Navy ships generate a variety of liquid wastes: bilge water, black water, gray water, shipboard “industrial” wastes, and solid residuals from existing treatment systems. Many of the current Navy waste treatment systems would benefit from the efficient removal of solids (e.g., oily waste ceramic membranes are susceptible to face plugging and mechanical failure). Available solids removal technologies have not been particularly effective, necessitating the development of improved solids removal technologies. The High-Shear Rotary Membrane System (HSRMS), which consists of stacked, rotating membrane disks, has shown superior abilities to separate and concentrate Navy and non-Navy wastewater solids (e.g., oily wastes, underwater hull cleaning sludge, non-skid deck cleaning wastewater, metal hydroxide suspensions). However, the HSRMS has been confined to land-based applications where space is not a critical design consideration.

This project sought to overcome limitations and improve the HSRMS for shipboard wastewater and solids residual treatment by addressing the following objectives:

  1. Increase HSRMS permeate flux to decrease system size by employing backpulsing and continuous membrane physical surface cleaning. In addition to increasing the flux, these improvements should allow the use of larger membrane disks, decrease cleaning/maintenance frequency/residuals, and increase membrane life.
  2. Increase “active” membrane packing density (active membrane surface area/system ft2 and ft3) by increasing the membrane diameter and/or employing overlapping disks. Single-shaft rotating membranes increase shear induced scouring intensity at distances farther from the disk center. The inner portion of the disk has less shear applied, thus less permeate produced compared to the more “active” outer regions. Overlapping disks should create a more uniform distribution of turbulence/shear over the membrane surface, increasing the “active” membrane area.
  3.  Conceptually design, fabricate, and test a laboratory-scale HSRMS that incorporates a combination of backpulsing, continuous membrane cleaning, larger disks, and disk overlap. An improved HSRMS placed shipboard will have increased waste treatment throughput, a smaller footprint, and may be constructed of lighter weight/cheaper materials.

Technical Approach

Four Navy wastewaters were selected for HSRMS evaluation and prioritized by the Naval Sea Systems Command (NAVSEA) Technical Warrant Holder (TWH) for Environmental Systems based on current and anticipated regulatory requirements for discharge to the sea (highest to lowest priority): (1) bilge water, (2) black water (black water-gray water), and (3) plasma arc waste destruction system (PAWDS)/metal hydroxide and biosolids (tie). Commercially available membrane disks of the following materials and pore sizes were selected for baseline evaluation with the four Navy wastes: stainless steel (SS) (0.1 micrometer (µm), 0.5 µm, and 3 µm pore sizes), polytetrafluoroethylene (PTFE) (1-2 µm pore size), and ceramic (7 nanometer (nm), 30 nm, 60 nm, 0.2 µm, 0.5 µm, and 2 µm pore sizes). As a result of the baseline tests and initial backpulsing evaluations performed during bilge water treatment, the 60 nm pore size ceramic disk was selected to investigate the effects of membrane disk diameter, rotational speed, backpulsing, and continuous surface cleaning on HSRMS treatment of Navy waste streams.


The 60 nm ceramic membrane manufactured by Keramische Folien GmbH (KERAFOL) produced the greatest flux for bilge water and black water. For the metal hydroxide waste, flux performance was influenced by pore size. Although the stainless steel and PTFE membranes with larger pores produced higher fluxes, the ceramic membranes with larger pore sizes (0.2 µm, 0.5 µm, and 2.0 µm) may perform as well as the other membrane materials and future investigation would be warranted for this waste. Biosolids were not evaluated for all membranes; input from the NAVSEA TWH for Environmental Systems indicated that black water results would likely be representative of biosolids.

Numerical results varied by wastewater, but the permeate flow rate benefit of increasing membrane diameter was demonstrated for bilge water, black water, and metal hydroxide wastewater. Larger membranes at lower rotations (e.g., 14.7-inch at 500 revolutions per minute (rpm)) produced permeate at flow rates equal to or greater than smaller membranes at higher rotations (e.g., 10.5-inch at 1150 rpm). KERAFOL manufactures membrane disks in 12.3-inch and 14.7-inch diameters (both disk sizes have inner diameters of 3.6 inches) and recommends rotational speeds no greater than 500 rpm for both. The system benefits of fewer 14.7-inch disks (2.1 ft2/disk using 5-inch hub) versus more 12.3-inch disks (1.4 ft2/disk using 5-inch hub) would need to be weighed in designing the improved HSRMS. An overlapping disk design, such as the system commercialized by Andritz Separations, should also be considered as a way to increase the active membrane packing density.

The effects of backpulsing and continuous surface cleaning on membrane performance also varied by wastewater. For treatment of bilge water, the highest priority wastewater, backpulsing did not improve flux. For single-disk HSRMS bilge water treatment, continuous surface cleaning showed great promise in increasing flux for all rotations evaluated (100, 250, 500, 750, and 1150 rpm). However, when scaled up to a multi-disk system, continuous surface cleaning had a significant flux benefit at the lower rotations, but only increased flux by 10% at 500 rpm. Therefore, for improved HSRMS treatment of bilge water, the continuous surface cleaning flux benefit could be matched by increasing membrane area. The decision is best left up to the system manufacturer as each option possesses obstacles that would need to be overcome; the former would have additional maintenance/consumables and require additional evaluation to determine attachment lifespan, while the latter may involve increased motor and/or membrane stacks to meet permeate production requirements. A conservative flux value for bilge water treatment at 500 rpm is 150 gallons per square foot per day (GFD).

The effects of backpulsing and continuous surface cleaning on single-disk HSRMS treatment of black water and metal hydroxide wastewater were not determined, but based on the bilge water results, it is recommended that multi-disk HSRMS evaluation of black water and metal hydroxide wastewater using backpulsing and continuous surface cleaning be performed before proceeding with improved HSRMS design for these wastes.

At 500 rpm, single-disk baseline black water flux at 500 rpm was approximately 75 GFD and was increased approximately 55% with backpulse and approximately 37% with continuous surface cleaning. At 500 rpm, single-disk baseline aluminum hydroxide flux was approximately 150 GFD and increased approximately 20% with backpulse; continuous surface cleaning was not investigated. The effects of scale-up on the flux benefits of backpulsing and continuous surface cleaning could influence the design of an improved HSRMS for black water or inorganic particulate waste treatment.


The project results demonstrate the ability of the HSRMS to effectively treat a variety of solids-bearing Navy wastes and the impacts of potential flux enhancements (continuous surface cleaning, backpulsing), disk diameter, rotation, membrane selection, and membrane surface turbulence enhancements (wagon wheels) on minimizing system footprint and volume. The use of advanced HSRMS technology will enable more efficient solids removal prior to existing shipboard treatment processes, direct replacement of problematic treatment systems with the more robust, higher efficiency system, and concentration of sludge, waste oil, and process residuals. The ultimate benefit to the Department of Defense is a robust “barrier” technology that is easy to operate, not labor intensive, is capable of being cleaned in place, can withstand harsh environments, and is potentially mobile.