Modeling the Effects of Biofouling On Full-Scale Reverse Osmosis Desalination: Spacer Clogging Or Membrane Fouling?

Biofouling has long been considered one of the most detrimental performance limitations for desalination plants employing reverse osmosis (RO) membranes. This is primarily due to polyamide composite membrane degradation by chlorine, and hence, the inability to continuously chlorinate RO membrane feed water. Once deposited onto a RO membrane surface, bacteria may lie dormant for long periods of time, but upon introduction of a nutrient source (e.g., biodegradable carbon from an algae bloom, pre-treatment chemicals), bacteria can rapidly form a biofilm that shuts down a desalination plant. Over the past decade there has been some debate over the mechanisms of biofouling in full-scale RO spiral wound elements, leading a number of membrane scientists and desalination experts to ask the question, “Is it spacer clogging or membrane fouling?”

Towards answering this question, we extended a previously published abiotic, organic/colloidal fouling model already proven applicable to full-scale RO plants [1], to include biogrowth leading to biofilm formation and resultant RO system performance decline.  The model leverages classical substrate-limited bacterial growth and biofilm structure models, and is sufficiently parameterized to allow perturbation studies to be performed on the biofilm (substrate concentration, substrate utilization rate, biomass conversion rate, biofilm structure), the membrane module (water/salt/nutrient permeability, spacer thickness & porosity, spiral leaf length, width, and number), plant design (number of elements per vessel, number of stages, array configuration), and operating conditions (applied pressure, recovery, temperature, feed salinity).

The biofilm growth process is coupled to physically important transport processes such as (1) available open area through the spacer-filled cross flow channel, which affects axial pressure drop and mass transfer, (2) hydraulic resistance to water permeation, which affects trans-membrane pressure drop, and (3) hindered salt diffusion within the biofilm, leading to biofilm-enhanced osmotic pressure. The model considers local information (e.g. biofilm properties, mass transfer coefficient and hydraulic friction factor) and integrates it over the entire plant – enabling one to rapidly probe the externally visible signs of performance decline at the plant-level. In particular, the model provides insight into conditions under which a RO plant suffers from spacer clogging versus membrane fouling.

1. E.M.V. Hoek, J. Allred, B.H. Jeong, and T. Knoell,  “Modeling the Effects of Fouling on Full Scale Reverse Osmosis Processes,” Journal of Membrane Science 314 (2008) 33-49.