Analysis of the cyanobacterial hydrogen photoproduction process via model identification and process simulation

Abstract

Dynamic simulation models for cyanobacterial hydrogen production process. Parameter estimation via dynamic optimisation. Proposed modified models exhibit higher accuracy for real process simulation. Interpretation of higher performance of CSTR over PFR for this process. Fed-batch processes are proposed as the optimal reactor operation. a b s t r a c t Cyanothece sp. ATCC 51142 is considered a microorganism with the potential to generate sustainable hydrogen in the future. However, few kinetic models are capable of simulating different phases of Cyanothece sp. ATCC 51142 from growth to hydrogen production. In the present study four models are constructed to simulate Cyanothece sp. batch photoproduction process. A dynamic optimisation method is used to determine parameters in the models. It is found that although the piecewise models fit experimental data better, large deviation can be induced when they are used to simulate a process whose operating conditions are different from the current experiments. The modified models are eventually selected in the present study to simulate a two-stage continuous photoproduction process. The current simulation results show that a plug flow reactor (PFR) shows worse performance compared to a continuous stirred-tank reactor (CSTR) in the current operating conditions since it lowers the total hydrogen production. The finding is that nitrate and oxygen concentration change along the direction of culture movement in PFR, and hydrogen is only generated in the zone where both of them are low. The reactor area thereby is not well utilised. Additionally, as hydrogen production rate is primarily influenced by biomass concentration, which increases initially and decreases eventually along the direction of culture movement, the overall hydrogen production rate in a PFR may be lower than that in a CSTR. Finally, in this study fed-batch photoproduction processes are proposed containing only one photobioreactor based on the current simulation.

An efficient model construction strategy to simulate microalgal lutein photo‐production dynamic process

“…Moreover, to thoroughly explore the feasibility of applying ANNs into an offline optimal control framework where the entire process behavior of an unknown experiment is predicted before its implementation (del Rio-Chanona et al, 2015a), the current ANNs are used to simulate the processes Test 1 and Test 2. It follows the procedure that a single initial experimental point is supplied to the network, then the ANN computes the next state.…”

“…Hence, executing process optimization within a control scheme can remarkably improve the process profitability (del Rio‐Chanona et al, ). In order to resolve this challenge, it is essential to initially construct a robust mathematical model, which is able to accurately simulate and predict the dynamic performance of the long‐term photo‐production system, for both microalgae growth and lutein production throughout the entire process (Zhang et al, ).…”

Dynamic modelling of high biomass density cultivation and biohydrogen production in different scales of flat plate photobioreactors

“…A previous study (Zhang et al, ) presents a dynamic model for the simulation of Cyanothece sp. photo‐autotrophic growth under nitrogen and carbon (CO 2 ) sufficiency.…”

Reinforcement learning for batch bioprocess optimization

“…Due to different growth phases, nonmooth behaviour is observed for the physical system. To accommodate this difficulty, switching functions have been proposed [15,57]. In this work the nonsmooth behaviour is modelled using a sign(·) function.…”