Brian L. Wise, Nanostone Water Inc., 10250 Valley View Road, Suite 143, Eden Prairie, MN, USA, 55344, firstname.lastname@example.org, Ph; 612.294.3900
Aditya Kumar, Nanostone Water Inc., Eden Prairie, MN, USA
Dr. Stanton Smith, Nanostone Water Inc., Eden Prairie, MN, USA
Daniel Hugaboom P.E., Carollo Engineers, Boise, ID, USA
Ceramic membranes continue to make improvements in surface area and production costs which are making them cost competitive with polymeric membranes. The system designs employed by many ceramic membrane suppliers tend to use aggressive hydraulic cleaning methods and high rates of crossflow in order to increase flux in an effort to make the system overall more cost competitive. These aggressive designs are effective methods for flux maintenance but present a design challenge and a significant capital cost premium for full scale membrane plants.
The latest development of Nanostone Water is a ceramic monolith membrane design that improves the economics of ceramic membranes allowing comparable capital expenditures to full scale polymeric membrane systems without the use of aggressive system designs. Various hydraulic cleaning methods were studied during the optimization process at a variety of field applications including filter backwash water recovery, cooling tower blowdown recovery, and direct coagulation surface water treatment. Results of field studies show that a conventional backwash method at a moderate flux rate can be effective as flux maintenance for ceramic membranes.
This paper will present hydraulic cleaning design summary options for ceramic and polymeric low pressure membranes as well as field testing data comparing more conventional to more aggressive methods. Also presented will be several retrofit studies of Nanostone’s ceramic monolith membrane replacing polymeric hollow fiber modules.
Ceramic Ultrafiltration and Microfiltration (UF/MF) membranes continue to make improvements in surface area and lower prices resulting from innovative production techniques. These advancements allow ceramics to be cost competitive with polymeric UF/MF membranes. The invention of the high surface area monolith structure (Goldsmith, 1988) was a major breakthrough in ceramic module construction to reduce the cost and broaden the applications of ceramics. Additional new developments in ceramic membranes by Nanostone Water, Inc. include a patent pending design (Göbbert and Volz, 2010) where individual ceramic segments are potted together forming the monolith structure, significantly reducing the production costs compared to other ceramic membrane modules. These advances have made the membrane portion of a UF/MF system even more cost competitive.
Figure 1: Monolith CM Module
In many cases, a ceramic UF/MF membrane system can operate with less pretreatment such as eliminating the need for a clarification system which has a significant impact on the initial capital cost, operating cost, and foot print. With the higher total suspended solids (TSS) tolerance of ceramic membranes and the ability to utilize more aggressive chemical and hydraulic cleaning methods, the risk of irreversible fouling is less than with polymeric UF/MF membranes. However, in a more direct comparison of a polymeric to a ceramic UF/MF system, the hydraulic cleaning methods can have a significant impact in the overall cost of the system.
Hydraulic Cleaning Methods for UF/MF Membranes
The use of traditional polymeric UF/MF membranes (e.g. PVDF, PES…) has been established in the industrialized nations. UF/MF membrane systems have seen increasing use in recent years due to decreasing prices resulting from manufacturing advances and technological improvement in packing densities and hollow fiber designs. For most industrial and small to mid-size municipal projects, the pressurized hollow fiber UF/MF membrane module is a commonly applied technology. The pressurized UF/MF modules cover a wide range of membrane types as well as both inside / out and outside / in flow configurations. The hydraulic cleaning methods used by polymeric hollow fiber UF/MF membranes are quite varied.. With most outside/in membranes, the backwash cycles typically use air scour to help break up the cake layer within the fiber bundle. A backwash sequence is also used where permeate is pumped in reverse direction through the module to remove suspended solids built up during filtration mode. Inside / out polymeric membranes typically rely on the backwash flow alone to remove suspended solids between filtration cycles. Backwash flow rates have a wide range of variation between manufacturers with some requiring as little as 10% of filtration flow while others require up to 400% of filtration flow. Feed forward flush is another option often used as part of the backwash sequence to help move suspended solids out of the membrane modules. Some suppliers use a drain down step as well. Backwash and drain down steps are used simultaneously with or in sequence with air scour. Due to the potential for fiber breakage or other damage to the polymeric hollow fiber UF/MF membranes, backwash pumps are typically controlled with a variable frequency drive (VFD) to provide a gentle increase in flow and pressure during backwash. Because the hydraulic cleaning design is well controlled with smooth transitions between cycles, it is common to use plastic piping such as Polyvinyl chloride (PVC) which helps keep the capital costs low.
With many pressurized inside/out type ceramic UF/MF membrane systems, the design approach typically utilizes a high flux design with high crossflow velocities. For smaller industrial systems, this approach can help keep the overall costs lower when the ceramic membranes represent a significant portion of the system cost. With these designs, there is typically a concentrate bleed rate that sets the recovery rate of the system. The crossflow velocity helps keep suspended solids from building up on the membrane surface too quickly. Periodically, a backwash cycle is employed to reverse direction of permeate flow while the crossflow stream continues to operate. The periodic backwash moves suspended material off the membrane surface where it is swept away by the high crossflow in the channels. This backwash method is often driven by compressed air either directly or through a diaphragm valve or hydro pneumatic vessel to deliver a fast step change increase in both pressure and flow for short durations.
Figure 2: Typical Flow Schematic of Crossflow Ceramic UF/MF System
For small scale systems with high fouling potential or high suspended solids the, crossflow design can be effective. However, it is generally more expensive considering the additional crossflow pump cost. The typical crossflow velocities designed by ceramic UF/MF suppliers is somewhere between 1 and 3 meters (3 to 10 feet per second). To achieve this velocity through the capillary channels of 2 to 3 millimeters, the recirculation rate is often 10 to 35 times higher than the filtration flow rate. System designs with high crossflow rates typically have some amount of ceramic membranes in series to minimize the pumping costs. In the example above for a system with a feed flow of 500 GPM (114 m3/hr) and producing 475 GPM (108 m3/hr) of permeate the design outline has two ceramic modules in series. Even with two modules in series a recirculation rate of 2,800-8,500 GPM (636-1,930 m3/hr) is required to reach a crossflow velocity of 1-3 meters per second (3-10 feet per second). The crossflow pump and piping involved at these recirculation flow rates would be very costly as compared to the piping and pump designed only for the throughput volume of the plant.
In other cases ceramic membrane systems can be designed for dead end filtration without high velocity crossflow. Flux rates in these applications tend to be lower as compared to designs with high crossflow but may be more cost effective without the high crossflow equipment required. Hydraulic cleaning methods for ceramic membranes in dead end mode of operation will also typically employ a fast step change with high pressure, high flow backwash for short durations typically with compressed air as the motive force. And in some cases the system designs use compressed air to purge the dislodged solids from the capillary channels in a fast draining step. While these high pressure hydraulic cleaning methods are effective for ceramic membranes, they do represent design and cost challenges to the overall system. For example, in cases where compressed air is introduced into the piping system for backwash, the materials of construction need to be appropriately rated which generally requires non-plastic pipe. Compared to a polymeric UF/MF system designed with nearly 100% plastic piping, this can be a significant cost difference. In addition, with the high pressure and high rate backwash flow, the size of the permeate and drain pipes for the system need to be significantly larger than they would be for systems with lower backwash rates.
For ceramic membranes the question is what is the right balance between effective hydraulic cleaning and overall cost of the system? Nanostone water conducted a series of experiments to evaluate different methods of hydraulic cleaning to try and answer this question.
Hydraulic Cleaning Pilot Study #1 for Ceramic Membrane
The first pilot test location used to study hydraulic cleaning methods is a direct river water feed location in the western USA. The river water feed is generally < 15 NTU turbidity with total organic carbon (TOC) values < 4 mg/L. The local drinking water utility allowed Nanostone to setup their pilot system on site for long term product and process design development work. Carollo Engineers was contracted to support the effort with consultation and periodic operations of the pilot system. The pilot study began with several experiments to study different ceramic membrane configurations and various operating procedures including hydraulic and chemical cleaning. Earlier experiments at the site, not discussed in this paper, showed that fast step change high pressure, high flow backwash methods had similar results for both one and eight second durations. Earlier experiments also showed that a feed water flush following backwash was more effective than without the flush. The results of these experiments provided a base-line operation using the fast step change high pressure / high flow backwash methods for short durations followed by a feed water flush at 100% of filtration flow for longer durations.
The pilot system used is a 20-foot containerized system with one process train designed to operate with a single ceramic test membrane with 3m2 (32ft2) of active surface area housed in a 100mm (4-inch) diameter PVC vessel. This is a representative version of the full scale module from Nanostone with 20m2 (215ft2) of active surface area housed in a 10 BAR (150 PSI) rated fiberglass reinforced plastic (FRP) membrane vessel. In both cases the membrane has a nominal rating of 30 nm (0.03 micron).
Figure 3: Nanostone Water Ceramic Pilot System
The first test condition was a fast step change high pressure / high flow backwash for a short duration followed by a feed water flush. The filtration cycle was set for dead end filtration at a 60 minute cycle. The filtration flux was set constant at 170 L/hr per m2 of membrane (LMH) or 100 Gallons per day per ft2 of membrane (GFD). 1 mg/L as Al+3 of polyaluminium chloride (PACl) is used as a flux enhancing coagulant. A drinking water plant using a polymeric UF membrane system operating on this same water source uses the same dosage of coagulant. The backwash method used for the test was compressed air set for 5 BAR (72 PSI) for a duration of 1 second. The estimated backwash flow rate was 14 times higher than the filtration flow or 2,400 LMH (1,400 GFD). Following the backwash cycle was a feed water flush at 100% of filtration flow for a 40 second duration. The estimated water recovery in this scenario is 98%. The test was run for 120 hours with an average pressure increase of 0.0017 BAR per hour or 0.025 PSI per hour.
Figure 4: Pilot Test Results With Fast Acting High Rate Backwash Method
The next test condition was a conventional method used by most polymeric UF/MF membranes with a gentle increase of low pressure / low flow backwash for a longer duration followed by a feed water flush. The filtration cycle was set for dead end filtration at a 20 minute cycle. The filtration flux was set at the same 170 LMH (100 GFD) with 1 mg/L as Al+3 of PACl coagulant dosage. The backwash method used for the test was a backwash pump set to reach a flow of 2 times the filtration flow or 340 LMH (200 GFD) for a duration of 11 seconds. Following the backwash cycle was a feed water flush at 100% of filtration flow for a 10 second duration. The estimated water recovery in this scenario is 97%. With the shorter filtration cycles in this test the pressure increase during each filtration cycle is lower than with the first test. The test was run for just over 2 days with an average pressure increase of 0.003 BAR per hour or 0.04 PSI per hour. The operation is stable and would be manageable with routine maintenance cleanings. The more conventional backwash method has a higher pressure increase over time in comparison to the more aggressive high flow backwash data even though the recovery rate was 1% lower. The tradeoff in the conventional backwash method is slightly more frequent cleaning and less recovery. However, the system designed with the conventional backwashing method would allow smaller sized plastic pipes to be used to reduce the cost of the overall system.
Figure 5: Pilot Test Results Conventional Low Rate Backwash Method
The third test incorporated the longer duration, frequent backwash at low flow while introducing periods of the fast step change backwash method described in the first test condition. This fast step change of backwash water pressure and flow can be accomplished by ramping the backwash pump ramp against a closed valve to build pressure prior to opening the valve. For the pilot system, however, this scenario was accomplished with the same compressed air backwash system described for the first test condition but at a much lower pressure. The backwash flow rate was only 2 times the filtration flow rate or 340 LMH (200 GFD) for a duration of 10 seconds. The filtration cycle was set for dead end filtration for 20 minutes. The filtration flux was set at the same 170 LMH (100 GFD) with 1 mg/L as Al+3 of PACl coagulant dosage. Subsequent to the backwash cycle, a feed water flush at 100% of filtration flow for a 40 second duration was applied. The estimated water recovery in this scenario is 95%. The test was run for just over 100 hours with an average pressure increase of 0.002 BAR per hour or 0.03 PSI per hour. This hybrid design approach is an effective way to get the benefits of the fast step change pressure and flow increase but without the high capital cost impact of high flow and high pressure backwash systems.
Figure 6: Pilot Test Results With Hybrid Fast Step Change Low Flow Rate Backwash Method
For the purposes of these experiments the operating flux was conservatively held at a consistent level of 170 LMH (100 GFD). Future experiments are planned for this pilot site to further explore the hybrid backwash operation at higher recoveries and higher flux rates showing longer term stability with a maintenance cleaning regime in place.
Table 1: Pilot Study #1 Results Summary Of Hydraulic Cleaning Methods
|Test #1||Test #2||Test #3|
|Filtration Cycle Duration||60 min||20 min||20 min|
|Backwash Method||Fast Step Change Pressure and Flow Increase||Slow Pressure and Flow Increase||Fast Step Change Pressure and Flow Increase|
|Backwash Pressure||5 BAR (72 PSI)||0.8-1.0 BAR (12-15 PSI||0.8-1.0 BAR (12-15 PSI|
|Backwash Flux||2400 LMH (1400 GFD)||340 LMH (200 GFD)||340 LMH (200 GFD)|
|Backwash Duration||1 second||11 second||10 second|
|Feed Flush Flow||100% of Filtration Flow||100% of Filtration Flow||100% of Filtration Flow|
|Feed Flush Duration||40 second||10 second||40 second|
|NDP Pressure Increase||0.0017 BAR/hr
Hydraulic Cleaning Pilot Study #2 for Ceramic Membrane
At a second pilot location at a power plant in the southwest USA hydraulic cleaning methods for ceramic membrane was also studied. In this case, a pilot study used the same ceramic ultrafiltration membrane as was used in pilot study #1. In pilot study #2 the UF membrane is used in a polishing step after a chemical precipitation process to reduce silica from the cooling tower blowdown stream. This water is characterized by high pH, mineral scaling potential, and high turbidity. Turbidity values over 400 NTU resulting from the precipitation process were consistently observed. Silica values in the blowdown stream reached 120 mg/L and the pilot regularly demonstrated product water < 25 mg/L of silica, which was considered adequate to feed a downstream reverse osmosis system to allow the plant to reuse the water. Stable operation was demonstrated at 170 LMH (100 GFD) filtration flux operating with just 20% of concentrate recycle to keep the high TSS moving through the membrane and going back to the chemical reaction tank located upstream of the membrane.
Backwash operation for initial pilot work at this location used the fast step change high pressure / high flow method with short duration. After stable operation and water quality goals were demonstrated, a series of experiments were performed to investigate hydraulic cleaning methods. In the initial test results shown below, the pilot used the conventional backwash method with a standard pump running at 4 times filtration flow with backwash frequencies at 20 minutes and backwash duration at 10 seconds. A feed flush at 100% of filtration flow for 46 seconds was also utilized.
Figure 7: Post Chemical Precipitation Pilot Test Results with Conventional Slow Ramp Up
Backwash Pump At 4 Times Filtration Flow
The system was stable over time with CEB frequencies every 24 hours but the pressure increase between CEB cycles was higher than desired. After close examination of the backwash flow profile, it appeared that the slower pressure and flow increase of a standard backwash pump was not as effective as the faster step change high pressure and flow backwash tests used earlier. For a better comparison of the two methods, the hybrid fast step change backwash test condition was implemented but with lower pressures than previous operation delivering an effective flow rate at only 4 times the filtration flow point. The pressure required to reach the 4 times filtration flow was only 0.8-1.0 BAR (12-15 PSI) as compared to previous fast step change backwashes with 5 BAR (72 PSI). The results show that a faster application of backwash water was more effective than a slow increase of backwash flow and pressure in this application. No other process parameters were changed with the membrane or the chemical precipitation so the improvement in operation is assumed to be primarily from the change in backwash method. The fast step change of backwash water pressure and flow can be delivered with a compressed air tank or by using a pump ramping up to pressure prior to valve activation. The CEB cycles were maintained at 24 hour periods. However, with only minor pressure increases observed, there is opportunity to extend the acid CEB cycles beyond 24 hours.
Figure 8: Post Chemical Precipitation Pilot Test Results With Fast Ramp Up Backwash at 4 Times Filtration Flow
Table 2: Pilot Study #2 Results Summary Of Hydraulic Cleaning Methods
|Test #1||Test #2|
|Filtration Cycle Duration||20 min||20 min|
|Backwash Method||Slow Pressure and Flow Increase||Fast Step Change Pressure and Flow Increase|
|Backwash Pressure||2-3 BAR (30-40 PSI)||0.8-1.0 BAR (12-15 PSI)|
|Backwash Flux||680 LMH (400 GFD)||680 LMH (400 GFD)|
|Backwash Duration||10 second||10 second|
|Feed Flush Flow||100% of Filtration Flow||100% of Filtration Flow|
|Feed Flush Duration||46 second||46 second|
|Average NDP||1 BAR (14.5 PSI)||0.52 BAR (7.5 PSI)|
The experiments show the ceramic UF membranes demonstrated successful operation both in dead end mode and with minimal recycle rates (< 20% of filtration flow) for high TSS applications using backwash flow rates between 2 and 4 times the filtration flow. It is feasible, then, that this membrane may be able to retrofit some of the pressurized hollow fiber polymeric UF membranes currently in the market place.
Process Design Comparison: Nanostone Ceramic UF to Various Polymeric UF/MF
Incorporating the experimental results previously described, the analysis below is a paper exercise comparing the process designs of the ceramic UF to several polymeric pressurized UF modules to demonstrate the feasibility of retrofits. The design scenario outlined assumes a theoretical system producing a peak flow of 500 GPM (114 m3/hr). The selected peak flux is a nominal value of 60 LMH (35 GFD) for polymeric UF membranes that would be applicable for a variety of application. A system design outline incorporates parameters published by the manufacturer including surface area and design outlines for backwash.
Table 3: Process Design Comparisons of Nanostone ceramic UF to various polymeric UF/MF for 500 GPM (114 m3/hr) Peak Flow System
|Case #1||Case #2||Case #3|
|Hollow Fiber #1||Nanostone CM-131TM||Hollow Fiber #2||Nanostone CM-131TM||Hollow Fiber #3||Nanostone CM-131TM|
|Number of Modules||32||32||40||36||32||32|
|Area per module (m2/ft2)||60/645||20/215||46/500||20/215||77/829||20/215|
|Module Height (Inch/mm)||66/1676||67.5/1715||63/1600||67.5/1715||93/2359||67.5/1715|
|Clean Permeability (LMH/BAR)||350-400||800-900||350-400||800-900||200-250||800-900|
|Backwash Flow (m3/hr / GPM)||435/1917||435/1917||372/1639||372/1639||336/1478||336/1478|
|Backwash Flux (LMH/GFD)||230/135||690/405||200/117||512/301||136/80||516/303|
|Backwash Flow Multiple of Filtration Flow||3.8||3.8||3.3||3.3||2.9||2.9|
|Air Scour Used||No||No||No||No||Yes||No|
In case #1, the PES inside/out model has 60 m2 (645 ft2) or 3 times the area of the Nanostone ceramic membrane. The permeability of the ceramic membrane is typically 800-900 LMH/BAR and the PES membrane is typically 350-400 LMH/BAR. Given the higher permeability of the ceramic between 2 and 2.5 times the PES and a flux rate 3 times higher, the feed pressure is anticipated to be higher for the ceramic membrane in this example. Depending on the fouling profile of the application and design of the feed pump pressure an upgrade maybe required to a higher pressure design. The backwash rate for the UF module in case #1 is higher compared to other polymeric UF membranes. The flux multiple of 3.8 times the filtration flow is within the stated backwash range for the Nanostone ceramic membrane of 2-4 times. Therefore it is clear that the existing backwash system would be suitable for use in a ceramic retrofit.
In Case #2 another PES inside/out module is compared. In this case, the area is only 46 m2 (500 ft2). Consequently, fewer ceramic modules could be required to retrofit the polymeric modules in this model. In this case the flux of the ceramic membrane is 2.5 times the PES membrane which closely matches the permeability advantage. As a result the feed pump would likely have sufficient pressure to run the ceramic membrane at the stated fluxes without needing to increase the design pressure. The backwash rate for this case is lower than for case #1, but still yields a backwash flux multiple of 3.3 times the filtration flow. This is also in the design range of 2-4 times filtration flow stated above for the Nanostone ceramic membrane.
Case #3 is an outside/in PVDF hollow fiber membrane with 77 m2 (829 ft2) of area. In this case, the module count for the ceramic retrofit remains the same since the flux increase is 3.8 times higher. The permeability advantage of the ceramic membrane in this case is 3.5 to 4 times higher. Depending on the application and the fouling potential, the feed pump would likely provide adequate pressure. The backwash rate for this PVDF module is lower than the other two cases given that this module uses air scour in the backwash mode.. However, the backwash rate is 2.9 times the filtration flow rate and so does fall within the typical design range for the ceramic membrane. Again, it is practical to assume that the existing backwash system would be sufficient in a ceramic retrofit. There are several PVDF outside / in hollow fiber modules on the market with low backwash rates that rely primarily on air scour to dislodge solids from the fiber bundle. In these cases, changing the backwash system to a larger pump and potentially large pipe line would be required.
These three cases illustrate that a ceramic membrane retrofit could utilize existing conventional backwash designs in all three hollow fiber membranes modeled. Each retrofit would need to be evaluated on a case by case basis to determine if the feed pump and backwash system could be employed for optimum performance. The module heights are very similar. Depending on the design of the piping connections to the modules a direct module retrofit maybe possible but would need to be evaluated for each case. Replacing the existing membrane rack with a new ceramic membrane rack may be more practical to limit the complexity of field work. Regardless of the approach, much of the infrastructure of the incumbent system could be utilized.
Summary and Outlook
With more price competitive ceramic UF membrane modules now available, there is now opportunity to retrofit many mainstream water and waste water applications where polymeric UF/MF membranes are typically applied. Ceramic membranes have several performance advantages over polymeric membranes such as a longer life span, higher pressure tolerances and greater chemical stability. However, the overall system design around the ceramic membranes needs to be cost effective as well. The aggressive hydraulic cleaning methods typically used for ceramic membranes certainly are effective and in many specialty applications they are necessary. The hybrid hydraulic cleaning methods developed with conventional flows and pressures but with applying a faster step change represents an approach to achieve an optimized hydraulic cleaning method without high capital cost of the of the methods traditionally applied to ceramic UF/MF membranes.
Goldsmith, Robert (1988), US Patent 4781831
Göbbert, Christian and Volz, Manfred (2010), US Patent Application 20130153485 Priority Date 22 February, 2010.