BIOSYSTEMS CHARACTERIZATION, MODELLING AND OPTIMIZATION

There is a growing worldwide demand to move to renewable and clean energy, in this case biofuel. This interdisciplinary project aims to support decision making on an intermediate aspect of biofuel production from Nannochloris and Ooscystis algae: lipid extraction. It is intended to meet this aim through the coordinated use of statistical experimental design and mathematical optimization. Lipid extraction in this project is seen as a two-phase experimental process: (i) algal cell growth and (ii) lipid extraction. In the first phase, it is necessary to minimize the cell growth time while maximizing the number of viable cells. In the second stage, the amount of lipid extracted per cell will be maximized as a proxy to efficiency. These three aspects –time to cell growth, number of viable cells, and efficiency- will be mapped to the experiment parameters on each stage including formulation, setup and processing conditions.The parameters proposed are light, sucrose and salinity, which are the most commonly controlled in these type of experiments. As indicated previously, this project will involve techniques from two different disciplines: statistical design of experiments and mathematical optimization. Statistical design of experiments will be used to create the functions connecting the purposeful variation of parameters with the resulting variation of the three responses –time to cell growth, number of viable cells, and efficiency- through a central composite design. Then moving on to mathematical optimization, the first measurement is absorbance, which is a direct measurement of amount of cells. The characterization of the absorbance slope is a curve that its initial steepness is the time to grow cells and later moving on to a steady state plateau that represents the maximum amount of cells. For the characterization ratio of lipid fluorescence with cells data that also has a curve behavior the main goal is to obtain the steady state plateau that indicates de maximum amount of lipids in each cell. It is proposed to first take a step back from this direct measure and focus con the derivative of the functions of the characterization, this will be linear functions: absorbance slope, absorbance and the ratio of lipid and the ratio. With the empirical functions available, a mathematical optimization problem will be built to ultimately drive up the production of lipids. This is an ongoing project that is on the verge of entering to the experimental phase.