gPROMS advanced process modelling platform provides powerful abilities for applying models to real applications. The platform encompasses many enabling features including:
The powerful equation-oriented numerical solvers underpinning the gPROMS platform enable not just steady-state but dynamic modelling. This is key for capturing the true dynamics of a food manufacturing process such as fouling in membrane ultrafiltration.
A key feature of the gPROMS platform is the ability to create high-fidelity custom models of processes to capture corporate knowledge and create competitive advantage.
Unlike traditional flowsheeting packages, you can easily create and maintain custom models using all the power of the gPROMS platform\'s advanced process modelling language – with no need for programming. Models can then be published in libraries for other users across the organisation.
An important part of using advanced modelling approaches is to validate the model being used. There are, broadly, two parts to this process; the first is to fit empirical model parameters using experimental or pilot plant data and the second is to confirm the predictive nature of the model. gPROMS has a dedicated model validation ability that enables the user to easily use their data to validate the model(s) of the application.
Optimisation is a key technology for process organisations to create value and competitive advantage in process design and operations.
In particular, large-scale optimisation based on high-fidelity models has the ability to create significant value from 'already-optimised' processes.
gPROMS family products take full advantage of the optimisation capabilities in the gPROMS platform, which make it possible to apply rigorous optimisation to the design and operation of individual unit operations, plant sections, entire plants or manufacturing lines, or even multi-site applications.
The main use of process simulation and modelling tools is to analyse "what-if" scenarios in order to improve process design and operation. Currently this is done manually as a point activity, using repeated simulation runs.
Global systems analysis (GSA) allows the comprehensive exploration of the behaviour of a system over domains of any user-selected subset of its input variables ('factors'), and output variables ('responses').
This provides a quick, easy and systematic way to explore the complex process design and operational decision space using high-fidelity models.
All the above features can be achieved via drag-and-drop flowsheeting model libraries to create a flowsheet model of the process in question. This flowsheet can then be used in all the above activities. This gives access to powerful modelling techniques to users who maybe do not have experience in advanced process modelling.
Solutions can also be deployed outside of the gPROMS modelling environment through a company intranet for instance. There are many options for building non-expert interfaces including the ability to build Excel based applications.
gPROMS FormulatedProducts is PSE’s modelling library built specifically to tackle the needs of the formulated products industries including the pharmaceutics, fast moving consumer goods and food industries. Advanced phase description and properties modelling allow for high-fidelity modelling of manufacturing lines for synthesis, crystallization, product manufacture and product performance.
gPROMS food processing model library (a module of gPROMS FormulatedProducts) combines the strength of the gPROMS platform, gPROMS FormulatedProducts and process-product interaction models developed by NIZO and PSE. The concepts behind models available are explained in more detail below:
Heat treatment of food products is basically a trade-off between product safety and heat induced detrimental effects.
The heat treatment model library therefore not only comprises basic mass and energy balances, but also models for heat induced effects such as inactivation of micro-organisms, enzyme inactivation, protein denaturation, protein-mineral fouling, bio-fouling, changes in viscosity and taste.
An array of building blocks (CSTR, isothermal tube, tubular heat exchanger, plate heat exchanger, flash reactor, steam injector and steam infuser) is available for in-silico implementation of heat treatment processes. These process models can subsequently be coupled with the process-product interaction models mentioned above that are relevant for the product-process combination under consideration. In case physical (protein-mineral) fouling plays a role the effect of the fouling process on heat transfer during the process is also taken into account.
The model library has been validated and used extensively in industrial settings e.g. to optimise runtime by minimizing fouling, to minimize heat induced detrimental effects such as protein denaturation or off-flavour formation or to re-design or design novel heat treatment processes.
The mass balance equations (for total mass and per component) and the constitutive relations form the basis of the membrane separation model library.
Relations for membrane flux and retention factors complete the model.
Membrane resistance, which determines the flux, is assumed to consist of a static (empirical) and a dynamic part.
The dynamic part enables the user to incorporate fouling behaviour of the membranes in the model.
The membrane model predicts permeate and retentate compositions as a function of time. It is suitable for single and multi-stage membrane separation systems and can be used for continuous and batch processes. The model also has built-in scheduling options to include operational changes during production and cleaning procedures in the model calculations. The combined key features mentioned above enable customers to use the model both for designing membrane processes and for optimising operational and scheduling procedures.
The falling film evaporator model library contains all required building blocks for evaporators using thermal vapour recompression (TVR) and/or mechanical vapour recompression (MVR). Besides the basic mass and energy balances the model library also comprises a wide range of process-product interaction related aspects such as product temperatures, boiling point elevation, (minimum) wetting rates, product viscosity, vapour flows and heat transfer coefficients. Coupling with heat treatment models, e.g. to predict heat induced effects on specific components or fouling, is also possible. The dynamic implementation of the model makes it suitable for predicting and controlling the behaviour of the evaporator in transient situations such as start-up. The model library has frequently been used in industrial settings for various purposes such as (re)designing evaporators, trouble shooting and process optimisation.
The cheese model library in covers all process steps in the production process of semi-hard cheese types such as Gouda: cheese milk standardisation, thermisation and pasteurisation, renneting, curd preparation, brining and ripening.
The cheese model library consists of a set of empirical and mechanistic models.
Examples of mechanistic models are the salt/moisture diffusion models the protein breakdown (proteolysis) model.
Models for other parameters such as acidification and taste attributes are empirical in nature.
All models have been developed and validated using an extensive amount of industrial scale cheese production and product analyses data.
The cheese model library has been used in industrial cheese production settings to optimize cheese yield and improve cheese quality. Another application is minimizing variation in cheese moisture content. This enables producers to increase the average moisture content of the cheese and thereby increase profit.
The spray drying model library can be used for single-stage dryers and multi-stage dryers with internal and/or external fluid beds.
Additional models for cyclones and bag filter units etc. are available to complete the flowsheets.
Several modelling approaches, at different fidelity levels (described here), are available for the droplet drying process itself. For spray drying efficiency and powder quality, however, not only powder moisture, but also several other parameters such as stickiness and a range powder quality attributes such as bulk density, solubility etc., are relevant. The spray drying library therefore also includes a range of (semi-)empirical and mechanistic models for stickiness and a range of powder properties such as particle density, bulk density and insolubility index. Model parameters in these models are mostly formulation dependent and can, in case they are not already available, be estimated or determined using existing data or by performing lab, pilot and/or industrial scale experiments.
The spray drying model is used frequently in industrial environments for trouble shooting and optimisation purposes. For example, when sticky products are dried using ambient air, the maximum capacity of the dryer depends strongly on variations in ambient air humidity. The spray drying model library, which includes particle drying and stickiness models, makes real time optimisation with respect to drying capacity and product quality possible. Depending on the case, this can result in 10 to 20% capacity increase, while improving powder quality at the same time.
This is a lumped type mass and energy balance model with a psychrometric basis. By combining the model with the appropriate sorption isotherm for the product under consideration (that can be determined experimentally) the model can predict equilibrium moisture content.
CFD simulations can be coupled with a compartmental spray drying model to capture the hydrodynamic and particle trajectories more accurately. A more detailed particle evolution prediction can then be made.
For powder quality properties the average moisture content of the powder is just as important as the equilibrium moisture content. Because the internal moisture diffusion time scale exceeds the particle residence time inside the dryer, the average moisture content of spray dried particles is higher than the equilibrium moisture content at outlet conditions. The near-equilibrium model, which is an extension of the psychrometric model described above, has a built-in feature to calibrate the difference between the average moisture content and the equilibrium moisture content, using readily available process and product data. The calibration approach has been extensively validated by NIZO using large scale production tests. Besides moisture calibration the model also has an option to calibrate heat loss.
Kinetic drying models
As an alternative to the near-equilibrium approach, it is also possible to use kinetic drying model options. In this case the internal moisture diffusion and external moisture transport from the particles to the surrounding air during the drying process is explicitly taken into account. The initial droplet size distribution and residence time (distribution) of the particles are required as input values for the model. Besides that, the formulation specific moisture diffusion coefficient as a function of moisture content and temperature has to be known or determined experimentally.