Development of a dynamic model for cleaning ultra filtration membranes fouled by surface water
Introduction
Ultra filtration is increasingly used as a technology to purify surface water. UF membranes have high selectivity and are economically attractive. A major drawback, however, is the limitation in performance as a result of fouling. Consequently, frequent cleaning of the membrane is necessary. In the short term, fouling is controlled by means of hydraulic cleanings (backwashes), but in the long term, fouling that cannot be removed by means of backwashes builds up, and treatment of the membrane with chemicals is required. It is expected that cleaning costs can be reduced by means of optimization.
The main objective of this paper is the development of a cleaning model that is suitable for (dynamic) optimization of the chemical cleaning procedure. A three parameter model that describes dynamic behavior of a chemical cleaning procedure is proposed. After construction of the model, a protocol for experimental verification of the model is developed and tested. The model parameters are determined by a nonlinear least squares algorithm using experimental runs. The fitted parameters are consequently used to validate the model for additional experimental runs. It was found that the model describes the cleaning process well.
Section snippets
Requirements of the model
Chemical cleaning of membranes is widely studied in a qualitative way [1], [2], [3], [4], [5], [6]. In these studies, the characterization of fouling and the interactions that occur between fouling, cleaning agent and membrane surface are described in detail.
The number of authors who published on the modeling aspects of chemical cleaning of membranes is limited. Bird [7] developed a microscopic cleaning model for the caustic cleaning of ceramic micro-filtration filters fouled by whey
Equipment and material
Fouling and subsequent cleaning experiments were performed with a laboratory scale dead end ultra filtration unit. The membrane used in the experiments is a poly ether sulphone Norit-Xiga RX300 PSU hollow fiber ultra filtration module with a membrane surface of m2.
The setup consists of the ultra filtration membrane module, a filtration and a backwash pump, flow indicators to monitor fluxes, temperature indicators to correct trans membrane pressure for changes in temperature and
Results
Six experimental cleaning runs were performed at different sodium hydroxide concentrations. Table 1 shows the settings and the effectiveness of the cleaning experiments. Runs 1–4 are used to determine the model parameters, using a nonlinear least squares algorithm. The model parameters were determined to be , , . The value is small, which implies that the cleaning agent is not consumed during the cleaning reaction, which is in agreement with the
Conclusions
A simple cleaning model, based on component balances was proposed. The model predicts the irreversible fouling decay as function of cleaning time and cleaning agent concentration. The model parameters can be determined by a simple experimental protocol, measuring trans membrane pressure, pH and turbidity. The execution of the protocol can be done relatively fast, which facilitates frequent online adaptation of the model parameters, when process conditions and water quality vary with time.
Acknowledgements
The financial support of NWO/STW, Aquacare Europe, Hatenboer-Water, Norit Membrane Technology and Vitens Laboratory & Process Technology is gratefully acknowledged.
References (9)
- et al.
Chemical cleaning of reverse osmosis membranes fouled by whey
Desalination
(2004) - et al.
The chemical cleaning of polymeric UF membranes fouled with spent sulphite liquor over multiple operational cycles
J. Membr. Sci.
(2003) Study on the cleaning of new ultra filtration spiral-woundmodules to prevent membrane fouling (including biological fouling)
Desalination
(2005)- et al.
Optimization of membrane physical and chemical cleaning by a statistically designed approach
Journal of Membrane Science
(2003)