This series of five articles discusses cleaning of reverse osmosis (RO) systems. The first article, published in the January/February issue of Industrial WaterWorld, discussed cleaning criteria and normalization of RO systems.
This second article addresses design and operation of a clean-in-place (CIP) skid and its integration into an RO system.
This is critical as an incorrectly designed CIP skid can shorten the life of the membrane elements since cleaning won’t be effective. In addition, the RO system also needs to be correctly designed to allow effective cleaning.
A CIP skid typically includes a tank, cleaning pump, cartridge filter and a heating/cooling device. Figure 1 illustrates a flow diagram of a CIP system.
The CIP system is connected with the RO system either with flexible hoses or fixed piping (stainless steel or FRP). Large RO systems typically have fixed piping since the cleaning flow quantities are such that flexible hoses shouldn’t be used either due to handling issues (heavy) and/or safety reasons.
The materials used for the CIP system should withstand a pH range of 1-13 and temperatures up to 122°F (50°C). They should be non-corrosive.
Cleaning tank design
A rule of thumb in sizing a cleaning tank total the empty pressure vessels’ volume and then add the volume of the feed and return hoses or pipes. For example, to clean ten 8-inch diameter pressure vessels with six elements per vessel, the following calculation would apply:
Therefore, the capacity of the cleaning tank should be 1.25 times the volume of the required cleaning solution which is approximately 700 gallons (1.5 m3).
Flat bottom tanks aren’t recommended since it isn’t possible to drain these tanks completely. The residual amount of liquid remaining in the tank becomes more contaminated over time. The cleanings will become less effective since fresh cleaning solution is contaminated with the residual liquid. In addition, the contaminated cleaning solution may cause a decline of the membrane element performance since additional foulant is introduced into the RO system.
It’s recommended to inspect the CIP tank prior to cleaning. It may be necessary to clean the tank first so any foulant is removed.
Cleaning is carried out at low pressures to minimize permeate production and redeposition of dirt on the membrane element. Little permeate should be produced in order to achieve high cross-flow velocities across the membrane surface, i.e., it should be run at low recovery. High cross-flow velocities are essential to remove foulant effectively from the membrane surface.
Sizing of the cleaning pump is dependent on the membrane element type. The cleaning pump should be sized for the flows and pressures as recommended by the membrane manufacturer, while making allowances for pressure losses in the piping and across the cartridge filter. The pump should be constructed of 316SS or nonmetallic composite plastics.
Table 1 contains information on feed pressures and flow rates that can be used for sizing of the cleaning pump.
Example cleaning pump sizing:
A two-stage RO system consists of 10 pressure vessels in stage 1, or the first bank of vessels, and five pressure vessels in stage 2. The pressure vessels contain six, 8-in., 400-sq. ft. elements each.
As per Table 1, stage 1 would require a cleaning flow of 400 gpm (91 m3/hr) and stage 2 would require 200 gpm (45.5 m3/hr) at 60 psi (4 bar). The cleaning pump can be selected based on these flow rates. Pressure of the cleaning pump will have to be higher to compensate for pressure losses in the piping and across the cartridge filter.
Cleaning at elevated temperatures (above 86°F, or 30°C) is important to remove organic fouling, biofouling and colloidal fouling effectively. Cleaning at temperatures below 68°F (20°C) isn’t recommended because of the very slow chemical kinetics at low temperatures. In addition, cleaning chemicals such as sodium lauryl sulfate might precipitate at low temperatures.
Cooling may also be required in certain geographic regions, so both heating/cooling requirements must be considered during design of the RO system and CIP skid.
The maximum temperature used for cleaning is dependent on the element type and pH of the cleaning solution. The maximum allowed temperature is 113°F (45°C) for most thin film composite membranes.
The third article of this series will discuss selection of temperature and pH for cleaning of RO membrane elements.
In addition to a correctly designed and operated CIP skid, it’s also very important the RO system is properly designed to allow effective cleaning. The following design considerations are important for optimal cleaning results.
The RO system should be designed such that concentrate and permeate produced during the cleaning can be sent to drain (during initial phase of the cleaning) and recycled back to the cleaning tank (after initial phase of the cleaning).
It’s not recommended to close the permeate valve to avoid permeate production during cleaning. Otherwise, permeate pressure will be higher than feed pressure which results in membrane damage and subsequent salt passage increases.
The cleaning solution enters the RO system in the same direction as feed water during normal operation. Reverse flow cleaning (cleaning direction is from concentrate to feed side) isn’t recommended since it can cause membrane damage.
The stages should be separately cleaned:
1. This is critical to avoid foulant removed from the first stage being deposited in the last stage. Otherwise, performance of the cleaned RO system may be the same or worse when compared with before cleaning.
2. Cleaning of the stages separately also ensures the proper cleaning flow rates for each stage.
3. Many RO systems contain different types of foulant. For instance, there may be biofouling in the first stage while the last stage has calcium carbonate scaling. In this case, an alkaline cleaning is required for stage 1 while the last stage needs an acid cleaning. An acid cleaning for the first stage isn’t recommended since the acid will react with biofilm and causes further performance decline. This topic will be addressed in subsequent articles in this series.
RO permeate should be used for both flushing and preparation of the cleaning solution. Prefiltered raw water or RO feed water should be avoided since its components may react with the cleaning solution: precipitation of foulants may occur in the membrane elements.
The return lines for both concentrate and permeate should extend into the cleaning tank such that they’re immersed. This is very important when detergents or surfactants are used as they typically cause foaming. Immersion of the return lines minimizes splashing and therefore less foam is produced. It’s important to minimize foaming since cleaning efficiency will be less effective otherwise. Anti-foam agents shouldn’t be used as they’ll foul the membrane elements, which results in permeate flow reduction.
In case of automated process control for both RO operation and cleaning, provisions need to be made in the process automation to allow for alternating circulation and soaking of the cleaning solution during the cleaning.
The design of the clean-in-place skid and the integration with the reverse osmosis system is important to obtain good cleaning results and subsequently increase the lifetime of reverse osmosis membrane elements.
What to do with an existing system that does not have a properly designed CIP system? One can make modifications to the clean-in-place skid or use one of the many off-site membrane cleaning service companies that have the proper equipment.
About the Author: Jantje Johnson is a senior development specialist at FilmTec Corp., a subsidiary of The Dow Chemical Company. With over 21 years in liquid separations, Johnson has extensive experience in membrane applications, trouble-shooting, cleaning, and system design. She holds a degree in chemical engineering from H.T.S. Groningen, The Netherlands. Contact: firstname.lastname@example.org