Assessment of Integrated Microalgal-Bacterial Ecosystems for Bioenergy Production - Optimization-based Methodology

Lead Research Organisation: Imperial College London
Department Name: Chemical Engineering

Abstract

Many microbial ecosystems, as part of their normal activity, have the potential to provide services to society and improve environmental quality. Some can degrade organic or trace contaminants that pollute water, air or soil. Others can transform waste materials into valuable renewable resources, including bioenergy, biomaterials and high-value products. This generic capability opens the possibility for combining several microbial ecosystems in integrated bioprocesses, where various types of bioenergy or biomaterials are produced and multiple sources of pollution are treated, all at the same time.

The focus in this project is on integrated bioprocesses that couple a microalgae photobioreactor with an anaerobic digester. While microalgae are currently considered one of the most promising feedstocks for biofuels due to their high productivity of carbon-rich lipids, the said combination with anaerobic digestion provides an efficient means of recycling the nutrients present in the waste algal biomass--after lipid extraction. On the whole, integrated microalgal-bacterial ecosystems will be capable of producing bioenergy in the form of biofuel and biogas, while treating both flue gas and wastewater. However, unlike in traditional biorefineries where a spectrum of bio-based products and energy are obtained by processing an available biomass feedstock, growing the feedstock becomes an integral part of the process in an integrated microalgal/bacterial system. Therefore, a photobioreactor can no longer be designed and operated separately from the algal downstream processing. Special attention must also be paid to the microbial adaptation to environmental and operational changes as well as the strong interactions between the various kinds of microorganisms. This intricacy makes it extremely challenging to design and operate these processes solely based on engineering intuition.

It is a principal aim of this project to investigate integrated microalgal-bacterial processes by applying systematic methods of process analysis, design and operation, that are based on mathematical models. Our objective is twofold: (i) make an assessment of integrated microalgal-bacterial systems for sustainable bioenergy production and CO2 capture; and (ii) determine reliable design and operation strategies. An important challenge is the presence of process variability and modeling uncertainty, which challenges the current state-of-the-art of optimization under uncertainty. It is therefore another principal aim of this proposal to develop the crucial methods and tools needed for the analysis and optimization of integrated microalgal-bacterial systems.

While many experimental research and demonstration programs are being carried out in the UK and worldwide to identify the most suitable algae strains and expand algal biofuel production to a major industrial process, this project will be the first of its kind to apply a systematic, model-based optimization methodology, that takes full account of operational issues as well as their interplay with design decisions. It is expected that operational considerations will bring a first element of response regarding critical design decisions, such as the need to operate algae growth and lipid production in separate bioreactors, and whether or not to extract the lipids before the anaerobic digestion step. The ability to identify operational bottlenecks will also provide valuable insight and guidance for strain improvement, e.g. via genetic engineering. Finally, it has been argued that economically sustainable production of microalgae for biofuels may only be achieved if combined with production of bulk chemicals, food, and feed ingredients. While the coproduction of multiple compounds from microalgae remains a challenge, the methodology developed through this project will bring on key insight on the best way to achieve such biorefining.

Planned Impact

It is anticipated that the project will also has a large number of non-academic potential users and beneficiaries. Because the impetus for this research comes from both the exploitation of microbial ecosystems and the optimization technology, the potential impacts reflect this dichotomy.

- The most direct non-academic users of the project results will be companies developing and exploiting microbial ecosystems, such as microalgae culturing and anaerobic digestion processes. More generally, potential applications can be for the production of bioenergy (e.g., hydrogen, methane, ethanol or biodiesel), the treatment of gas and liquid wastes (industrial and/or municipal), or the output of high-value products (e.g., chemical or nutraceutical products). This research therefore has strong ties with the emerging paradigm of biorefineries, defined by the International Energy Agency (IEA) as "the sustainable processing of biomass into a spectrum of bio-based products and bioenergy." In particular, there has been much speculation about the fact that such biorefineries may play a major role in producing chemicals and materials that are traditionally produced from petroleum. Coming from a different angle, biodiesel from microalgae is currently considered one of the most promising alternatives since it has the potential to completely displace petroleum-derived transport fuels without adversely affecting the supply of food and other crop products. It is therefore not surprising that energy giants, such as BP, Shell and ExxonMobil, have recently started large research initiatives on algae-derived biofuel.

- Chemical and pharmaceutical companies will also be a major beneficiary of the optimization technology developments throughout this project. Many Chemical processes are hard to model accurately due to complex and intricate physicochemical phenomena and there is substantial uncertainty concerning resource availability, product prices and demands, plant/process unit availability and reliability, etc. Managing uncertainty efficiently is thus a major incentive in such industry for designing and operating safer and more environmentally benign plants, and at the same time improving product quality and decreasing production costs in an increasingly demand-driven and competitive market. The Industrial Research Consortium of the Centre of Process Systems Engineering (CPSE), which brings together major chemical companies that heavily rely on optimization methods, will serve as a means for dissemination of the developed technology and software. Ultimately, the scope of this research will not be limited merely to chemical industries. It is believed that other industrial sectors--such as energy production, transportation and logistics, aeronautics and aerospace--can also greatly benefit from this technology.

Publications

10 25 50
 
Description Through the research funded on this grant, we have pioneered efficient and fully-validated bounding techniques for parametric/uncertain dynamic systems, which find applications in numerous scientific and engineering fields, e.g., for robust optimization, estimation or control. The developed techniques enjoy high-order convergence properties in order to reduce conservatism, and general conditions under which the computed bounds are stable over infinite time horizons could be derived for the first time.

This improved bounding capability has enabled the rigorous solution of optimization and estimation problems that could not tackled previously. In the field of global optimization, we have been able to optimize dynamic systems with up to 10 decision variables, while previous approaches were limited to a handful of variables only. Another key area is guaranteed parameter estimation, whereby one wants to determine all the parameter values of a model that are consistent with a set of experimental observations/data. Here, we have been able to tackle problems with up to seven uncertain parameters despite the presence of complex dynamics, thereby greatly expanding the scope of applications for these techniques.

A third key contribution through the funded research has been the development of branch-and-lift, the first algorithm of its kind to provide a certificate of global optimality for a wide class of optimal control problems, or more generally infinite-dimensional optimization problems. These algorithmic advances open the perspective of applying global and robust optimization technology to industrially-relevant problems.
Exploitation Route The numerical methods that have been developed during this project for bounding and optimizing undcertain dynamic systems have been implemented in computer programs. All these programs can be downloaded by third-parties, free of charge, from our group website (http://www3.imperial.ac.uk/environmentenergyoptimisation/software). In doing so, our hope is that it will help spread the developed technology both faster and more widely. We are also hopeful that these techniques will be incorporated into commercial process simulators, such as gPROMS or Modelica. Moreover, the Centre for Process Systems Engineering (CPSE) runs an industrial consortium, whose member companies have direct access to the tools and support for using them.
Sectors Chemicals

Digital/Communication/Information Technologies (including Software)

Energy

Environment

 
Description The developments of new optimization methods has led to the creation of prototype software, which are made freely available to the scientific community. There have not been any significant economic or societal impact of the grant to date.
First Year Of Impact 2015
Sector Digital/Communication/Information Technologies (including Software)
 
Description Faculty of Engineering Kick-Start Funds
Amount £10,000 (GBP)
Organisation Imperial College London 
Sector Academic/University
Country United Kingdom
Start 02/2013 
End 03/2013
 
Description Marie Curie Career Integration Grant
Amount € 100,000 (EUR)
Funding ID PCIG09-GA-2011-293953 
Organisation European Research Council (ERC) 
Sector Public
Country Belgium
Start 08/2011 
End 08/2015
 
Description Collaboration of robust optimal control methods and tools 
Organisation ShanghaiTech University
Country China 
Sector Hospitals 
PI Contribution During his post-doctoral appointment at Imperial College, Dr Houska initiated the development of a new approach to global optimal control called branch-and-lift, alongside novel set-valued integration techniques for uncertain dynamic systems. Our on-going collaboration with Dr Houska (now Assistant Professor at ShanghaiTech University) entails the continuation of these developments and their implementations in the form of computer software.
Collaborator Contribution Dr Houska provided regular support to 2 PhD students at Imperial College (Mario E Villanueva and Jai Rajyaguru), which has resulted in several joint publications. Dr Houska is also supervising a PhD student on the topic of "rigorous robust optimal control" at ShanhaiTech University (JC Li), which has led to another joint publication
Impact Papers: - J Rajyaguru, ME Villanueva, B Houska, B Chachuat, Chebyshev model arithmetic for factorable functions, Journal of Global Optimization (DOI: 10.1007/s10898-016-0474-9) - ME Villanueva, R Quirynen, M Diehl, B Chachuat, B Houska, Robust MPC via min-max differential inequalities, Automatica 77: 311-321, 2017 (DOI: 10.1016/j.automatica.2016.11.022) - B Houska, JC Li, B Chachuat, Towards rigorous robust optimal control via generalized high-order moment expansion, Optimal Control Applications & Methods (DOI: 10.1002/oca.2309)
Start Year 2015
 
Description Collaboration on modeling of acclimation in microalgae 
Organisation The National Institute for Research in Computer Science and Control (INRIA)
Department Sophia Antipolis Research Centre
Country France 
Sector Academic/University 
PI Contribution We applied robust estimation techniques to calibrate and validate a new model of microalgae growth that acccounts for photoproduction, photoinhibition and photoacclimation all together (see below)
Collaborator Contribution The said model was developed by INRIA, incollaboration with one of my graduate students (A. Nikoalou) that I seconded for 3 months.
Impact No significant outcome to date.
Start Year 2011
 
Description Collaboration on modeling of microalgae growth under limited light conditions 
Organisation University of Padova
Country Italy 
Sector Academic/University 
PI Contribution I hosted a PhD student from Padova (A. Bernardi) during 8 months (May 2013-Feb 2014), and then sent one of my own PhD students (A. Nikolaou) to Padova for 2 months (June-July 2014). I myself spent 1 month at this university in June 2014 as visiting scientist position. This has led to the submission of a proposal for H2020 call LCE-11-2015. For the first time, a mathematical model capable of quantitative prediction of chlorophyll-a fluorescence (a proxy of photoproduction) and accounting for the mechanisms of photoinhibition and photoregulation in microalgae has been developed and experimentally validated. These findings have been submitted to Journal of Biotechnology.
Collaborator Contribution The lab of Prof. T. Morosinotto in the Department of Biology has state-of-the-art experimental facilities for production and analysis of microalgae. The validation experiments for the aforementioned fluoresence model were conducted in this lab.
Impact This is a multidisciplinary collaboration between biologists (experimental) and systems engineers (modeling). The main outcomes of this collaboration to date have been in the form of publications (2 papers submitted for publication) and conference presentations (2 conference presentations at IFAC World Congress 2014 and AIChE Annual Meeting 2014).
Start Year 2013
 
Title CRONOS (Complete seaRch sOlutions for NOlinear Systems) 
Description CRONOS provides a collection of C++ classes for global optimization and constraint propagation of nonlinear and dynamic systems using complete search. 
Type Of Technology Software 
Year Produced 2016 
Open Source License? Yes  
Impact No significant impact to date 
URL https://bitbucket.org/omegasoftware/cronos
 
Title MC++ 
Description A function is called factorable if it can be formed from a finite recursive composition of unary and binary operations. Typically, the latter are binary sums and binary products, and the former are outer compositions with a univariate intrinsic function, such as 'inv', 'exp', 'log', 'pow', 'sqrt', etc. Nearly every function that can be represented finitely on a computer is factorable. Given a factorable function, MC++ provides methods for computing bounds on that function, in the form of convex/concave relaxations, Taylor models, and spectral bounds. A main objective in developing MC++ has been to make the bounds computation as simple and natural as possible, similar to computing function values in real number arithmetic. In particular, MC++ can be quite useful for the fast prototyping and testing of new algorithms and ideas, for instance in such areas as global and robust optimization. 
Type Of Technology Software 
Year Produced 2013 
Open Source License? Yes  
Impact - Library incorporated into software ACADO Toolkit (http://acado.github.io/) - Library used in CRONOS (https://bitbucket.org/omegasoftware/cronos) 
URL https://bitbucket.org/omegasoftware/mcpp