Advanced Powder Technology xxx (xxxx) xxx

Contents lists available at ScienceDirect

Advanced Powder Technology

journal homepage: www.elsevier.com/locate/apt

Perspective article

Flowsheet simulation of solids processes: Current status and future

trends
Maksym Dosta a,⇑, James D. Litster b, Stefan Heinrich a
a
Institute of Solids Process Engineering and Particle Technology, Hamburg University of Technology, Germany
b
Department of Chemical and Biological Engineering, University of Sheffield, England, UK

a r t i c l e i n f o a b s t r a c t

Article history: Complex manufacturing processes are nowadays applied for production of various solid products. It is
Received 11 September 2019 very common that for production of particles with desired properties several transformation steps like
Received in revised form 4 December 2019 drying, milling, classification, granulation, etc. should be involved. This leads to the process structures
Accepted 11 December 2019
consisting of different apparatuses or transformation substeps connected with material and energy bal-
Available online xxxx
ances. Consequently, development of new processes or optimization of already existing, as well as an
optimal control, is a very challenging task, which can be partially solved using numerical modelling.
Keywords:
For the simulation of modern production processes, the flowsheet calculations can be effectively used.
Flowsheet simulation
Solids processes
Starting from the 80 s a lot of work focused on the flowsheet simulation of liquid-vapor systems has been
Process modelling done and as result various well-established systems exist today. With respect to the solid processes the
Simulation frameworks intensive research has been started much later. In this contribution we present our view about a current
role of flowsheet simulation for modeling of particulate materials and specify the open fields which can
be covered in future research.
Ó 2019 The Society of Powder Technology Japan. Published by Elsevier B.V. and The Society of Powder
Technology Japan. All rights reserved.

1. Why flowsheet simulation? is very inefficient and can be applied only for very simple
structures.
Most of industrial processes which are used to manufacture Generally, flowsheet simulation can be performed in steady-
products in form of particulate materials involve many transforma- state or in dynamic mode. From the computational point of view,
tion steps occurring in different equipment. In Fig. 1 the structures the simulation, as well as the development of models, is simpler
for such processes such as cement manufacturing [1], continuous for the steady-state analysis. However, despite the fact that the
fluidized bed granulation [2,3], concentrator plant [4] as well as dynamic modeling is a more challenging task, process behavior
continuous tablet manufacturing [5] are illustrated. Nowadays, during start-up or shut-down phases, transient process behavior,
the major part of manufacturing processes is operated in the con- development of dynamic control strategies, etc. can be investigated
tinuous mode. Even in the relatively conservative areas, such as only in this mode. In the ideal case, the advanced simulation
pharmaceutical industry, the migration from batch-wise to contin- framework should provide a possibility to combine steady-state
uous operation mode can be observed [6–8]. and dynamic analysis modes [9].
The modeling of the behavior of integrated production pro- Not all unit operations in solids process engineering show tran-
cesses cannot be done based only on the information of single pro- sient behavior that have a noticeable influence on the dynamics of
cess units. Due to the existence of recycle streams and application the whole process. Many unit operations with relatively small
of strategies for process control and plant wide optimisation, the holdup mass such as screens, some types of mills or mixers can
behavior of each single unit can have a strong influence not only be efficiently treated as steady-state models during dynamic sim-
on the downstream processes but on the entire process. Therefore, ulation [2,3]. However, in some cases, dynamics of the same clas-
it is necessary to treat the integrated system as a whole. For this sification or solid-fluid separation units cannot be neglected and
purpose, flowsheet simulation can be effectively applied. Alterna- play an important role [10,11].
tive approaches, such as a manual subsequent calculation of units, The flowsheet calculations can be used for four main purposes:
modeling of process behavior, sensitivity analysis, process
optimization and process control. In most cases, empirical or
⇑ Corresponding author. semi-empirical models are used for the simulations [2,3,12,13].

https://doi.org/10.1016/j.apt.2019.12.015
0921-8831/Ó 2019 The Society of Powder Technology Japan. Published by Elsevier B.V. and The Society of Powder Technology Japan. All rights reserved.

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
2 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx

Fig. 1. Different solids manufacturing processes with complex process structures. Reprinted from [1, 4 and 5] with permission from Elsevier.

Therefore, only macroscopic states like mass flows, particle size comprehensive material characterization are not considered. Many
distributions, temperatures, etc. can be predicted. At the same such properties like an internal granule microstructure, mechani-
time, numerous product properties which play important role for cal properties like stiffness or strength, spatial distribution of com-

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx 3

ponents, etc. are neglected or estimated only partially. From the sequence, more recently new calculation approaches and models
computational point of view there already exist methods to treat have been developed especially for treatment of solids.
such information such as multidimensional population balance For the correct processing of multidimensional distributed
models or transformation matrices [14]. The main challenge is in parameters during steady-state and dynamic calculations transfor-
the derivation of mechanistic or first-principles models, where mation matrices have been effectively applied [2,23,29]. Instead of
knowledge of microscale processes like a formation of nuclei, their explicit calculation of all output variables, the transformation laws
further agglomeration, densification [15], etc. is properly consid- are formulated for each model and a transformation matrix is gen-
ered in a macroscale model. erated. This matrix is afterwards applied to transform holdup and
Important benefit which can be gained from the use of flow- input streams implicitly. Such implicit calculation makes it possi-
sheet simulation is to answer the question ‘‘what if?”. Modelling ble to extend applicability of model significantly. In this case the
the variation of process parameters or even process structures models developed for strictly limited dimensions of parameters
can be effectively used for a wide range of tasks, starting with per- space can be used for a larger number of dimensions [14].
sonnel training [16] ending with process optimization. Using From the computational point of view, there exist equation-
flowsheet-based sensitivity [17] or bifurcation [18] analysis, the oriented (simultaneous) and sequential-modular (modular)
influence of process parameters can be investigated and, for exam- approaches [9]. For an application of the equation-oriented
ple, stable steady-state regions identified. Furthermore, flowsheet approach, open-form models are needed [30], whereby in the case
simulation plays an important role in the development of control of modular approach ‘‘black box” type of models can be modeled
strategies [19–22]. [31]. This advantage can play an important role for the simulation
Nowadays, there are several robust software systems which can of solids processes, where mathematical models of single process
be applied for the flowsheet simulation of solids processes. Most of units are often of heterogeneous nature and contain discontinu-
them such as Aspen Plus (Aspen Technology Inc.), gPROMS Formu- ities. Equation-oriented approaches are standard in commercial
lated Products (Process Systems Enterprise Ltd.), JKSimMet (JKTech software platforms such a gProms. They provide fast and robust
Pty Ltd.), CHEMCAD (Chemstations Inc.) are commercial products. solutions and are amenable to optimization and global sensitivity
In contrast, the Dyssol framework which was especially developed analysis. Flowsheeting packages designed for solids processing
for solids processes is an open-source system [23]. All of these sys- can easily handle distributions of one internal ordinate and
tems contain library of units that allow the users to get an access to solution techniques are suitable for one dimensional population
the up-to-date knowledge database. balances [31]. However, some processes like crystallization, granu-
lation, drying etc. may be described by multi-dimensional popula-
tion balance equations which are partial integro-differential
2. Complexity of solids
equations [32]. As a result, the combining of all models into a sin-
gle equation set and application of equation-oriented approach is a
The first flowsheet simulation methods and frameworks were
rather challenging task. Thus, some of the flowsheet frameworks
developed for modeling of liquid-vapor systems [24,25]. The
such as Dyssol [23] are based on the sequential-modular approach.
necessity to distinguish between processes for manufacturing of
For the modular dynamic flowsheet simulation of solids, a
products in fluid or solid form has been pointed out by different
waveform relaxation (Picard-Lindelöf iteration) approach [33] has
authors [2,23,26–28]. One of the most decisive differences is the
been proposed [2]. Using waveform relaxation, the whole simula-
characterization of the material. In case of vapor-liquid systems,
tion time is divided into smaller time windows. Afterwards all
the material can be completely described by a set of bulk parame-
models are iteratively solved on this window and the convergence
ters, typically thermodynamic properties. In contrast, a compre-
is analyzed. If the convergence is reached the calculations are
hensive description of particulate materials requires the use of
started from a new window. This method can be effectively applied
multidimensional distributed parameters. When one property
to perform simulation on several abstraction levels:
changes, other secondary attributes also change. For example, dur-
ing an agglomeration process not only particle enlargement takes
 to couple different simulation frameworks [34];
place, but also the distribution of particles over other property
 to perform the process simulation within one framework [2];
coordinates such as porosity, form factor or chemical composition
 for component-based simulation on the scale of a single process
are simultaneously changed (Fig. 2). Therefore, solids processes
unit [35].
require more complex models for different process units. As a con-

3. Models of single process units

The models of single process units play the crucial role in the
flowsheet simulations. The models can be generally distinguished
by detailing levels and application purposes. Werther et al. [36]
have proposed three main levels:

 level 1: short-cut models, where no knowledge of specific appa-

ratus is required. This type of models can be used for rough esti-
mation at an early stage process or product development;
 level 2: semi-empirical and physical models, which can be applied
for process optimization, de-bottlenecking or unit design;
 level 3: comprehensive physical models. These models are based
on microscale simulation approaches like CFD, DEM, MD, etc.
and allow to perform detailed unit design or optimization.

Only the models of the first and second level can be directly
Fig. 2. Two-dimensional particle distribution after agglomeration process [14]. solved with a flowsheet simulator. For more sophisticated models

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
4 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx

of the third level the incorporation of other type of simulators is ment of future models. The measurements may be done to charac-
needed. terize raw materials, products, intermediates and especially key
Compared to the models developed for liquid-vapor systems, process parameters which influence the efficiency [49]. The
the major part of models for solids processes have more complex straightforward strategy to use experimental data is the estimation
behavior and reveals heterogeneous nature. This is caused due to of unknown model parameters, where adjusted parameters are
the fact, that: varied with a goal to minimize discrepancy between numerical
and experimental results [50–51]. In most cases a raw experimen-
 many models for particulate processes are based on various tal data should be pre-processed to reduce volume and dimension-
empirical correlations which have been obtained for a ality of data and to treat problems caused by measurement error
strictly limited parameter space. As a result, combination of [52].
two or more correlations into one model can lead to large While few models are fully predictive, robust, mechanistically
discontinuities; based models exist for most unit operations and flowsheeting tools
 even small deviations in process conditions can cause the tran- are now mature enough for a move to Model Driven Design
sition between rate processes occurring in a unit operation. For (MoDD) for many particulate processes and products. Here the
example, increased flow rate of suspension mass flow injected simulation drives the design and experiments are used primarily
into fluidized bed granulator, can change particle growing to refine and validate the model. Robust, general workflows for
mechanism from coating to aggregation. Often, these transi- MoDD are still being developed. These workflows should minimize
tions can be linked to different transformation regimes that the number of model parameters that needed to be backfitted from
are expressed by dimensionless regime maps or design spaces pilot or full-scale experiments. Sensitivity analysis using the flow-
[37]. sheet simulation is a powerful tool for choosing key parameters
and designing appropriate experiments [48,53].
4. Future directions Other strategies to use experimental data are based on the gen-
eration of semi-parametric or nonparametric models. They can be
4.1. Multiscale simulations generally classified into three main groups, depending on the type
of knowledge which models are based on [54]:
One of the most promising future directions is the multiscale
process treatment, where submodels from different time and  parametric: ‘‘conventional” type of models which are formu-
length scales are linked together to obtain detailed process lated a priori, based on knowledge about process;
description [36,38]. Most of currently available unit operation  nonparametric: formulated exclusively from data;
models are empirical or semi-empirical nature. In order to improve  hybrid (semi-parametric): combines parametric and nonpara-
these models and to estimate unknown model parameters [39], metric in the parallel or in serial arrangement [55].
microscale simulation techniques like DEM, CFD, SPH etc. can be
applied to describe processes occurring within single process unit For the generation of a nonparametric model different types of
with a higher detailing grade. The central role here plays model artificial neural networks (ANNs), such as radial basis function net-
decomposition and inter-scale relations [40]. Depending on the work and multilayer perceptron can be used. These types of mod-
way how the models are linked, Ingram et al. [41] have proposed els have been applied for different apparatuses from solids process
to distinguish five main categories: multidomain, embedded, par- engineering like crystallizer [56], mills [57], granulators [58] etc.
allel, serial and simultaneous. These linking strategies have been For training of ANN’s not only real experimental data can be used,
used for investigations of different processes like granulation or but also the data generated from the microscale simulations, like
classification [10,42–45]. for example, from DEM calculations [58].
Nowadays, there are two main limitation factors which hinder Finally, the structured gathering of the process data and incor-
the further industrial usage of multiscale simulations: poration of this data into virtual image (flowsheet model) of phys-
ical process allows to build digital twins. Such coupling between
 large computational effort caused due to the usage of the micro- real process and its digital representation can be realized in both
scale models; directions. On the one hand, the data flow is going from a physical
 simultaneous usage of various simulation approaches and auto- object to digital model through sensor updates to mirror behavior
matic inter-scale data transfer. of a corresponding twin [59]. On the other hand, the flow in reverse
direction allows to use a twin as a controlling instance. For exam-
The first problem can be partially solved using the modern com- ple, the digital twin can be used for in silico experiments for tuning
puter architectures, such as GPUs and applying coarse-graining sophisticated control strategies such as model predictive control
methods, such as the multiphase particle in cell MPPiC [46] without need for expensive and material intensive experiments
method. To solve the second challenge, different numerical meth- on the real plant (see Fig. 3).
ods such as MD, CFD, DEM, PBM, etc. should be directly integrated
into one general software framework and the whole calculation 4.3. Simulation frameworks
procedure consisting of model decomposition, inter-scale data
transfer, analysis of the inter-scale convergence should be auto- In the next decade we expect further intensive developments
mated. Brief overview about some of proposed frameworks and focused on multiscale process treatment. Thus, the interoperability
interfaces can be found in Groen et al. [47]. Such frameworks are between different simulation frameworks and the data exchange
beginning to become available even in commercial software plat- between them will be critical. In recent years, attempts have been
forms [48]. made on the integration of different tools. For example, Shopfer
et al. [30] have proposed a component-based platform for integra-
4.2. Incorporation of experimental data tion of modeling tools. Especially for the chemical process engi-
neering, the CAPE-OPEN standard has been developed and
The incorporation of the available off-line and on-line measured integrated into several simulation frameworks [61]. For the multi-
process data into the mathematical models of integrated process or disciplinary data exchange, Fillinger et al., [62] have proposed to
into single process steps is one of key directions in the improve- use functional mock-up interface.

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx 5

Fig 3. Using a process simulation digital twin for twin screw granulation of pharmaceuticals to tune a model predictive controller [60]. Integration of gPROMS and
PharmaMV. Reprinted with kind permission from Gavin Reynolds.

With respect to further development and implementation of granular materials. In many cases, distribution of particles over
simulation frameworks we expect increased research in the fol- only one property coordinate, namely size, is predicted with a high
lowing fields: accuracy. Meanwhile, many properties such as particle porosity,
internal structure, surface properties, etc. are being left behind.
 intensified usage of modern computer architectures like graphic However, this information is of major importance for many appli-
processor units to improve calculation performance for solution cations of granular product design, such as the development of
of complex multidimensional problems; pharmaceutical products where information of internal structure
 migration from desktop simulator to the cloud-based solutions. allows estimation of a priori dissolution time [63]. Furthermore,
The cloud computing and development of Web-based user incorporating additional information about solid product into
interfaces can significantly simplify usage of flowsheeting tools. flowsheet simulation is a necessary step to solve inverse problems:
find optimal process configuration and parameters for a given pro-
duct specification (see Fig. 4).
4.4. Linking process and product models
With respect to the detailed characterization of a final solid pro-
duct, the further research should be performed in three main
The currently available unit operation models and process sub-
areas:
steps do not allow to describe multidimensional properties of

Fig. 4. Combining product and process engineering for particulate product design and manufacture [67].

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
6 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx

 extension of mathematical models for single process units, [10] F. Elskamp, H. Kruggel-Emden, M. Hennig, U. Teipel, Discrete element
investigations of process models for batch screening under altered
where such effects like influence of the thermal conditions onto
operational conditions, Powder Technol. 301 (2016) 78–95.
granule morphology will be considered [64]; [11] M. Gleiss, S. Hammereich, M. Kespe, H. Nirschl, Application of the dynamic
 development of numerical methods for proper treatment of flow sheet simulation concept to the solid-liquid separation: Separation of
multidimensional distributed parameters. This can include sol- stabilized slurries in continuous centrifuges, Chem. Eng. Sci. 163 (2017) 167–
178.
vers of multidimensional PBM, like finite volume method [65], [12] S.-Y. Park, S.C. Galbraith, H. Liu, H. Lee, B. Cha, Z. Huan, T. O’Connor, S. Lee, S.
calculation approaches like transformation matrices [14] or Yoon, Prediction of critical quality attributes and optimization of continuous
alternative methods like Monte Carlo approach [66]; dry granulation process via flowsheet modeling and experimental validation,
Powder Technol. 330 (2019) 461–470.
 development of a much better library of product performance [13] V. Skorych, M. Dosta, E.-U. Hartge, S. Heinrich, R. Ahrens, S. Le Borne,
models to link with the process simulations. Investigation of FFT-based solver applied to dynamic flowsheet simulation of
agglomeration processes, Adv. Powder Technol. 30 (2019) 555–564.
[14] V. Skorych, N. Das, M. Dosta, J. Kumar, S. Heinrich, Application of
transformation matrices to the solution of population balance equations,
5. Concluding remarks Processes 7 (8) (2019).
[15] S.M. Iveson, J.D. Litster, K. Hapgood, B.J. Ennis, Nucleation, growth and
breakage phenomena in agitated wet granulation process: a review, Powder
In the recent years increased interest to the flowsheet simula-
Technol. 4 (2001) 3–39.
tion of solids processes can be observed. The fact that particulate [16] D.S. Pattle, Z. Ahmad, G.P. Rangaiah, Operator training simulator in the
materials should be treated differently compared to the liquid- chemical industry: review, issues, and future directions, Rev. Chem. Eng. 30
(2014) 199–216.
vapor systems lead to the development of new models, calculation
[17] D. Schwier, E.-U. Hartge, J. Werther, G. Gruhn, Global sensitivity analysis in the
approaches and implementation of simulation frameworks. As a flowsheet simulation of solids processes, Chem. Eng. Process. 49 (2010) 9–21.
result, there exist different systems which already allow to simu- [18] R. Radichkov, T. Müller, A. Kienle, S. Heinrich, M. Peglow, L. Mörl, A numerical
late and to optimize different production processes with a rela- bifurcation analysis of continuous fluidized bed spray granulation with
external product classification, Chem. Eng. Process.: Process Intensification
tively high detailing grade. Simulation of solids flowsheets have 45 (2006) 826–837.
now matured to the point where they can be used as powerful [19] J. Zhang, J.D. Litster, F.Y. Wang, I.T. Cameron, Evaluation of control strategies
tools for design, optimization and control in industry. for fertiliser granulation circuits using dynamic simulation, Powder Technol. 8
(2000) 122–129.
In the near future we expect intensification of research in the [20] A. Bück, R. Dürr, M. Schmidt, E. Tsotsas, Model predictive control of continuous
four main fields: layering granulation in fluidised beds with internal product classification, J.
Process Contr. 45 (2016) 65–75.
[21] G.C. Pereira, S.V. Muddu, A.D. Roman-Ospino, D. Clancy, B. Igne, C. Airiau, D.
 multiscale process treatment direct inclusion of microscale sim- Muzzio, M. Ierapetritou, R. Ramachandran, R. Singh, Combined feedforward/
ulation methods (DEM, CFD, SPH, etc.) to macroscale flowsheet feedback control of an integrated continuous granulation process, J. Pharm.
model; Innov. 1–27 (2018).
[22] C. Neugebauer, E. Diez, A. Bück, S. Palis, S. Heinrich, A. Kienle, On the dynamics
 incorporation of experimental data to generate nonparametric
and control of continuous fluidized bed layering with screen-mill-cycle,
or hybrid models and to build digital twins; Powder Technol. 354 (2019) 765–778.
 more sophisticated product models that make use of more [23] V. Skorych, M. Dosta, E.-U. Hartge, S. Heinrich, Novel system for dynamic
flowsheet simulation of solids processes, Powder Technol. 314 (2017) 665–
detailed structural characterization of the product, and more
679.
sophisticated process models to provide this structural [24] V. Hlavacek, Analysis of a complex plant steady-state and transient behavior,
information; Comp. Chem. Eng. 1 (1977) 75–100.
 further development of simulation software into directions of [25] M. Shacham, S. Macchieto, F. Stutzman, P. Babcock, Equation oriented
approach to process flowsheeting, Comput. Chem. Eng. 6 (1982) 79–95.
unified multiscale frameworks, cloud computing and Web- [26] A.P. Rossiter, J.M. Douglas, Design and optimization of solids processes. Part 1.
based user interfaces. A hierarchical decision procedure for process synthesis of solids systems,
Chem. Eng. Res. Des. 64 (1986) 175–195.
[27] F. Muzzio, T. Shinbrot, B.J. Glasser, Powder technology in the pharmaceutical
industry: the need to catch up fast, Powder Technol. 124 (2002) 1–7.
[28] E.U. Hartge, M. Pogodda, C. Reimers, D. Schwier, G. Gruhn, J. Werther,
References Flowsheet simulation of solids processes, KONA 24 (2006) 146–156.
[29] M. Pogodda Development of an advanced system for the modelling and
simulation of solids processes. Diss. Aachen 2007.
[1] U. Kääntee, R. Zevenhoven, R. Backman, M. Hupa, Cement manufacturing using
[30] G. Schopfer, A. Yang, L. von Wedel, W. Marquardt, CHEOPS: A tool-integration
alternative fuels and the advantages of process modelling, Fuel Process.
platform for chemical process modeling and simulation, Int. J. Softw. Tools.
Technol. 85 (2004) 293–301.
Technol. Transf. (2004) 186–202.
[2] M. Dosta, S. Heinrich, J. Werther, Fluidized bed spray granulation: analysis of
[31] E. Gavi, G.K. Reynolds, Systems modeling of a tablet manufacturing process,
the system behaviour by means of dynamic flowsheet simulation, Powder
Comput. Chem. Eng. 71 (2014) 130–140.
Technol. 204 (2010) 71–82.
[32] D. Ramkrishna, Population balances. Theory and applications to particulate
[3] I.M. Cotabarren, D.E. Bertín, V. Bucalá, J. Piña, A validated flowsheeting tool for
systems in engineering, Academic Press, London, 2000.
the study of an industrial granulation process, Ind. Eng. Chem. Res. 52 (2013)
[33] E. Lelarasmee, The waveform relaxation method for time domain analysis of
15198–15210.
large scale integrated circuits: Theory and application, University of California,
[4] A. Remes, J. Aaltonen, H. Koivo, Grinding circuit modeling and simulation of
Berkley: Diss., 1982.
particle size control at Siilinjärvi concentrator, Int. J. Miner. Process. 96 (2010)
[34] V. Kulikov, H. Briesen, R. Grosch, A. Yang, L. Wedel, W. Marquardt, Modular
70–78.
dynamic simulation for integrated particulate processes by means of tool
[5] R. Singh, M. Ierapetritou, R. Ramachandran, Hybrid advanced control of flexible
integration, Chem. Eng. Sci. 60 (2005) 2069–2083.
multipurpose continuous tablet manufacturing process via direct compaction,
[35] M. Dosta, Modular based simulation of single process units, Chem. Eng. Techn.
Proc. of 23rd Europ. Symp. Comp. Aided Proc. Eng. (2013) 757–761.
42 (2019) 699–707.
[6] K. Plumb, Continuous processing in the pharmaceutical industry. Changing the
[36] J. Werther, S. Heinrich, M. Dosta, E.-U. Hartge, The ultimate goal of modeling—
mind set, Chem. Eng. Res. Design 83 (2005) 730–738.
Simulation of system and plant performance, Particuology 9 (2011) 320–329.
[7] I.R. Baxendale, R.D. Braatz, B.K. Hodnett, K.F. Jensen, M.D. Johnson, P. Sharratt,
[37] D. Kayrak-Talay, S. Dale, C.R. Wassgren, J.D. Litster, Quality by design for wet
J.-P. Sherlock, A.J. Florence, Achieving continuous manufacturing: Technologies
granulation in pharmaceutical processing: Assessing models for a priori design
and approaches for synthesis, workup, and isolation of drug substance May
and scaling, Powder Techn. 240 (2013) 7–18.
20–21, 2014 Continuous Manufacturing Symp, J. Pharm. Sci. 104 (2015) 781–
[38] K.-U. Klatt, W. Marquardt, Perspectives for process systems engineering –
879.
Personal views from academia and industry, Comput. Chem. Eng. 33 (2009)
[8] S. Byrn, M. Futran, H. Thomas, E. Jayjock, N. Maron, R.F. Meyer, A.S. Myerson, M.
536–550.
P. Thien, B.L. Trout, Achieving continuous manufacturing for final dosage
[39] M. Dosta, S. Antonyuk, E.-U. Hartge, S. Heinrich, Parameter estimation for the
formation: challenges and how to meet them May 20–21 2014 Continuous
flowsheet simulation of solids processes, Chemie Ingenieur Technik 86 (7)
Manufacturing Symp, J. Pharm. Sci. 104 (2015) 792–802.
(2014) 1073–1079.
[9] W. Marquardt, Dynamic process simulation – Recent progress and future
[40] A. Yang, W. Marquardt, An ontological conceptualization of multiscale models,
challenges, Chem. Process Control CPC-IV, CACHE Publications (1991) 131–
Comput. Chem. Eng. 33 (2009) 822–837.
180.

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015
 M. Dosta et al. / Advanced Powder Technology xxx (xxxx) xxx 7

[41] G.D. Ingram, I.T. Cameron, K.M. Hangos, Classification and analysis of [55] M.L. Thompson, M.A. Kramer, Modeling chemical processes using prior
integrating frameworks in multiscale modeling, Chem. Eng. Sci. 59 (2004) knowledge and neural networks, AIChE J. 40 (1994) 1328–1340.
2171–2187. [56] V. Velàsco-Mejía, V. Vallejo-Becerra, A.U. Chávez-Ramírez, J. Torres-González,
[42] J.A. Gantt, I.T. Cameron, J.D. Litster, E.P. Gatzke, Determination of coalescence Y. Reyes-Vidal, F. Castañeda-Zaldivar, Modeling and optimization of a
kernels for high-shear granulation using DEM simulations, Powder Technol. pharmaceutical crystallization process by using neural networks and genetic
170 (2006) 53–63. algorithms, Powder Technol. 292 (2016) 122–128.
[43] C.R. Wassgren, B. Freireich, J.F. Li, J.D. Litster, Incorporation particle flow [57] P. Kazemi, M.H. Khalid, J. Szlek, A. Mirtič, G.K. Reynolds, R. Jachowicz, A.
information from discrete element simulations in population balance models Mendyk, Computational intelligence modeling of granule size distribution for
of mixer-coaters, Chem. Eng. Sci. 66 (2011) 3592–3604. oscillating milling, Powder Technol. 301 (2016) 1252–1258.
[44] M. Sen, A. Dubey, R. Singh, R. Ramachandran, Mathematical development and [58] D. Barasso, A. Tamrakar, R. Ramachandran, A reduced order PBM-ANN model
comparison of a hybrid PBM-DEM description of a continuous powder mixing of a multi-scale PBM-DEM description of wet granulation process, Chem. Eng.
process, Powder Technol. (2013). Sci. 119 (2014) 319–321.
[45] S. Heinrich, M. Dosta, S. Antonyuk, Multiscale analysis of a coating process in a [59] F. Tao, J. Cheng, Q. Qi, M. Zhang, H. Zhang, F. Sui, Digital twin-driven product
Wurster fluidized bed apparatus, Adv. Chem. Eng. 46 (2015) 83–135. design, manufacturing and service with big data, Int. J. Adv. Manuf. Technol. 94
[46] M. Andrews, P. O’Rourke, The multiphase particle-in-cell (MP-PIC) method for (2018) 3563–3576.
dense particulate flows, Int. J. Multiphas. Flow 22 (1996) 379–402. [60] G. Reynolds, Application of hybrid models for Advanced Process Control of a
[47] D. Groen, J. Knap, P. Neumann, D. Suleimanova, L. Veen, K. Leiter, Mastering the Twin Screw Wet Granulation Process, Presentation to the ADDoPT Digital
scales: a survey on the benefits of multiscale computing software, Phil. Trans. Design Showcase, London, 2019.
R. Soc., A 377 (2019) 20180147. [61] J. van Baten, M. Pons, CAPE-OPEN: Interoperability in industrial flowsheet
[48] L.G. Wang, J.P. Morrissey, K. Hanley, J.Y. Ooi, J.D. Litster, Multiscale modelling simulation software, Chemie Ingenieur Technik 86 (2014) 1052–1064.
of twin screw granulation using a DEM–PBM coupling framework, Proc. 9th Int. [62] S. Fillinger, E. Esche, G. Tolksdorf, W. Welscher, G. Wozny, J.-U. Repke, Data
Granulation Workshop, Lausanne, 2019. exchange for process engineering – Challenges and opportunities, Chemie
[49] L.L. Simon, H. Pataki, G. Marosi, et al., Assessment of recent process analytical Ingenieur Technik 91 (2019) 256–267.
technology (PAT) trends: A multiauthor review, Org. Process Res. Dev. 19 [63] F. Stepánek, Computer-aided product design. Granule dissolution, Chem. Eng.
(2015) 3–62. Res. Design 82 (2017) 1458–1466.
[50] A. Braumann, M. Kraft, P.R. Mort, Parameter estimation in a multidimensional [64] E. Diez, K. Meyer, A. Bück, A. Tsotsas, S. Heinrich, Influence of process
granulation model, Powder Technol. 25 (2010) 196–210. conditions on the product properties in a continuous fluidized bed spray
[51] D. Barasso, S. Oka, A. Muliadi, J.D. Litster, C. Wassgren, R. Ramachandran, granulation process, Chem. Eng. Res. Des. 139 (2018) 104–115.
Population balance model validation and prediction of CQAs for continuous [65] J. Saha, N. Das, J. Kumar, A. Bück, Numerical solution for multidimensional
milling process: toward QbD in pharmaceutical drug product manufacturing, J. fragmentation problems using finite volume methods, Kinet. Relat. Mod. 12
Pharm. Innov. 8 (2013) 147–162. (2019) 79–103.
[52] C. Reimers, J. Werther, G. Gruhn, Flowsheet simulation of solids processes. [66] G. Kotalczyk, F.E. Kruis, Fractional Monte Carlo time steps for the simulation of
Data reconciliation and adjustment of model parameters, Chem. Eng. Process. coagulation for parallelized flowsheet simulation, Chem. Eng. Res. Design 136
47 (2008) 138–158. (2018) 71–82.
[53] J.D. Litster, I.D.L Bogle, Smart process manufacturing for formulated products, [67] J.D. Litster, Design and processing of particulate products, Cambridge
Engineering, in press. https://doi.org/10.1016/j.eng.2019.02.014 University Press, 2016.
[54] M. von Stosch, R. Oliveira, J. Peres, S.F. de Azevedo, Hybrid semi-parametric
modeling in process system engineering: Past, present and future, Comput.
Chem. Eng. 60 (1994) 86–101.

Please cite this article as: M. Dosta, J. D. Litster and S. Heinrich, Flowsheet simulation of solids processes: Current status and future trends, Advanced Pow-
der Technology, https://doi.org/10.1016/j.apt.2019.12.015