations.

Sergey: “It is rather difficult to evaluate the governmental influence from the BAT point-of-view (in Russia at least). There are certain environmental protection requirements and a demand for BAT practices in place. But, according to my experience, industrial experts are usually more knowledgeable about modern solutions for water/wastewater treatment applications than so-called "governmental experts". So, industrial experts are capable of proving that their solutions meet the BAT practice in most cases. That is why there is no real help, nor any hindrance to the overall treatment system design and chosen equipment and technologies for high-purity water treatment (power, microelectronics, and pharma) in Russia from my viewpoint.”


Mike:  “What are the key measurements to be aware of feedwater quality before the pretreatment system? What are optional, but helpful measurements? Please share your thoughts and experiences.”

Vasanthy: “Thanks for posting/starting this discussion. It is very specific to regions. The fouling index is very crucial. Hardness, TDS, and pH are also important.”

Mike:  “Vasanthy, Thanks for your response. I agree it is specific to the area where a facility is located.”

Jean-Francois:I agree with the measurements already listed by Vasanthy and her comments about the local feedwater quality are quite relevant. In Europe, for instance, feedwater conductivity can vary from 50 µS/cm at 25°C in Scotland to 3,000 µS/cm at 25°C in Barcelona area. 

I consider that the measurements to perform on feedwater depend of course largely of the nature of the purification techniques that will be used to produce pure or high-purity water. Feedwater conductivity will always be important. The measure of chlorine level in feedwater is important if you use RO membranes that are chlorine sensitive (such a polyamide-based RO membranes), but is not so important to measure if you use only DI resins for the purification process first step. The fouling index is important if you use RO membranes or DI resins as a first step in the purification process...

Those measures will help shape the nature and the magnitude of the pretreatment. For instance, if you use polyamide RO membranes as the first purification step, you know that you need to monitor free chlorine levels in feedwater to the RO membrane, as the chlorine will attack the RO membrane material; you will therefore need to install upstream the RO membrane a pretreatment step that will reduce chlorine levels at a value below the maximum chlorine concentration acceptable for the RO membrane. This information is generally provided by the RO membrane manufacturer in the membrane specifications. Based on this information, you will know that activated carbon is an acceptable pretreatment solution (among other solutions). The feedwater chlorine level (combined with information on the flowrate) will help you to define the size of the activated carbon pretreatment and also the frequency of replacement of the pretreatment.”

Ravichandran: “1. In some specific places of India, we do find CO2 levels beyond 20 ppm. (sometimes even 50 ppm.) How to treat such water as electrodeionization (EDI) will not work very efficiently.

2. Is there a way to bring down the TDS beyond RO ( RO wastes lot of water and can an NF work with 50% ionic rejection and 90 percent water recovery)?

3. Can a self-cleaning (or an automated) UF treat high silt density index (SDI) and bring it down to <10 for lab water systems. How practical and what will be the size of such systems?”

Craig:You can use gas-removal membrane technology to remove the CO2 and the Liquiflux crossflow UF design.”

Jean-Francois:There are several ways to answer question 1 asked Ravichandran. Using GTM (gas transfer membrane) technology on the permeate of the RO is probably the most elegant solution. GTM = hydrophobic membranes operating under vacuum to remove CO2 from water. Liquicel (as indicated by Craig) is one of the manufacturers of these membranes.

Ravichandran:Thanks Jean- Francois and Craig for your suggestions.

James:For a two-pass RO, increasing the pH of the first pass permeate will covert CO2 gas to the bicarbonate or carbonate form, which can be removed by the second pass.

What are your thoughts? We welcome new members to the Linkedin Ultrapure Water group, click here to join.

Log in or Subscribe to Access the Full Article

To read or download full-length articles you need a subscription to Ultrapure. Please log in or subscribe below.