Draft:Boundary flux

Boundary flux is a concept used in pressure-driven membrane filtration to distinguish operating regimes according to the development of membrane fouling. It was introduced in 2014 by Marco Stoller and Javier M. Ochando-Pulido as a common framework for the earlier concepts of critical flux and threshold flux.

The concept separates membrane operation into a lower-flux region, in which fouling is absent or develops at a relatively low and approximately constant rate, and a higher-flux region characterized by much faster fouling development. Independent studies have subsequently applied the boundary-flux concept to microfiltration, nanofiltration and reverse osmosis, while independent review literature has discussed boundary-flux modelling among approaches for describing membrane fouling.

Background

Membrane fouling is one of the main drawbacks of membrane technology and causes a progressive reduction in membrane permeability. In pressure-driven membrane processes, one approach to limiting fouling is to identify operating conditions in which the fouling rate remains sufficiently low.

The critical flux concept was introduced by Field and co-workers in 1995. It proposed the existence, for a given membrane-feed system and set of operating conditions, of a permeate flux below which time-dependent flux decline caused by fouling is not observed.

For many practical membrane systems, particularly complex feeds, measurable fouling can occur even at relatively low permeate flux. The later threshold flux concept was proposed to distinguish a lower-flux region with a low and approximately constant fouling rate from a higher-flux region in which the fouling rate increases markedly.

In 2014, Stoller and Ochando-Pulido introduced the term boundary flux, denoted as $J_{b}$, to describe critical- and threshold-flux behaviour within a single framework. In this formulation the critical-flux case can be regarded as the limiting case in which the permeability-loss rate below the boundary approaches zero. The original paper presented the boundary-flux concept as a simplification and unification of the two previous approaches rather than as an additional physical theory of fouling.

Lan and co-workers later described the development as an attempt to merge the critical- and threshold-flux approaches into the single concept of boundary flux.

Concept and parameters

The boundary-flux framework divides membrane operation into sub-boundary and super-boundary conditions.

Under sub-boundary conditions, the permeate flux $J_{p}$ is lower than or equal to $J_{b}$. In this region, permeability may remain approximately constant, corresponding to critical-flux behaviour, or decrease at a small approximately constant rate, corresponding to threshold-flux behaviour.

The permeability-loss rate in the sub-boundary region is represented by the parameter $\alpha$, termed the sub-boundary fouling rate index:

$${\frac {dm}{dt}}=-\alpha ,\qquad J_{p}(t)\leq J_{b}$$

where $m$ is membrane permeability and $t$ is time.

When $\alpha$ approaches zero, permeability becomes approximately constant with time and the formulation reduces to the critical-flux case.

For sub-boundary operation, integration of the permeability-loss rate gives the permeate-flux evolution:

$$J_{p}(t)=J_{p}(t_{1})-\alpha \int _{t_{1}}^{t}TMP(\tau )\,d\tau$$

where $TMP$ is the transmembrane pressure and $\tau$ is the integration variable.

For constant TMP this becomes:

$$J_{p}(t)=J_{p}(t_{1})-\alpha \,TMP\,(t-t_{1})$$

so that a slow decrease in permeate flux can occur even while the membrane remains under sub-boundary conditions when $\alpha >0$.

Above the boundary flux, an additional fouling contribution becomes significant. The parameter $\beta$, termed the super-boundary fouling rate index, is used to characterize the higher-fouling regime. Later modelling work treated $\beta$ as dependent on operating pressure rather than as a universal constant.

Dependence on feed composition and time

The boundary flux is not a fixed intrinsic property of a membrane. Its value depends on the membrane-feed system and on operating variables including hydrodynamics, temperature, feed composition and membrane properties.

For systems in which feed composition evolves during operation, the original boundary-flux formulation introduced a more general dependence on a representative feed parameter, denoted $KP$, and on operating time:

$$J_{b}=J_{b}(KP,t)$$

The model assumes that membrane permeability and osmotic-pressure effects can be approximated as functions of the selected feed parameter. Under these assumptions, a second-order expression for the boundary flux was derived:

$$J_{b}(KP,t)=m_{0}P_{b}-\alpha tP_{b}-\left(m_{0}p_{1}-\alpha p_{1}t+m_{1}P_{b}\right)KP+m_{1}p_{1}KP^{2}$$

where $m_{0}$ represents the initial membrane permeability, $P_{b}$ is the boundary operating pressure, and $m_{1}$ and $p_{1}$ describe the dependence of permeability and osmotic pressure on the selected feed parameter.

The expression makes the boundary flux a time-dependent operating limit rather than a constant value. For $\alpha >0$, membrane permeability progressively decreases even during sub-boundary operation, causing the corresponding $J_{b}(KP)$ curve to shift toward lower flux values with time.

The model also defines a concentration limit $KP^{*}$, corresponding to the lower positive root of the $J_{b}(KP,t)$ relation. As $KP$ approaches this limit, the boundary flux approaches zero, so that even very low positive permeate fluxes correspond to super-boundary operation and rapid fouling development. Values of $KP$ above $KP^{*}$ fall outside the physically meaningful range of the fitting relation.

For a fixed value of $KP$, the model predicts a finite operating time $t^{*}$ at which the corresponding boundary flux reaches zero. The original formulation associates this limiting behaviour with the progressive accumulation of irreversible fouling and/or membrane ageing.

Thus, remaining below the boundary flux does not necessarily imply constant membrane performance indefinitely: when the sub-boundary fouling rate is non-zero, the operating boundary itself progressively decreases with time.

Experimental determination

Boundary flux can be determined experimentally using procedures derived from critical-flux measurement methods. One approach is the transmembrane-pressure stepping method, in which the applied pressure is varied through successive steps and the resulting permeate-flux and permeability behaviour is analysed.

The method first requires characterization of the permeability-loss rate in the lower-fouling region. Pressure is then progressively increased until the measured flux behaviour departs from that expected from the sub-boundary permeability loss. This transition is used to identify the boundary pressure and the corresponding boundary flux.

An independent study by Zhu and co-workers used a TMP-step method to determine both the boundary flux and the sub-boundary fouling-rate index during constant-pressure microfiltration of activated-sludge suspensions.

Applications

Microfiltration

Zhu and co-workers applied the boundary-flux concept to activated-sludge suspensions from submerged membrane bioreactors. They experimentally determined $J_{b}$ and $\alpha$ and investigated the influence of suspended-solids concentration, stirring rate and membrane pore size on the measured boundary flux.

Nanofiltration

Li and co-workers applied boundary-flux theory to nanofiltration of municipal wastewater in 2016. The theory was used to predict the evolution of permeate flux during extended membrane operation and as part of the analysis of long-term membrane fouling behaviour.

Reverse osmosis

In 2024, Wen and co-workers applied boundary-flux theory to reverse-osmosis purification of sugar-mill condensate. Under the operating conditions studied, they identified a boundary flux of 5.92 L m⁻² h⁻¹ and compared membrane fouling under sub-boundary and super-boundary conditions.

The study reported different fouling behaviour under the two operating regimes and used boundary-flux operation as a criterion for examining fouling control in that reverse-osmosis system.

Process modelling and control

Boundary-flux modelling has also been discussed in independent review literature. A 2019 review by Ahmed, Hashaikeh, Diabat and Hilal in Desalination included the boundary-flux approach in its discussion of mathematical modelling of nanofiltration and membrane fouling.

The framework has also been used as the basis for membrane-process control strategies intended to maintain operation within lower-fouling conditions. In 2017, Stoller and Serrão Mendes developed an advanced control approach based on a boundary-flux fouling model. The work included implementation of a custom membrane-process unit in Aspen HYSYS and validation using experimental data from ultrafiltration and nanofiltration treatment of olive-mill wastewater.

See also

• Membrane fouling
• Cross-flow filtration
• Microfiltration
• Ultrafiltration
• Nanofiltration
• Reverse osmosis
• Membrane bioreactor

Further reading

• Stoller, Marco; Ochando-Pulido, Javier Miguel (2014). The Boundary Flux Handbook: A Comprehensive Database of Critical and Threshold Flux Values for Membrane Practitioners. Elsevier. ISBN 9780128015896.