Future Foods: Coding Caviar To Avoid Species Extinction

Ed
g d or he e. lis rs em b e pu e ich is ngin y@ k ch E se .u hi al res .co , w mic : jc dia e e tc he ns in f C ptio inem o ri nl d re n c i ea tio bs ma m pp stitu , Su s@ co a n y. st e I .org hen fir th da e tep le to m tic by che el.s tce ar hly i . ig is t w t@ n w Th on ket ing: m uc is w rt d : a ve al Ad ri ito
the chemical engineer|issue 824|february 2010
future foods
coding caviar to avoid species extinction
CHEMICAL ENGINEERS WHO CHANGED THE WORLD | PROCESS SIMULATION | ARSENIC REMOVAL
PROCESS SIMULATION &CAREERS OPTIMISATION
tce
Mark Matzopoulos explains new physical property modelling techniques for strongly associating systems and polymers
hysical properties are at the heart of virtually every calculation performed by practising chemical engineers. The size of vessels, operating conditions, temperatures, energy requirements, safety limits and many other critical design and operating quantities all depend on the intrinsic properties vapour pressure, density and phase partitioning, for example of the fluids being processed. The last 20 years have seen significant research into advanced molecular thermodynamic methods aimed at improving the representation of pure component and mixture properties, making it easier to determine properties for new components and apply these within process modelling tools. Much of this effort centres on statistical associating fluid theory (SAFT). The SAFT approach allows many different thermodynamic properties of mixtures to be determined accurately based on models of molecules and their interactions. These are much more sophisticated than those used by more well-known techniques such as the cubic equations of state. The resulting fluid model can then be used to accurately predict a wide range of properties over a range of operating conditions. SAFT is particularly good for stronglyassociating systems, such as the azeotropic hydrogen fluoride-water mixture used in the production of refrigerants, aqueous solutions of non-ionic surfactants and even strong electrolyte solutions. It also accurately quantifies the behaviour of systems involving high molecular weight
No its not a caterpillar! The SAFT view of a nonionic surfactant made of five segments; smaller spheres indicate electron donor and acceptor association sites
thermodynamic methods is that because the underlying molecular model has a much more physically meaningful basis, SAFTs predictions can be extrapolated much more widely a particular advantage for systems where there is limited experimental data in components, such as polymer-gas systems the areas of operation being studied. all typically poorly represented by Also, traditional equations of state cannot conventional equations adequately represent the behaviour of of state. systems involving complex materials such as The development of SAFT has been given polar solvents, hydrogen bonded fluids and a strong focus at Imperial College London, polymers. SAFTs associating fluid approach where a team of around 25 are working which can account for the electrostatic, polar on molecular systems engineering (MSE) and other association forces that need to be funded by a 3.6m ($5.8m) EPSRC grant. considered provides an unprecedented The key focus of Imperials SAFT group, led capability for modelling such systems by George Jackson, Amparo Galindo accurately. and Claire Adjiman, is in One challenge that needs providing thermodynamic to be overcome is that of modelling tools for computational efficiency. industry. In typical process simulations, physical background properties are At its heart SAFT responsible for about is an equation of 70% of computational state, a term likely to time. The complexity send shivers down of the underlying the spine of any SAFT equations, chemical engineer which require solution who endured a course of a system of nonin thermodynamics. linear equations and SAFT embodies evaluation of its partial significant advances derivatives for every Water molecule showing over the traditional van property calculation, brings association sites der Waals equation or cubic additional demands as well engineering equations such as as potential robustness issues. Peng-Robinson, which are better suited to These are important concerns since a typical near-spherical molecules. simulation will involve thousands of calls to The main difference is that the SAFT the physical property package. equation is underpinned by a physically how it works realistic representation of the molecule that includes its shape, size and the interactions In SAFT, molecules are modelled as chains between molecules for example, hydrogen formed from spherical segments, and its bonding within a mixture. possible to add so-called association sites on the segments to treat as hydrogen bonds or The major advantage over other
odel M viour beha

february 2010 www.tcetoday.com 43
tce
PROCESS SIMULATION & OPTIMISATION

650 strongly polar interactions. The basic SAFT approach starts by describing ideal molecular behaviour in terms of the usual translational, 550 rotational and vibrational kinetic energy terms. The various non-ideal terms are T/K then introduced separately: terms for the 450 forces of attraction and repulsion between segments making up the molecules, chain bonding representing the contribution due to the formation of chains, and forces of 350 association for example, those resulting 0 0.2 0.4 0.6 WPEG from hydrogen bonds in aqueous mixtures. There are various implementations of Above: Closed loop regions of immiscibility the underlying SAFT model, which differ of aqueous solutions of PEG of molecular principally in terms of the mathematical weight from 2180 g/mol (smallest loop) to 1,020,000 g/mol (largest loop) representation of the molecule and the specific forms of the various contributions. For mixtures, predictions based on data The approach taken by the Imperial team fitted in one area of operation remain is known as SAFT-VR, where VR stands for valid over a wide range of operating states. variable range . For example, liquid-liquid equilibrium Key to the behaviour of molecules that behaviour can be predicted accurately form hydrogen bonds, such as water or the using parameters fitted solely against amine solvent MEA (monoethanolamine), is vapour-liquid data, decreasing the need for the definition of association sites electron complicated and often-unreliable laboratory donors or electron acceptors which are LLE measurements. used to quantify hydrogen bond formation This is the closest that we have come to between molecules. a truly universal tool for predicting pure Its these association sites that component and mixture properties , provide SAFTs unique says Jackson. What is more it capability in dealing predicts phase equilibrium with aqueous solutions, properties within a few percent surfactants and other for a vast array of complex strongly-associating systems. components, allowing engineers to design integrated and optimise approach processes that have Not only does SAFT SAFT ethanol never been able to simplify standard engineering benefit from the application design problems, it has opened up new of modern model-based approaches that could revolutionise the engineering techniques before. design of industrial processes. Because Parameters defining all the relevant properties models can be closely integrated quantities that define the model for a with process models its possible to given molecule for example, the number incorporate molecular-level decisions and diameters of spherical segments, the (including molecular architecture, number and type of association sites are microstructure, solvent selection, mixture estimated from experimental data. Typically formulation and so on) within the process this data includes saturated vapour pressure design procedure and take decisions and density measurements at different simultaneously rather than sequentially. temperatures. Where available, mixture data A new approach is to design molecules is also used. One of the Imperial teams main and processes hand-in-hand, so that the activities is to build the growing databank of whole system is optimal a significant pure component and mixture parameters. challenge because of the complexity of the other key advantages physical description and the wide range of length-time scales involved, says Adjiman. There are other advantages. For example, its possible to transfer parameters from The integrated approach is exemplified one molecule to another within the same by Imperials work on the development of class (for example, from n-alkanes to efficient and cost-effective solutions for CO2 polyethylene), thus reducing or eliminating capture a key component of most practical the need for measurement although, carbon abatement systems through says Galindo, good experimental data simultaneous optimisation of process and always helps. solvent design. Traditionally, decisions on
case study
The Imperial team worked with ICI and Ineos Fluor to apply SAFT-VR techniques to chemical processes for refrigerants such as the hydrofluorocarbons (HFCs) being used as replacements for the CFC refrigerants responsible for the depletion of ozone in the upper atmosphere. HFC production involves mixtures containing hydrogen fluoride (HF) which, because of the extremely strong association between HF and water molecules, are particularly difficult to characterise with standard thermodynamic methods. SAFT-VR provides an excellent description of the vapour-liquid and liquid-liquid equilibria in these systems, which was used to design and optimise operation of the HFwater separation process. At the other end of the molecular scale, SAFT-VR also provided an excellent description of the phase equilibria of aqueous solutions of hydrocarbons and micellar solutions of alkyl polyoxyethylene surfactants and polyethylene glycol (PEG) in work with BP Exploration, Schlumberger and ICI. These systems are key ingredients in numerous applications ranging from personal care formulations to enhanced oil recovery that can extend well lifetime by a factor of two or three. In a recent collaboration with the Borealis Group, SAFT-VR was used to analyse phase behaviour (adsorption) in gas phase polymerisation reactions. The analysis suggested that replacement of the inert nitrogen gas by less volatile n-pentane would significantly increase the catalytic activity of the reaction. Borealis has subsequently verified at bench scale a 30% increase in the yield of polythene compared to the conventional mixture, with no change required in the reactor conditions.
44
www.tcetoday.com
SAFT has opened up new approaches that could revolutionise the design of industrial processes
february 2010
PROCESS SIMULATION &CAREERS OPTIMISATION

solvents are taken first then the process is designed around the constraints posed by the solvent choice, potentially resulting in seriously sub-optimal performance. new approach known as SAFT-, a group contribution method that embodies the concepts of the original UNIFAC approach developed by Fredenslund, Prausnitz and co-workers in the 1970s and 80s, but applying the molecular model principles of SAFT. SAFT-is immensely powerful in terms of property prediction; once the basic functional groups defining a molecule have been characterised, pure component and mixture behaviour can be predicted very accurately with no need for experimental data-fitting whatsoever. Its early days for SAFT- says Jackson, , but initial results are very promising. We can predict mixture properties over a wide range of conditions without any data. This will significantly reduce the time and effort required to ensure accurate physical property calculations in a wide range of industrial process design projects in the future. The MSE group has just won a 150,000 Research Excellence award from Imperial that will be used to accelerate technology transfer to industry. tce Mark Matzopoulos (m.matzopoulos@ psenterprise.com) is coo of London-based process modelling software supplier PSE
tce
commercialisation
Imperial has licensed the software to Process Systems Enterprise (PSE), which has very close links with the chemical engineering community at Imperial (the company was an early spin-out of the college), and has a significant chemical industry user base and an already-large number of academic users of its software . SAFT-VR is already interfaced with PSEs gPROMS suite, and the company is offering free trials of the capability to selected gPROMS industrial users in the chemical and polymers sectors.
This is a tool that will significantly reduce the time and effort required to ensure accurate physical property calculations
reducing reliance on experimental data

Work at Imperial continues in a number of promising directions. Most of the research is related to refining the underlying molecular model, to make it as easy as possible to predict properties for existing and new molecules, and to make those properties as accurate as possible. One major area of research relates to a
ADVANCED CUSTOM MODELLING IN A FLOWSHEETING ENVIRONMENT

PSEs gPROMS provides unprecedented power for high-delity predictive modelling for any complex process or product, within a sophisticated owsheeting environment. And our expert ModelCare service helps you maximise return, whatever youre modelling.
Accelerate innovation Manage risk

Operations in UK, USA, Japan, Korea, Saudi Arabia and India. +44 20 8563 0888 www.psenterprise.com
february 2010
www.tcetoday.com
45

Future Foods: Coding Caviar To Avoid Species Extinction

Uploaded by

Copyright:

Available Formats

Future Foods: Coding Caviar To Avoid Species Extinction

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Future Foods: Coding Caviar To Avoid Species Extinction

Uploaded by

Copyright:

Available Formats

Ed

the chemical engineer|issue 824|february 2010

PROCESS SIMULATION &CAREERS OPTIMISATION

odel M viour beha

PROCESS SIMULATION & OPTIMISATION

PROCESS SIMULATION &CAREERS OPTIMISATION

reducing reliance on experimental data

ADVANCED CUSTOM MODELLING IN A FLOWSHEETING ENVIRONMENT

You might also like