In this study, a fluorescent silica nano particle is used as the surrogate for challenging test of membrane surface integrity. The particles are functionalized by a fluorescent dying agent so that as an ultraviolet light is imposed a bright fluorescent image from the particles can be taken. If a membrane surface is damaged and has a compromised part larger than the size of surrogate the fluorescent particles would pass through and contained in the permeate. An operator can directly notice whether the membrane surface is damaged or not by detecting a fluorescent image taken from the permeate. Additionally, the size of compromised part is estimated through analysing the fluorescent image in which we surmise the mass of particles included in the permeate by calculating an average RGB value of the image. The pilot scale experiments showed that this method could be applied successfully to determine if a membrane surface had a damaged parts regardless of the test condition. In the testing on the actual damaged area of 4.712 mm2, the lowest error of estimating the damaged area was –1.32% with the surrogate concentration of 80 mg/L, flux of 40 L/m2/hr for 25 minutes of detection. A further study is still going on to increase the lowest detection limit and thus decrease the error of estimation.

In this study, a novel method was proposed to test the integrity of water treatment system specifically equipped with membrane filtration process. We applied the silica particles coated with a fluorescent agent (rhodamine B isothiocyanate) as a surrogate to detect a membrane process integrity and evaluate the reliability of effluent quality in the system. Additionally, a series of experiments was conducted to evaluate the sensitivity of the method through the laboratory scale experiment. The laboratory scale experiments showed that the feasibility of application of proposed method to detect a breach or damaged part on the membrane surface. However, the sensitivity on predicting the area of a breach was significantly influenced by the testing conditions such as a concentration of surrogate, filtration flux, and detection time. The lowest error of predicting the area of breach was 3.5 % at the testing condition of surrogate concentration of 80 mg/L injected with flux of 20 L/m2/hr for 10 minutes of detection time for the breach having the actual area of 7.069 mm2. However, the error of estimation was increased at the small breach with area less than 0.785 mm2. A future study will be conducted to estimate a damaged area with more accuracy and precision.