Bioresource Technology 329 (2021) 124828

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Review

Integrated membrane bioreactors modelling: A review on new

comprehensive modelling framework
Giorgio Mannina a, *, Marion Alliet b, Christoph Brepols c, Joaquim Comas d, e, Jerome Harmand f,
Marc Heran g, Nesrine Kalboussi h, Jacek Makinia i, Ángel Robles j, Taise Ferreira Rebouças a,
Bing-Jie Ni k, Ignasi Rodriguez-Roda d, e, María Victoria Ruano j, Giorgio Bertanza l, Ilse Smets m
a
Engineering Department, Palermo University, Viale delle Scienze, Ed.8, 90128 Palermo, Italy
b
Laboratoire de Génie Chimique, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
c
Erftverband, Am Erftverband 6, 50126 Bergheim, Germany
d
Catalan Institute for Water Research (ICRA), Emili Grahit 101, 17003 Girona, Spain
e
LEQUiA, Laboratory of Chemical and Environmental Engineering, University of Girona, Campus Montilivi, 17071 Girona, Spain
f
LBE, Univ. Montpellier, INRAE, Narbonne, France
g
IEM, Univ. Montpellier, CNRS, ENSCM, Montpellier, France
h
Université de Carthage, Institut National ds Sciences Appliquées et de Technologie & Université de Tunis El Manar, Ecole Nationale d’Ingénieurs de Tunis, Laboratoire de
Modélisation Mathématique et Numérique dans les sciences d’ingénieur, Tunis, Tunisia
i
Faculty of Civil and Environmental Engineering, Gdansk University of Technology, ul. Narutowicza 11/12, 80-233 Gdansk, Poland
j
Departament d’Enginyeria Química, Escola Tècnica Superior d’Enginyeria (ETSE-UV), Universitat de València, Avinguda de la Universitat s/n, 46100 Burjassot,
València, Spain
k
Centre for Technology in Water and Wastewater, School of Civil and Environmental Engineering, University of Technology Sydney, Sydney, New South Wales 2007,
Australia
l
Departament of Civil, Environmental, Architectural Engineering and Mathematics, Brescia University, via Branze 43, 25123 Brescia, Italy
m
Department of Chemical Engineering, KU Leuven, Celestijnenlaan 200F Box 2424, 3001 Heverlee, Belgium

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• A literature review on integrated MBR

modelling and control is presented.
• The use of integrated MBR models
should be more encouraged.
• A new framework was proposed to pur-
sue good practice for MBR modelling.
• Integrated MBR modelling applications
to real case studies is needed.

A R T I C L E I N F O A B S T R A C T

Integrated Membrane Bioreactor (MBR) models, combination of biological and physical models, have been
representing powerful tools for the accomplishment of high environmental sustainability. This paper, produced
Keywords:
High environmental sustainability
by the International Water Association (IWA) Task Group on Membrane Modelling and Control, reviews the
Biological processes state-of-the-art, identifying gaps for future researches, and proposes a new integrated MBR modelling framework.

* Corresponding author.
E-mail address: [email protected] (G. Mannina).

https://doi.org/10.1016/j.biortech.2021.124828
Received 2 December 2020; Received in revised form 30 January 2021; Accepted 3 February 2021
Available online 17 February 2021
0960-8524/© 2021 Elsevier Ltd. All rights reserved.
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Performance indicators In particular, the framework aims to guide researchers and managers in pursuing good performances of MBRs in
Modelling framework terms of effluent quality, operating costs (such as membrane fouling, energy consumption due to aeration) and
mitigation of greenhouse gas emissions.

1. Introduction involving integrated MBR modelling identifying gaps for future re-
searches. The second goal focusses on establishing the basis to propose a
The application of mathematical models to wastewater treatment framework focused on best modelling practice of aerobic MBRs, also
plants (WWTPs) or water resource recovery facilities (WRRFs) is widely highlighting future needs to achieve this goal. A framework for good
used in view of providing more accurate answers for design, manage- MBR modelling practice can be employed for the design, operation,
ment, optimization and control issues (Zhang et al., 2021; Zuthi et al., optimisation and management of MBR systems involving multiple-
2012). Indeed, since the 80 s they have been representing a powerful targets: effluent quality, membrane fouling, energy consumption and
tool to understand the several features related to the treatment process overall operating costs, and mitigation of GHG emissions, among others.
(Martin & Vanrolleghem, 2014; Hamedi et al., 2019). The Activated
Sludge Model (ASM) series represent the highlight concerning WWTP 2. Motivation of integrated MBR modelling
modelling since they are considered a reliable tool to assess the main
features related to wastewater treatment. They were first applied to In this section, we present a number of arguments to illustrate the
traditional activated sludge (AS) systems, consisting of a bioreactor and necessity of considering coupled biological and physical principles to
a secondary clarifier, and evolved to accompany more advanced tech- come up with what will be named hereafter “integrated models”.
nologies, such as aerobic membrane bioreactors (MBRs). MBRs came to Continuous MBRs do not operate “continuously” but often “intermit-
light as an advanced treatment technology with the ability to provide an tently” mainly due to the necessary backwash or relaxation phases
effluent with higher quality to comply with strict regulation limits, subject but also due to fluctuating influent flows. Such intermittently
reduce sludge production, require less space, and have potential for operated processes, also known as production/regeneration systems,
upgrading existing WWTPs and implementing water reuse (Judd, 2006; present the fundamental characteristic of never working at steady-state.
Krzeminski et al., 2017; Zuthi et al., 2017; Zheng et al., 2018). Recent Thus, during the treatment phase, the biomass leads to a set of
studies have reported that the implementation of aerobic MBRs is complex phenomena which may cause a progressive fouling of the
increasing over the years due to their many advantages over conven- membrane; during backwash/relaxation phase, the matter accumulated
tional treatment technologies (Deng et al., 2016; Meng et al., 2017). onto the membranes is washed off out in order, ideally, to recover its
Attempts to combine membranes with the activated sludge process were original filtering properties. In practice, the presence of “irreversible
already made 50 years ago (Smith et al., 1969). The first full-scale and/or irrecoverable fouling” phenomena leads to the fact that after the
commercial applications of aerobic MBRs were reported in the late backwash we do not find exactly the initial filtration properties but
1970s in North America (Smith et al., 1969), and early 1980s (Japan). conditions in which these properties slowly degrade backwash after
MBRs use micro- or ultra-filtration membranes for the phase (solid- backwash. In the long term, such a phenomenon must be solved by
–liquid) separation. The nominal pore size of membrane filters is typi- operators. However, to remain as pedagogic as possible, this is not
cally 0.02–0.2 μm. While smaller dissolved fractions (e.g. humic and explicitly considered in the very simple model (as the one represented
low-molecular-weight substances) pass through the membrane, larger above by Eqs. (1)–(6)). During these phases, complex feedbacks - of both
particles are mostly retained. Typical examples of such large particles biological and physical natures - are exerted on the variables of the
include flocs, bacteria and organic colloids (Fenu et al., 2010). system and influence their dynamics. Biological dynamics are highly
Following the spread of aerobic MBR technology, mathematical dependent on different factors, such as the nature of the influent (i.e.
models advanced to include the unit process of membrane filtration. flow and content in particulate matter), the structure of the biomass, the
From a modelling point of view, MBR differs from conventional acti- aeration power, etc. These characteristics highly influence the fouling
vated sludge systems (CAS) not only due to the physical separation but propensity of the bioreactor content. Conversely, the application of a
also due to the need for integrating the biological and physical treatment transmembrane pressure (TMP) leads to the attachment of the particu-
features, which is one of the biggest challenges faced by researchers late matter and of certain molecules on the membrane and in its pores,
(Zuthi et al., 2012; Naessens et al., 2012). To address this issue, inte- gradually blocking the outlet flow. Due to the corresponding elevated
grated MBR models were introduced in the literature (Zarragoitia- TMPs, more severe cross flow velocities of the sludge mixture or higher
González et al., 2008; Di Bella et al., 2008). air scouring flows need to be applied which affect the bioflocculation
Despite the existence of such models, there is a pressing demand for structure of the biomass and, hence, its activity and again its filterability.
the updating of integrated models to consider the numerous new targets Indirectly, biological parameters (sludge concentration, kinetics, etc…)
that were brought into the current MBR’s situation, such as fulfilment of may also be influenced. Thus, biology – of the biological process – in-
more and more restrictive legal effluent requirements, or improvements fluences the physical phenomenon – the filtering of the medium through
for several key performance indicators (KPI): energy consumption and the membrane – and vice versa. To illustrate the direct influence of the
overall operating costs, greenhouse gas (GHG) emissions, land use, so- filtration process on the dynamics of biological variables, consider the
cial acceptance, or resource consumption (Puyol et al., 2016; Bozkurt simplest school case in which a biomass X would grow on a single
et al., 2016; Atanasova et al., 2017; Mannina et al., 2017c). However, substrate S in a bioreactor with a perfect membrane separating soluble
many questions still remain about the connection between biological (S) and particulate matter (X). Assume the system is operating at con-
and physical processes, especially due to the complexity of the mem- stant TMP and that the flux decreases over time due to the attachment of
brane fouling phenomenon (Hamedi et al., 2019). matter onto the membrane (named M). The simplest model of such a
Some reviews have been published in the past to provide some situation would be written as follows during the filtration phase:
enlightenment regarding integrated MBR modelling (Ng & Kim, 2007;
Zuthi et al., 2012, 2013), which have not been updated ever since. Ẋ = μ(S)−αD(M))X (1)
Additionally, as far as the authors are aware, none of these works have
1
presented any kind of review focused on MBR multi-objective assess- Ṡ = (Sin − Sout )D(M) − μ(S)X (2)
Y
ments through integrated modelling. Therefore, the first goal of this
paper is to provide an up-to-date review of the current scenario

2
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Ṁ = βαD(M)X (3) account to finally come up with what will be named hereafter an “in-
tegrated model”. Looking at Fig. 1, it can be noticed that direct feed-
where, μ is the kinetics of the biomass, Sin is the input substrate con- backs of the physical compartment on the biology of the AS could also be
centration, Y is a yield coefficient, D(M) is the flow rate going through considered in the sense, for instance, where the quantity of attached
the reactor (assumed to be a decreasing function of M bounded by biomass M could directly impact the biomass growth kinetics in the
0 when M equals a limit value M*), α and β are kinetics and scaling medium. The modelling of such feedback obviously opens up instigating
parameters (notice that X will rather be in concentration and M in mass). research perspectives that are, however, out of the scope of the present
During the backwash/relaxation phase, this simplest model would be paper.
written as follows:
3. Integrated MBR models
Ẋ = ρM (4)
3.1. Classification
Ṡ = 0 (5)
Since the MBR invention, several efforts have been employed to
Ṁ = − γM (6)
improve their operation strategies and to prevent their known limita-
where, ρ and γ are kinetics and scaling parameters. tions (mainly membrane fouling issues) from affecting the viability of
There are many hypotheses behind this simple model (no oxygen the technology expansion. One of the challenges was to understand how
limitation, no biomass decay, etc…), but assuming an alternating the biological treatment communicates with the physical one, which led
functioning of the system, we immediately see that: to the necessity of combining biological models with membrane filtra-
tion models (Wintgens et al., 2003; Di Bella et al., 2008). From this point
1. Except Xeq = 0, the model does not exhibit any constant steady state: and based on the available literature, four types of aerobic MBR models
whatever the initial conditions, the system evolves such that M tends are generally described in the literature: biological, hybrid, physical,
towards M*, then D = 0 and the system is switched to the regener- and integrated.
ation mode until the membrane recovers (M be small enough), and Biological models refer to the unmodified ASM models, formerly
then the system is switched back to the production mode, etc… developed to describe both the kinetics and the stoichiometry for bio-
2. Because of the feedback of the membrane (the physical device) on logical nutrient removal activated sludge systems (Henze et al., 1987;
the flux, the membrane exerts a direct feedback on the dynamics of 1995), which are also used for modelling aerobic MBRs (Fenu et al.,
the biological variables. 2010). The called hybrid models in fact are just modified versions of the
3. Notice finally that even if the biomass is considered to be at what is ASM family, which were adapted including new state variables to take
called a “pseudo steady state”, the fact that D depends explicitly on M into account mainly the soluble microbial products (SMP) and extra-
– which can precisely be seen as a feedback of physical phenomena cellular polymeric substances (EPS) formation and degradation pro-
on biology – introduces a “continuous dynamical behaviour”. cesses (Zuthi et al., 2012), thus we will consider them as biological
models in this paper. Other authors (Galinha et al., 2018) define “MBR
As a consequence, such a system cannot be modelled by a “biological hybrid modelling” as the combination of a mechanistic model (ASM
compartment” followed by a model describing the physical behaviour of based) with non-mechanistic models (for membrane filtration mainly),
the membrane: the coupling of both must be necessarily taken into but we have focussed our review on mechanistic models, so we will not
take this definition into consideration. On the other hand, the physical

Fig. 1. Example of a schematic representation of integrated MBR models and correlations (Where Su:= soluble undegradable organics; SB:= soluble biodegradable
organics; SNHx:= soluble biodegradable ammonia; XCB:= particulate biodegradable organics; XCB:= particulate biodegradable organics; Xue:= Particulate unde-
gradable endogenous products ; YOHO:= yield coefficient of heterotrophic biomass; YANO:= yield coefficient of autotrophic biomass; Xcb:= particulate biodegradable
carbon, XUE:= particulate substances; XCBN:= particulate nitrogen.

3
G. Mannina et al. Bioresource Technology 329 (2021) 124828

models consider the liquid–solid separation process promoted by the leads to a comprehensive prediction of the system’s behaviour. Among
membrane filtration. In particular, the following processes are generally its main advantages, the integrated MBR models can provide credible
taken into account (Mannina et al., 2011a): i) cake layer formation estimations for full-scale facilities and validate results obtained on a
during the membrane filtration, permeate backwashing phases, and laboratory scale (Zuthi et al., 2012). They can also surpass the limita-
aeration; ii) pre-filtration throughout the cake layer (biological mem- tions of experimental results, which offer a limited universe of possi-
brane effect) leading to removal of organic matter (COD); iii) particle bilities while the model can present a wide range of scenarios to be
retention by the membrane (physical membrane effect) leading to considered during a decision-making process (Mannina & Cosenza,
removal of organic matter (COD); and iv) irreversible membrane fouling 2013; Monclús et al., 2012). In other words, by the use of integrated
(specifically pore narrowing; pore blocking; and influence of SMP on MBR models, managers are able to explore a variety of operating con-
pore fouling). Finally, the integrated MBR models are the combination ditions prior to their application on-site, avoiding waste of environ-
of the biological models (whatever the nomenclature used by the au- mental, physical and chemical resources which is reflected over plant’s
thors) with the physical model. performance indicators (e.g., operating costs, energy consumption,
Despite the simplified definition provided above, the concept of in- effluent quality, etc.).
tegrated MBR model is controversial, since the literature still struggles On the other hand, one of their major problems is related to the se-
with an agreement regarding their composition. Indeed, a historical lection of default values for crucial information (e.g., biomass growth
overview points out that integrated MBR models were first considered to and decay rates, formation/degradation coefficients, compounds indi-
be a correlation between biological and physical models (Ng & Kim, vidual fractioning, etc.), because available literature references are
2007), without a clear definition regarding the concept of the first scarce/limited and usually related to plants with different characteris-
model. Indeed, according to Ng & Kim (2007), an integrated MBR model tics among them (sometimes related only to CAS processes) (Zuthi et al.,
can be defined as the connection between a biological or hybrid models 2012). Another important issue regards to the model’s calibration and
and a physical model. According to Di Bella et al. (2008) and Mannina uncertainty, which are complex procedures that require time and
et al. (2011a), the integrated MBR models are a combination of bio- trustable data (Fenu et al., 2010; Zuthi et al., 2012; Mannina et al.,
logical models to describe biomass behaviour and physical models to 2010). In addition, models that were calibrated considering lab-scale
describe membrane fouling. information may provide underestimated results when applied to full-
In the second part of a review paper, Naessens et al. (2012) presented scale facilities. These liabilities reflect in results that may fit with the
three types of integrated models for MBR systems: i. models that couple researcher’s data but fail when applied to other researches (Ni and Yuan,
biological and filtration models (with and without the estimation of SMP 2015). Sensitivity and uncertainty analysis as well as comprehensive
and EPS); ii. models that integrate population mass balances into the calibration protocols are limited for MBR modelling and their applica-
integrated framework; iii. models that integrate cost models into the tion is needed in view of getting good modelling results (Freni and
integrated framework. No formal concept of the integrated approach Mannina, 2010; Mannina et al., 2010; Mannina and Viviani, 2009).
was presented by Naessens et al. (2012) in this review. The major Indeed, recently Mannina et al. (2017a); (2018c;)) carried out a study on
question retrieved from this work is which boundaries should be given the assessment of the sensitivity and uncertainty analysis for both an
while defining the integrated MBR models since the categorization ASM1 and ASM2d integrated (biological and physical) MBR model.
seems to encompass several aspects related to the whole MBR technol- Authors were able to pin down the source of uncertainty and most
ogy (e.g., calibration of half-saturation coefficients and costs), instead of influential parameters in the MBR modelling. The study showed that for
focusing on the biological and physical aspects that are considered by the gaseous model outputs 88–93% of the measured data lays inside the
the MBR modelling. Indeed, a few years later, Mannina et al. (2018a) confidence bands showing an accurate model prediction. Future studies
stated that integrated MBR models combine physical and physical should focus on the estimation of uncertainty in order to provide a
models in order to predict MBR behaviour. This short definition is in quality of model prediction as well as gaining insights in the dominated
agreement with previous literature (Di Bella et al., 2008; Mannina et al., processes.
2011a; Zuthi et al., 2012). Considering the overall discussion above reported, there is a need to
From the aforementioned definitions, one can understand that the simplify the definition of what is called integrated modelling that has a
use of integrated MBR models is the most comprehensive way to model non-clear and unique definition causing misunderstanding among
MBR systems. Some authors employ different approaches not simulating modellers. In view of that, here we define integrated MBR modelling as
EPS/SMP and obtaining acceptable results (Wintgens et al., 2003; the combination of biological and physical models to describe mem-
among others). For some other authors, the hybrid models are important brane filtration (Fig. 2), assuming that biological models refer to ASM
because SMPs and EPS are considered one of the main causes of mem- family (adding or not state variables and other processes). Special
brane fouling (Drews et al., 2008; Meng et al., 2009; among others). emphasis should be given to the link between physical and biological
Their formation/degradation processes are known to happen during the models which define the interactions in the integrated model which are,
biological treatment by means of substrate utilization, biomass decay in turns, the main features. Further, there is a continuous exchange of
and cell hydrolysis (Zuthi et al., 2012). The introduction of these aspects information from the biological to physical model and in some cases vice
in a hybrid model may provide for some cases a reasonable approach to versa (Mannina et al., 2010). The main advantages of integrated MBR
evaluate membrane fouling (Ahn et al., 2006). Some reviews were models consist in getting a more comprehensive approach and in having
published on this matter with the aim to address the aspects of MBR the possibility to better understand the overall involved phenomena
biomass kinetic modelling (Zuthi et al., 2012; Scholes et al., 2016), (both biological and physical) in view of an optimization. On the other
which can be consulted to a more detailed approach. hand, main disadvantages of integrated MBR models are the complexity
As for the physical model, literature states that most of them were and larger data set needed for model application respect to simple
developed to address membrane fouling issues considering resistance- modelling approaches.
based equations (Wintgens et al., 2003). In fact, models related to
physical aspects mostly differ in terms of complexity. While some 3.2. Historical evolution
consider aspects such as resistance-in-series (RIS) and permeate flux
(Lee et al., 2002; Wintgens et al., 2003; Sarioglu et al., 2012; Robles Fig. 3 summarises the historical evolution that led to the current
et al., 2013), others assess carbon removal by the cake layer (deep-bed integrated MBR models.
filtration theory), and the effects of reversible and irreversible fouling As mentioned before, integrated models derive from the ASM-types.
over permeate flux (Zuthi et al., 2013). The ASMs have evolved from the assessment of biological carbon
The idea of coupling biological and physical models in a single model removal, nitrification, and denitrification in the ASM1 (Henze et al.,

4
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Fig. 2. Schematic representation of integrated MBR models.

1987), to assess SMP and EPS (Orhon et al., 1989), then to include presented by Zarragoitia-González et al. (2008) and Di Bella et al.
biological phosphorus removal in the ASM2 (Henze et al., 1995), in- (2008). Zarragoitia-González et al. (2008) described a detailed hybrid
ternal cell storage compounds within heterotrophs in the ASM3 (Gujer ASM1-SMP model and simulated the biological-filtration link. The
et al., 1999), and GHG production (ASMN) (Hiatt and Grady, 2008). hybrid ASM-SMP model was based on the work of Lu et al. (2001) and
When MBR started to spread as an advanced treatment technology with Cho et al. (2003); the physical processes were modelled by the filtration
many advantages with respect to CAS (Suganthi et al., 2013; Xiao et al., model of Li and Wang (2006) where coarse bubble aeration was
2014), the development of specific MBR modelling tools was stimulated. considered both for the effects of its cycles on the attachment and
The first concerns about modelling MBR systems arose after the detachment of the cake layer formation and for the repartition of the
breakthrough for the MBRs proposed by Yamamoto et al. (1989). fouling along the height of the membrane. The latter model (Di Bella
Indeed, Chaize and Huyard (1991) applied an unmodified ASM1 to an et al., 2008) was similar to the one presented by Zarragoitia-González
MBR plant to assess its performance and found that the biological model et al. (2008), except for the physical model, which included the influ-
alone provided results in disagreement with the experimental ones, ence of backwashing in the attachment and detachment forces to the
since the ASMs were developed to consider the CAS systems as a refer- cake layer formation (instead of aeration) and, for the first time in an
ence, which rendered difficult applying an unmodified version to an integrated model, including hybrid EPS/SMP modelling, and the
MBR system. Lately, other attempts were made in view of modelling removal of COD based on deep-bed filtration theory (Bai and Tien, 2000;
membrane processes focused on the physical aspects. Nagaoka et al. Kuberkar and Davis, 2000). Lately, Mannina et al. (2011a) further
(1998) developed a mathematical model to simulate temporal changes modified the model proposed by Di Bella et al. (2008) including the
in suction pressure, flux, and filtration resistance. Additionally, the sectional approach for the resistance in series (Li and Wang, 2006) and
model considered the accumulation, detachment, and consolidation of the SMP model by replacing the Lu et al. (2002) model with Jiang et al.
EPS on the membrane surface. Despite considering EPS results, this (2008). Further, the model calibration was enhanced by considering the
model may not be considered hybrid since no biological assessment was protocol proposed by Mannina et al. (2011b) where the global sensi-
performed during model simulations. tivity analysis (Saltelli et al., 2004) was included to take into account the
Along with the development of the physical models, even more interactions among the model parameters. In Fig. 4 the main feature of
specific features were included in the biological ones, such as the for- Mannina et al. (2011a) model are reported.
mation of SMP and EPS (Lu et al., 2001, 2002). Indeed, the fact that Additional features were added to the ASM family. For example,
SMP/EPS kinetics were already under study by modellers (Namkung and excess sludge production, oxygen transfer rate, oxygen consumption
Rittmann, 1986; Orhon et al., 1989; Ahn et al., 2006; Aquino and rate started to be considered (Fenu et al., 2010). Meanwhile, the
Stuckey, 2008; Jiang et al., 2008), facilitated the integration of these assessment of SMP and EPS were being updated. Janus & Ulanicki
features into the biological models into a hybrid approach. The modi- (2010) divided the SMPs between soluble utilization associated products
fication of the ASM1 held by Lu et al. (2001) may be considered the first (SUAP) and soluble biomass associated products (SBAP) in accordance
application of a hybrid model. Lately, Lu et al. (2002) modified an ASM3 with the approach proposed by Laspidou and Rittmann (2002). Their
to include the estimation of SMP and EPS. From this point, the first at- purpose was to provide reliable values of SMP and EPS to be applied
tempts to develop an integrated MBR model were observed (Wintgens whilst modelling MBR processes. Results reported a strong correlation
et al., 2003; Saroj et al., 2008), although the link between the biological between SMP/EPS and mixed liquor volatile suspended solids (MLVSS)
and the filtration models was not clear yet. For example, Saroj et al. and sludge retention time (SRT), indicating a pathway to be followed in
(2008) applied an ASM3 with EPS dynamics, coupled with an EPS based the integration of biological and physical features.
filtration model (Ognier et al., 2004), without, however, linking the EPS Physical models were also updated, presenting important results that
dynamics with the filtration model. An extensive assessment among led to the current versions of the integrated models. Wu et al. (2012)
published literature demonstrates that the first complete versions of developed a combined cake layer and pore fouling model in view of
integrated MBR models including SMP/EPS were contemporaneously assessing the influence of solid, colloidal and soluble components over

5
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Fig. 3. Historical overview of the key developments for integrated MBR modelling evolution.

membrane fouling. The results were reported to be positive since the membrane performance. From this point, integrated MBR models star-
model successfully predicted the changes in TMP, and also in cake and ted to consider other targets, such as dynamic fouling.
pore resistance at various aeration intensities. Indeed, results reported Considering the dynamic fouling, for example, Charfi et al. (2015)
that aeration exerted a significant influence on fouling evolution, which and Zuthi et al. (2017) developed a semi-empirical mathematical model
stimulated other modelling approaches in the future. that accounts for cake formation and pore-blocking as the major
Since the integrated approach became more spread, several reviews contributor to membrane fouling. This integrated MBR model considers
were published to address some issues that still remained unclear the resistance due to pore blocking (RPB) and the loss of porosity in
(Naessens et al., 2012; Zuthi et al., 2012). One of the main findings of accordance with the approach of Bowen et al. (1997) and Busch et al.
Zuthi et al. (2012) was related to the production and degradation of SMP (2007). The model also considered that a pore size reduction is expected
within an MBR, which can be influenced by the operating specificities of due to the adsorption of soluble particles within the pores. The resis-
the treatment process, such as longer SRT, feeding ratios, total sus- tance due to cake layer formation (RCL) is obtained by considering the
pended solids, and MLSS among others. These results not only led to the attachment and detachment forces, which lead to a final thickness of the
spread of the integrated MBR models but also to understand that pro- cake after the application of filtration and backwashing fluxes. In the
moting changes in the operating parameters (i.e., to optimize the range end, the total resistance (RT) is modelled by means of the resistance-in-
of parameters set for MBR functioning) could lead to the optimization of series that accounts that RT is equal to the sum of the membrane intrinsic

6
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Fig. 4. Integrated MBR model as proposed by Mannina et al. (2011a).

resistance (RIt,M) plus RPB and RCL SMP was found to be the most while Model II considered nitrification as a two-step process (Pocquet
important cause of the biofilm formation onto the membrane surface. et al., 2016). The model was applied to understand the influence of both
A similar physical model was applied by Mannina et al. (2011a). nitrification approaches on GHG emissions. Results showed that Model
However, RCL was divided into two complementary resistances: the II had a better capacity to match experimental data as it considers a more
resistance of the stable (and irreversible) cake layer (RIv,CL) and of the comprehensive approach when it comes to nitrous oxide (N2O) forma-
dynamic (and reversible) cake layer (RRv,CL). These resistances are also tion by the ammonia-oxidizing bacteria (AOB). This application is
related to the presence of filtration and backwashing fluxes, but RRv,CL important to ensure the consideration of biomass metabolism while
relates to the amount of sludge that detaches from the membrane surface modelling GHG emissions and elevates the concerns regarding the
with the influence of aeration, whilst RIv,CL represents the irreversible application of more accurate kinetic values. Despite the advanced
cake that can be removed only by means of chemical cleaning. The studies presented by Mannina’s and co-workers, the models need to be
application of Zuthi et al. (2017) does not include the cross-sectional applied to real WWTPs to further verify their suitability.
approach that divides the membrane into equal fractions in order to The historical evolution of the integrated approach led to a more
consider the probability of cake formation according to how distant the consistent knowledge regarding the MBR’s functioning, and the appli-
section is from the aerator (Zarragoitia-González et al., 2008; Mannina cation of the integrated approach to assess MBRs operating conditions
et al., 2011a). This characteristic allows the model to correlate the has been proved as a trustworthy method that provides reliable and
aeration with membrane fouling, which is a subject that demands more realistic results (Saroj et al., 2008). Thus, the current step is to provide
attention from the literature Notations of the above mentioned filtration the models with tools that allow overcoming MBR’s most important
resistances have been amended according to literature (Brepols et al., obstacles in order to optimize their performance.
2020). Another application of the MBR modelling can be system optimiza-
Indeed, new applications can be found in literature considering the tion. Specifically, integrated models employed for MBR optimization are
comprehensive physical approach previously presented, with the com- still under careful studies and there are few available in the literature (Di
plete features of a hybrid model. From this point, the integrated models Bella et al., 2008; Zarragoïtia et al., 2009; Jang et al., 2006; Zuthi et al.,
were also able to estimate GHG emissions, since WWTPs were revealed 2013; González Hernández et al., 2015; Yang et al., 2017). Among the
to be responsible for almost 3% of the GHG emissions on a global scale available tools, an even smaller amount has assessed the optimization of
(IPCC, 2013; Koutsou et al., 2018). Mannina et al. (2017a) and Mannina an MBR by considering multi-objective targets, such as energy demand,
et al. (2018a) presented the first integrated models employing ASM1 and operating costs, etc. (Maere et al., 2011; Dalmau et al., 2013; 2014;;
ASM2d, respectively, and taken into account GHGs. Both models Gabarrón et al., 2015). For this reason, multi-objective performance
modified the nitrogen modelling by considering two steps and four steps assessment has become imperative to the development of the MBR
for nitrification and denitrification processes, respectively. Further, technology and the use of integrated MBR models may be an applicable
Mannina et al. (2018b) presented the comparison among two different solution. Further studies are thus needed in order to provide a more
integrated MBR models applicable to the assessment of nutrient comprehensive approach in MBR optimization and control. Future
removal, considering two different aspects of nitrification: the Model I studies should focus on the studying the interplay role among biological
applied the nitrification as a one-step process (Hiatt & Grady, 2008), and physical processes for enhancing design and operation of MBR

7
G. Mannina et al. Bioresource Technology 329 (2021) 124828

systems. the modelling targets can be considered by the integrated MBR models
in order to reflect on the results of each KPI presented in this paper.
4. Towards a framework for good MBR integrated modelling Some specific examples of MBR modelling projects include the
practice design of the required membrane surface or the specific aeration de-
mand device, the optimisation of biological nutrient removal and
An integrated modelling and simulation study involving MBR re- filtration processes in an integrated way, the control of membrane
quires a stepwise procedure to take into account all relevant phases of a fouling (by regulating filtration cycles and air scouring) or sludge resi-
modelling project in a similar way as activated sludge modelling is dence time. Similarly, additional performance criteria such as mem-
guided through the GMP guidelines (Rieger et al. 2013). Therefore, this brane fouling index or operational cost index, including air scouring,
protocol is taken as the key guidance to develop a framework for good may be considered.
practices in MBR integrated modelling projects. The rationale behind The second step, data collection and reconciliation, includes col-
consists on identifying the needs for extension or adaption of the acti- lecting the different type of data (influent and permeate quality, phys-
vated sludge GMP framework due to the particular characteristics of ical data related to the membrane compartment, and operational
MBRs. The MBR Modelling and Control Task Group from the Interna- parameters, such as filtration cycles or TMP set points for membrane
tional Water Association (IWA) is working towards the development of cleanings) and identify the data requirements (e.g. quantity and fre-
such a framework but some key ideas are already indicated here. The quency of the data), which might be very different depending on the
framework for MBR integrated modelling practice will be composed of modelling purpose and the dynamics of biological and filtration pro-
five steps: project definition, data collection and reconciliation, model cesses. The data reconciliation step should allow to identify gaps and
set-up, calibration and validation, and, simulation and results errors in the collected data and thus need for additional measuring
interpretation. campaigns for a proper validation. Typical additional data requirement
The project definition step includes defining the objective of the MBR would be dissolved oxygen and nutrient concentrations in the mem-
modelling study (e.g. design, operation/optimisation, control or opera- brane compartment to optimise nutrient removal and aeration control
tors training), the state variables, the modelling targets and the perfor- (Fenu et al., 2010).
mance indices to be used (permeate quality, membrane fouling or cost The third step requires the selection and set up of the models needed
evaluation criteria). Fig. 5 contains a schematic representation of how to describe MBR layout and performance; it means deciding on an

Fig. 5. Schematic representation of part of the framework for good MBR modelling practices. Where EQIgas = effluent quality index including the gas products (i.e.,
nitrous oxide, methane and CO2), OPEX = operating expenditure; CAPEX = Capital expenditure; TOC = total operating costs; MFI = modified fouling index; DFI =
Dimensionless Fouling Index; EQILIQ = Effluent Quality Index of the liquid phase).

8
G. Mannina et al. Bioresource Technology 329 (2021) 124828

influent model, reactor hydraulics, biological models, aeration and by integrating Life Cycle Assessment (LCA) based methodologies within
filtration models. As stated before, integrated models will be preferred the integrated MBR modelling. This poses a series of challenges. First,
here when results concerning the whole process are required. These the necessity to estimate a number of mass and energy flows throughout
modelling studies usually require additional state variables (with the plant, to build the input data set for LCA-based calculations. Sec-
respect to activated sludge modelling) for both the biological and ondly, the need of taking into account emerging contaminants of envi-
physical models (i.e. air flow rate for membrane scouring, SMP, EPS, ronmental concern (both in the sludge and the effluent wastewater).
resistance, TMP, permeate flow rate and chemical used for membrane This topic indeed hides another issue related to the growing con-
cleaning). Integrated MBR models have already shown a huge potential sciousness that chemical characterisation may fail in giving the real
in developing engineering solutions for MBR application and assessing picture of effluent properties, as measuring thousands of compounds and
various interactions between biological and physical models. As can be predicting the effect of their mixture is unfeasible (Pedrazzani et al.,
seen in Fig. 5, the integrated approach represents an important aspect of 2019a). This led to the development of bioassays for a more suitable
the modelling process, because it can correlate the biological and characterisation (Escher and Leusch, 2011; Gonzalez-Gil et al., 2016;
physical treatment by considering the most important operating vari- Papa et al., 2016a; Pedrazzani et al., 2020). Nevertheless, the direct use
ables that govern the MBR. In this case, a target can be set for the MBR of bioassays results in LCA-based procedures is not possible, unless
modelling within a treatment plant reality and it can be assessed by the proper conversion into equivalent pollutants mass flows is made (as
KPIs. It is necessary to select the most important KPIs, mainly related to suggested by Pedrazzani et al., 2019b; Di Trapani et al., 2015). Alter-
operating variables, depending on the aim of the modelling approach: native approaches have also been proposed for overcoming this limita-
design, operation optimization, or management. tion (Papa et al., 2013; 2016b).
For model selection, the general rule would be to keep model As for the second issue, which could be indeed considered as another
complexity as simple as possible to answer the modelling question. step in the decision making process, for assessing the plant suitability,
However, the feasibility of the framework application is related to the other areas should be taken into account and included: e.g. plant
available data set in order to balance among model accuracy and complexity and reliability; process flexibility; need of skilled personnel;
complexity. Sensitivity and uncertainty analysis can provide a good administrative and legal constraints, etc. Procedures for including these
response in finding a trade-off between model complexity and available items in a general evaluation and decision making framework, in the
data for model application (among others, Mannina et al., 2020a). case of MBR plants, are reported in the literature (Bertanza et al., 2017).
The fourth step, calibration and validation, has special features in
MBR modelling exercises. Depending on the complexity of the model, 6. Conclusion
many model parameters may be adjusted (namely, 122 from Mannina
et al. 2018c). Some of these parameters may be adjusted with external This review presented the remarks retrieved from peer-reviewed
experiments as respirometer or dead-end filtration test (Zarragoitia- papers regarding the integrated MBR modelling. On this behalf, a
González et al., 2008), however most of them are taken from the liter- clear simplified definition and a framework were proposed to pursue
ature. Still some are “manually adjusted” or just given (Janus, 2013), good practice for MBR modelling taking into account key process in-
others may be partially calibrated with optimisation protocols (Van- dicators such as effluent quality, membrane fouling, aeration, operating
rolleghem et al., 2003). A procedure advising to calibrate each target costs, energy consumption, and mitigation of GHG emissions. Literature
process (first biological nutrient removal processes, then filtration pro- review shows that the use of integrated mathematical models should be
cess or altogether simultaneously) and the default parameters for new more encouraged since they have the ability to provide comprehensive
processes encountered in MBRs is needed. results to gain more understanding concerning the functioning of an
The final step, simulation and results interpretation, refers to the MBR system.
definition, running and analysis of typical steady-state or dynamic sce-
narios in MBR modelling. It really depends on the objective of the
project. Declaration of Competing Interest
The final framework for good MBR modelling practice will also to
highlight the most relevant aspects in each step and provide guidelines The authors declare that they have no known competing financial
to support the application of this framework for different MBR model- interests or personal relationships that could have appeared to influence
ling applications. For example, suggesting the target variables, perfor- the work reported in this paper.
mance indicators, acceptable errors (for calibration), model type or
modelling scenarios as a function of the modelling objective. This sug- Acknowledgements
gestion may be illustrated with the application/benchmarking of
different MBR models for different data bases and case studies. This Giorgio Mannina holds the Chair of the Task Group on Membrane
proposed framework would benefit from including a discussion on the Bioreactor Modelling and Control of the International Water Associa-
potential benefits and limitations of using MBR models. Despite a first tion. This work is part of the activities carried out by the Task Group
attempt to apply a simplified scheme of the above proposed framework (http://www.iwa-network.org/groups/membrane-bioreactor-model-
has been provided for MBR pilot plants (i.e., Mannina et al., 2020b) ling-and-control/).
there is a need of more comprehensive applications to MBR systems in
real WWTPs. References

5. New perspectives Ahn, Y.T., Choi, Y.K., Jeong, H.S., Shin, S.R., 2006. Modeling of extracellular polymeric
substances and soluble microbial products production in a submerged MBR at
various SRTs. Water Sci. Technol. 53 (7), 209–216.
Despite a lot has been done, further work to make plant evaluation as Aquino, S.F., Stuckey, D.C., 2008. Integrated model of the production of soluble
wide (holistic) as possible is needed. Efforts may be addressed in two microbial products (SMP) and extracellular polymeric substances (EPS) in anaerobic
chemostats during transient conditions. Biochem. Eng. J. 38, 138–146.
different areas: 1) process performance assessment (which is more
Atanasova, N., Dalmau, M., Comas, J., Poch, M., Rodriguez-Roda, I., Buttiglieri, G., 2017.
properly linked to the main topic of this paper), and 2) evaluation of the Optimized MBR for greywater reuse systems in hotel facilities. J. Environ. Manage.
general plant suitability and sustainability. 193, 503–511.
As for the first point, instead of focusing on a few items (e.g. GHG Bai, R., Tien, C., 2000. Effect of deposition in deep-bed filtration: determination and
search of rete parameters. J. Colloid Interface Sci. 231, 299–311.
emissions, effluent quality index – EQI – based on a few parameters, Bertanza, G., Canato, M., Laera, G., Vaccari, M., Svanström, M., Heimersson, S., 2017.
etc.), a broader environmental footprint analysis should be performed, A comparison between two full-scale MBR and CAS municipal wastewater treatment

9
G. Mannina et al. Bioresource Technology 329 (2021) 124828

plants: techno-economic-environmental assessment. Environ. Sci. Pollut. Res. 24 Jiang, T., Myngheer, S., De Pauw, D.J.W., Spanjers, H., Nopens, I., Kennedy, M.D.,
(21), 17383–17393. Amy, G., Vanrolleghem, P.A., 2008. Modeling the production and degradation of
Bowen, W.R., Mohammad, A.W., Hilai, N., 1997. Characterization of nanofiltration soluble microbial products (SMP) in membrane bioreactors (MBR). Water Res. 42
membranes for perspective purposes – use of salts, uncharged solutes and atomic (20), 4955–4964.
force microscopy. J. Membr. Sci. 126, 91–105. Judd, S., 2006. The MBR Book: Principles and Applications of Membrane Bioreactors.
Bozkurt, H., van Loosdrecht, C.M., Gernaey, V., Sin, G., 2016. Optimal WWTP process Elsevier, Amsterdam.
selection for treatment of domestic wastewater – A realistic full-scale retrofitting Koutsou, O.P., Gatidou, G., Stasinakis, A.S., 2018. Domestic wastewater management in
study. Chem. Eng. J. 286, 447–458. Greece: greenhouse gas emissions estimation at country scale. J. Clean. Prod. 188,
Brepols, C., Comas, J., Harmand, J., Heran, M., Robles, Á., Rodriguez-Roda, I., Ruano, M. 851–859.
V., Smets, I., Mannina, G., 2020. Position paper – progress towards standards in Krzeminski, P., Leverette, L., Malamis, S., Katsou, E., 2017. Membrane bioreactors – A
integrated (aerobic) MBR modelling. Water Sci. Technol. 81 (1), 1–9. review on recent developments in energy reduction, fouling control, novel
Busch, J., Cruse, A., Marquardt, W., 2007. Modeling submerged hollow-fiber membrane configurations, LCA and market prospects. J. Membr. Sci. 527, 207–227.
filtration for wastewater treatment. J. Membr. Sci. 288 (1–2), 94–111. Kuberkar, V.T., Davis, R.H., 2000. Modeling of fouling reduction by secondary
Chaize, S., Huyard, A., 1991. Membrane bioreactor on domestic wastewater treatment membranes. J. Membr. Sci. 168, 243–257.
sludge production and modelling approach. Water Sci. Technol. 23, 1591–1600. Laspidou, C.S., Rittmann, B.E., 2002. A unified theory for extracellular polymeric
Charfi, A. Yang, Y. Harmand, Ben Amar J. N., Heran, M., Grasmick A. (2015). Soluble substances, soluble microbial products, and active and inert biomass. Water Res. 36,
microbial products and suspended solids influence in membrane fouling dynamics 2711–2720.
and interest of punctual relaxation and/or backwashing. Journal of Membrane Lee, Y., Cho, J., Seo, Y., Lee, J.W., Ahn, K.H., 2002. Modeling of submerged membrane
Science, 475, pp 156-166. bioreactor process for wastewater treatment. Desalination 146, 451–457.
Cho, J., Ahn, K.H., Seo, Y., Lee, Y., 2003. Modification of ASM No. 1 for a submerged Li, X., Wang, X., 2006. Modelling of membrane fouling in a submerged membrane
membrane bioreactor system: including the effects of soluble microbial products on bioreactor. J. Membr. Sci. 278, 151–161.
membrane fouling. Water Sci. Technol. 47, 177–181. Lu, S.G., Imai, T., Ukita, M., Sekine, M., Higuchi, T., 2002. Modeling prediction of
Dalmau, M., Monclús, H., Gabarrón, S., Rodriguez-Roda, I., Comas, J., 2014. Towards membrane bioreactor process with the concept of soluble microbial product. Water
integrated operation of membrane bioreactors: Effects of aeration on biological and Sci. Technol. 46 (11–12), 63–70.
filtration performance. Bioresour. Technol. 171, 103–112. Lu, S.G., Imai, T., Ukita, M., Sekine, M., Higuchi, T., Fukagawa, M., 2001. A model for
Dalmau, M., Rodriguez-Roda, I., Ayesa, E., Odriozola, J., Sancho, L., Comas, J., 2013. membrane bioreactor process based on the concept of formation and degra- dation of
Development of a decision tree for the integrated operation of nutrient removal soluble microbial products. Water Res. 35, 2038–2048.
MBRs based on simulation studies and expert knowledge. Chem. Eng. J. 217, Maere, T., Verrecht, B., Moerenhout, S., Judd, S., Nopens, I., 2011. BSM-MBR: A
174–184. benchmark simulation model to compare control and operational strategies for
Deng, L., Guo, W., Ngo, H.H., Zhang, H., Wang, J., Li, J., Wu, Y., 2016. Biofouling and membrane bioreactors. Water Res. 45, 2181–2190.
control approaches in membrane bioreactors. Bioresour. Technol. 221, 656–665. Mannina, G., Capodici, M., Cosenza., A., Di Trapani, A., Olsson, G. (2017c) Greenhouse
Di Bella, G., Mannina, G., Viviani, G., 2008. An integrated model for physical-biological gas emissions and the links to plant performance in a fixed-film activated sludge
wastewater organic removal in a submerged membrane bioreactor: model membrane bioreactor – Pilot plant experimental evidence, Bioresource and
development and parameter estimation. J. Membr. Sci. 322, 1–12. Technology, 241, 1145-1151.
Di Trapani, D., Di Bella, G., Mannina, G., Torregrossa, M., Viviani, G., 2015. Effect of C/N Mannina, G., Cosenza, A., 2013. The fouling phenomenon in membrane bioreactors:
shock variation on the performances of a moving bed membrane bioreactor. Assessment of different strategies for energy saving. J. Membr. Sci. 444, 332–344.
Bioresour. Technol. 189, 250–257. Mannina, G., Viviani, G., 2009. Separate and combined sewer systems: A long-term
Drews, A., Vocks, M., Bracklow, U., Iversen, V., Kraume, M., 2008. Does fouling in MBRs modelling approach. Water Sci. Technol. 60 (3), 555–565.
depend on SMP? Desalination 231 (1–3), 141–149. Mannina, G., Cosenza, A., Vanrolleghem, P., Viviani, G., 2011b. A practical protocol for
Fenu, A., Guglielmi, G., Jimenez, J., Sperandio, M., Saroj, D., Lesjean, B., Brepols, C., calibration of nutrient removal wastewater treatment models. J. Hydroinform. 13
Thoeye, C., Nopens, I., 2010. Activated sludge model (ASM) based modeling of (4), 575–595.
membrane bioreactor (MBR) processes: a critical review with special regard to MBR Mannina, G., Cosenza, A., Ekama, G., 2017a. Greenhouse gases from membrane
specificities. Water Res. 44, 4272–4294. bioreactors: Mathematical modelling, sensitivity and uncertainty analysis. Bioresour.
Escher, B., Leusch, F. (2011). Bioanalytical Tools in Water Quality Assessment. IWA Technol. 239, 353–367.
Publishing.ISBN13: 9781843393689. Mannina, G., Cosenza, A., Ekama, G., 2018a. A comprehensive integrated membrane
Fenu, A., Guglielmi, G., Jimenez, J., Sperandio, M., Saroj, D., Lesjean, B., Brepols, C., bioreactor model for greenhouse gas emissions. Chem. Eng. J. 334, 1563–1572.
Thoeye, C., Nopens, I., 2010. Activated sludge model (ASM) based modeling of Mannina, G., Cosenza, A., Ekama, G., 2018b. Mathematical modeling of greenhouse gas
membrane bioreactor (MBR) processes: a critical review with special regard to MBR emissions from membrane bioreactors: A comprehensive comparison of two
specificities. Water Res. 44, 4272–4294. mathematical models. Bioresour. Technol. 268, 107–115.
Freni, G., Mannina, G., 2010. Uncertainty in water quality modelling: The applicability of Mannina, G., Cosenza, A., Viviani, G., Ekama, G., 2018c. Sensitivity and uncertainty
Variance Decomposition Approach. J. Hydrol. 394 (3–4), 324–333. analysis of an integrated ASM2d MBR model for wastewater treatment. Chem. Eng.
Gabarrón, S., Dalmau, M., Porro, J., Rodriguez-Roda, I., Comas, J., 2015. Optimization of J. 351, 579–588.
full-scale membrane bioreactors for wastewater treatment through a model-based Mannina, G., Di Bella, G., Viviani, G., 2011a. An integrated model for biological and
approach. Chem. Eng. J. 267, 34–42. physical process simulations in membrane bioreactors (MBR). J. Membr. Sci. 376
Galinha, C, Sanches, S, Crespo, J. (2018). Fundamental Modelling of Membrane Systems. (1–2), 56–69.
Elsevier, 2018. Mannina, G., Di Bella, G., Viviani, G., 2010. Uncertainty assessment of a membrane
González Hernández, Y., Jáuregui Haza, U., Albasi, C., Alliet, M., 2015. Understanding bioreactor model using the GLUE methodology. Biochem. Eng. J. 52 (2–3), 263–275.
the Influence of Operating Parameters through in Silico Optimization of Energy Mannina, G., Ni, B.-J., Ferreira Reboucas, T., Cosenza, A., Olsson, G., 2020b. Minimizing
Consumption of Submerged Membrane Bioreactor for Urban Wastewater Treatment. membrane bioreactor environmental footprint by multiple objective optimization.
Desalin. Water Treat. 12, 1–13. Bioresour. Technol. 302, 122824.
Gonzalez-Gil, L. Papa, M. Feretti D., Ceretti, E. Mazzoleni, G. Steimberg, N. Pedrazzani, Mannina, G., Cosenza, A., Ferreira Reboucas, T., 2020a. Uncertainty and sensitivity
R. Bertanza, G.Juan M Lema, Carballa M. (2016): “Is anaerobic digestion effective analysis for reducing greenhouse gas emissions from wastewater treatment plants.
for the removal of organic micropollutants and biological activities from sewage Water Sci Technol 82 (2), 339–350.
sludge?” Water Research, 102, 211-220. Martin, C., Vanrolleghem, P., 2014. Analysing, completing, and generating influent data
Gujer, W., Henze, M., Mino, T., van Loosdrecht, M. (1999). Activated sludge model No. 3. for WWTP modelling: A critical review. Environ. Modell. Software 60, 188–201.
Water Sci. Technol. 39 (1), 183–193. Meng, F., Zhang, S., Oh, Y., Zhou, Z., Shin, H., Chae, S., 2017. Fouling in membrane
Hamedi, H., Ehteshami, M., Mirbagheri, Rasouli, S. A., Zendehboudi, S. (2019). Current bioreactors: An updated review. Water Res. 114, 151–180.
status and future prospects of membrane bioreactors (MBRs) and fouling Meng, F., Chae, S., Drews, A., Kraume, M., Shin, H., Yang, F., 2009. Recent advances in
phenomena: a systematic review. Can. J. Chem. Eng. 97, 32–58. membrane bioreactors (MBRs): membrane fouling and membrane material. Water
Henze, M., Grady, C.P.L.J., Gujer, W., Marais, G.v.R., Matsuo, T. (1987). Activated Res. 43, 1489–1512.
Sludge Model No. 1. IAWPRC Scientific and Technical Report No. 1, IAWPRC, Monclus, H., Buttiglieri, G., Ferrero, G., Rodriguez-Roda, I., Comas, Q., 2012.
London. Knowledge-based control module for start-up of flat sheet MBRs. Biore. Tech. 106,
Henze, M., Gujer, W., Mino, T., Matsuo, T., Wentzel, M.C.M., Marais G.V.R. (1995). 50–54.
Activated sludge model No. 2. IWA Scientific and Technical Report No. 3, London, Naessens, W., Maere, T., Ratkovich, N., Vedantam, S., Nopens, I., 2012. Critical review of
England. membrane bioreactor models – Part 2: Hydrodynamic and integrated models.
Hiatt, W.C., Grady, C.P.L., 2008. An updated process model for carbon oxidation, Bioresour. Technol. 122, 107–118.
nitrification, and denitrification. Water Environ. Res. 80, 2145–2156. Nagaoka, H., Yamanishi, S., Miya, A., 1998. Modeling of biofouling by extracellular
IPCC, Climate Change (2013) The physical science basis. In: Contribution of working polymers in a membrane separation activated sludge system. Water Sci. Technol. 39,
group I to the fifth assessment report of the intergovernmental panel on climate 497–504.
change. Cambridge University Press, Cambridge, New York, p 1535. Namkung, E., Rittmann, B.E., 1986. Soluble microbial products (SMP) formation kinetics
Jang, N., Ren, X., Cho, J., Kim, I.S., 2006. Steady-state modeling of bio-fouling potentials by biofilms. Water Res. 20, 795–806.
with respect to the biological kinetics in the sMBR. J. Membr. Sci. 284, 352–360. Ng, A.N.L., Kim, A.S., 2007. A mini-review of modeling studies on membrane bioreactor
Janus, T., Ulanicki, B., 2010. Modeling SMP and EPS formation and degradation kinetics (MBR) treatment for municipal wastewaters. Desalination 212 (1–3), 261–281.
with an extended ASM3 model. Desalination 261, 117–125. Ni, B., Yuan, Z., 2015. Recent advances in mathematical modeling of nitrous oxides
Janus, T., 2013. Modelling and Simulation of Membrane Bioreactor for Wastewater emissions from wastewater treatment processes. Water Res. 87, 336–346.
Treatment (Ph.D. thesis). De Montfort University.

10
G. Mannina et al. Bioresource Technology 329 (2021) 124828

Ognier, S., Wisniewski, C., Grasmick, A., 2004. Membrane bioreactor fouling in sub- Smith, C. W., Di Gregorio, D., and Talcott, R. M. (1969). The use of ultrafiltration
critical filtration conditions: a local critical flux concept. J. Membr. Sci. 229, membrane for activated sludge separation. Proceedings of the 24th Annual Purdue
171–177. Industrial Waste Conference, Purdue University, West Lafayette, Indiana, USA.
Orhon, D., Artan, N., Cimcit, Y., 1989. The concept of soluble residual product formation 1300–1310.
in the modelling of activated sludge. Water Sci. Technol. 21, 339–350. Suganthi, V., Mahalakshmi, M., Balasubramanian, N., 2013. Development of hybrid
Papa, M., Ceretti, E., Viola, G.C.V.D., Feretti, I., Zerbini, G., Mazzoleni, N., Steimberg, R. membrane bioreactor for tannery effluent treatment. Desalination 309, 231–236.
Pedrazzani, Bertanza, G., 2016a. The assessment of WWTP performance: towards a Vanrolleghem, P.A., Insel, G., Petersen, B., Sin, G., De Pauw, D., Nopens, I., Weijers, S.,
jigsaw puzzle evaluation? Chemosphere 145, 291–300. Gernaey, K., 2003. A comprehensive model calibration procedure for activated
Papa, Pedrazzani, R., Bertanza, G., 2013. How green are environmental technologies? A sludge models. In: Proceedings: WEFTEC 2003, 76th Annual Technical Exhibition
new approach for a global evaluation: the case of WWTP effluents ozonation” -. and Conference. October 11–15, 2003, Los Angeles, CA, USA (on CD-ROM).
Water Res. 47, 3679–3687. Wintgens, T., Rosen, J., Melin, T., Brepols, C., Drensla, K., Engelhardt, N., 2003.
Pedrazzani R., Baroni P., Feretti D., Mazzoleni G.,Steimberg N., Urani C., Viola G., Modelling of a membrane bioreactor system for municipal wastewater treatment.
Zerbini I., Ziliani E., Bertanza G. (2020) Methodological Protocol for Assessing the J. Membr. Sci 216, 55–65.
Environmental Footprint by Means of Ecotoxicological Tools: Wastewater Treatment Wu, J., He, C., Zhang, Y., 2012. Modeling membrane fouling in a submerged membrane
Plants as an Example Case. In: Ecotoxicological QSARs, Methods in Pharmacology bioreactor by considering the role of solid, colloidal and soluble components.
and Toxicology, Kunal Roy (ed.), https://doi.org/10.1007/978-1-0716-0150-1_14, J. Membr. Sci. 397–398, 102–111.
© Springer Science+Business Media, LLC, part of Springer Nature 2020. Xiao, K., Xu, Y., Liang, S., Lei, T., Sun, J., Wen, X., Zhang, H., Chen, C., Huang, X., 2014.
Pedrazzani, E., Ziliani, I., Cavallotti, E., Bollati, M., Ferreri, G. Bertanza, 2019b. Use of Engineering application of membrane bioreactor for wastewater treatment in China:
ecotoxicology tools within the environmental footprint evaluation protocols: the current state and future prospect. Front. Environ. Sci. Eng. 8, 805–819.
case of wastewater treatment plants. Desalin. Water Treat. 172, 2–14. Yamamoto, K., Hiasa, M., Mahmood, T., Matsuo, T., 1989. Direct solid-liquid separation
Pedrazzani, G., Bertanza, I., Brnardić, Z., Cetecioglu, J., Dries, J., Dvarionienė, A.J., using hollow fiber membrane in activated sludge aeration tank. Water Sci. Technol.
García-Fernández, A., Langenhoff, G., Libralato, G., Lofrano, B., Škrbić, E., Martínez- 30, 21–27.
López, S., Meriç, D. Mutavdžić, Pavlović, M., Papa, P., Schröder, K.P., Tsagarakis, C. Yang, M., Yu, D., Liu, M., Zheng, L., Zheng, X., Wei, Y., Wang, F., Fan, T., 2017.
Vogelsang, 2019a. Opinion paper about organic trace pollutants in wastewater: Optimization of MBR hydrodynamics for cake layer fouling control through CFD
Toxicity assessment in a European perspective. Sci. Total Environ. 651, 3202–3221. simulation and RSM design. Bioresour. Technol. 227, 102–111.
https://doi.org/10.1016/j.scitotenv.2018.10.027. Zarragoïtia, A., Schetrite, S., Jáuregui-Haza, U., Lorain, O., Albasi, C., 2009.
Pocquet, M., Wu, Z., Queinnec, I., Spérandio, M., 2016. A two pathway model for N2O Optimization of Wastewater Filtration Process in Submerged Membrane Bioreactors:
emissions by ammonium oxidizing bacteria supported by the NO/N2O variation. Applicability of a Dynamic Model to Scale Up. Computer Aided Chemical
Water Res. 88, 948–959. Engineering 27, 1545–1550.
Puyol, D., Batstone, D.J., Hülsen, T., Astals, S., Peces, M., Krömer, J.O., 2016. Resource Zarragoitia-González, A., Schetrite, S., Alliet, M., Jáuregui-Haza, U., Albasi, C., 2008.
Recovery from Wastewater by Biological Technologies: Opportunities, Challenges, Modelling of submerged membrane bioreactor: Conceptual study about link between
and Prospects. Front. Microbiol. 7, 1–23. activated slugde biokinetics, aeration and fouling process. J. Membr. Sci. 325,
Rieger, L., Gillot S., Langergraber, G., Ohtsuki, T., Shaw, A., Takács, I., and Winkler, S. 612–624.
(2013) Guidelines for Using Activated Sludge Models. Scientific and Technical Zhanga, G., Ngo, H.H., Peng, Y., Bux, F., Mannina, G. (2021) Biological nutrients removal
Report 22. London [u.a]: IWA Publ. and recovery, Bioresource Technology, Volume 320, Part B, 124377.
Robles, A., Ruano, M.V., Ribes, J., Ferrer, J., 2013. A filtration model applied to Zheng, F., Huang, L., Pan, J., Fan, L., Wang, S., Tan, C., 2018. Does influent surface
submerged anaerobic MBRs (SAnMBRs). J. Membr. Sci 444, 139–147. organic loading and aeration mode affect nitrogen removal and N2O emission in
Saltelli, A., Tarantola, A., Campolongo, F., 2004. Ratto (2004) Sensitivity analysis in subsurface wastewater infiltration systems? Ecol. Eng. 123, 168–174.
practice. John Wiley & Sons, Chichester, UK, A guide to assessing scientific models Zuthi, M.F.R., Guo, W.S., Ngo, H.H., Nghiem, L.D., Hai, F.I., Xia, S., Li, J., Li, J., Liu, Y.,
Probability and Statistics Series. 2017. New and practical mathematical model of membrane fouling in an aerobic
Sarioglu, M., Insel, G., Orhon, D., 2012. Dynamic in-series resistance modeling and submerged membrane bioreactor. Bioresour. Technol. 238, 86–94.
analysis of a submerged membrane bioreactor using a novel filtration mode. Zuthi, M.F.R., Ngo, H.H., Guo, W.S., 2012. Modeling bioprocesses and membrane fouling
Desalination 285, 285–294. in membrane bioreactor (MBR): A review towards finding an integrated model
Saroj, D., Guglielmi, G., Chiarani, D., Andreottola, G., 2008. Modeling and simulation of framework. Bioresour. Technol. 122, 119–129.
membrane bioreactors by incorporating simultaneous storage and growth concept: Zuthi, M.F.R., Ngo, H.H., Guo, W.S., Zhang, J., Liang, S., 2013. A review towards finding
an especial attention to fouling while modeling the biological process. Desalination a simplified approach for modeling the kinetics of the soluble microbial products
221, 475–482. (SMP) in an integrated mathematical model of membrane bioreactor (MBR). Int.
Scholes, E., Verheyen, V., Brook-Carter, P., 2016. A review of practical tools for rapid Biodeterior. Biodegrad. 85, 466–473.
monitoring of membrane bioreactors. Water Res. 102, 252–262.

11