Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy
This work presents a nonlinear model predictive control scheme with a novel structure of observers aiming to create a methodology that allows feasible implementations in industrial anaerobic reactors. In this way, a new step-by-step procedure scheme has been proposed and tested by solving two specif...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/15/22/8594 |
| _version_ | 1797465444762779648 |
|---|---|
| author | Luis G. Cortés J. Barbancho D. F. Larios J. D. Marin-Batista A. F. Mohedano C. Portilla M. A. de la Rubia |
| author_facet | Luis G. Cortés J. Barbancho D. F. Larios J. D. Marin-Batista A. F. Mohedano C. Portilla M. A. de la Rubia |
| author_sort | Luis G. Cortés |
| collection | DOAJ |
| description | This work presents a nonlinear model predictive control scheme with a novel structure of observers aiming to create a methodology that allows feasible implementations in industrial anaerobic reactors. In this way, a new step-by-step procedure scheme has been proposed and tested by solving two specific drawbacks reported in the literature responsible for the inefficiencies of those systems in real environments. Firstly, the implementation of control structures based on modeling depends on microorganisms’ concentration measurements; the technology that achieves this is not cost-effective nor viable. Secondly, the reaction rates cannot be considered static because, in the extended anaerobic digestion model (EAM2), the large fluctuation of parameters is unavoidable. To face these two drawbacks, the concentration of acidogens and methanogens, and the values of the two reaction rates considered have been estimated by a structure of two observers using data collected by sensors. After 90 days of operation, the error in convergence was lower than 5% for both observers. Four model predictive controller (MPC) configurations are used to test all the previous information trying to maximize the volume of methane and demonstrate a satisfactory operation in a wide range of scenarios. The results demonstrate an increase in efficiency, ranging from 17.4% to 24.4%, using as a reference an open loop configuration. Finally, the operational robustness of the MPC is compared with simulations performed by traditional alternatives used in industry, the proportional-integral-derivative (PID) controllers, where some simple operational scenarios to manage for an MPC are longer sufficient to disrupt a normal operation in a PID controller. For this controller, the simulation shows an error close to the 100% of the reference value. |
| first_indexed | 2024-03-09T18:21:34Z |
| format | Article |
| id | doaj.art-7652bd83769240cab0bedbe73fc4e2f7 |
| institution | Directory Open Access Journal |
| issn | 1996-1073 |
| language | English |
| last_indexed | 2024-03-09T18:21:34Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj.art-7652bd83769240cab0bedbe73fc4e2f72023-11-24T08:15:32ZengMDPI AGEnergies1996-10732022-11-011522859410.3390/en15228594Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification StrategyLuis G. Cortés0J. Barbancho1D. F. Larios2J. D. Marin-Batista3A. F. Mohedano4C. Portilla5M. A. de la Rubia6Departamento de Tecnología Electrónica, Escuela Politécnica, Universidad de Sevilla, 41011 Seville, SpainDepartamento de Tecnología Electrónica, Escuela Politécnica, Universidad de Sevilla, 41011 Seville, SpainDepartamento de Tecnología Electrónica, Escuela Politécnica, Universidad de Sevilla, 41011 Seville, SpainEfuels Technologies Ltd., 42-44 Bishopgate, London EC2N 4AH, UKDepartamento de Ingeniería Química, Campus de Cantoblanco, Universidad Autonoma de Madrid, 28049 Madrid, SpainFacultad de Minas, Universidad Nacional de Colombia, Robledo, Medellín 050034, ColombiaDepartamento de Ingeniería Química, Campus de Cantoblanco, Universidad Autonoma de Madrid, 28049 Madrid, SpainThis work presents a nonlinear model predictive control scheme with a novel structure of observers aiming to create a methodology that allows feasible implementations in industrial anaerobic reactors. In this way, a new step-by-step procedure scheme has been proposed and tested by solving two specific drawbacks reported in the literature responsible for the inefficiencies of those systems in real environments. Firstly, the implementation of control structures based on modeling depends on microorganisms’ concentration measurements; the technology that achieves this is not cost-effective nor viable. Secondly, the reaction rates cannot be considered static because, in the extended anaerobic digestion model (EAM2), the large fluctuation of parameters is unavoidable. To face these two drawbacks, the concentration of acidogens and methanogens, and the values of the two reaction rates considered have been estimated by a structure of two observers using data collected by sensors. After 90 days of operation, the error in convergence was lower than 5% for both observers. Four model predictive controller (MPC) configurations are used to test all the previous information trying to maximize the volume of methane and demonstrate a satisfactory operation in a wide range of scenarios. The results demonstrate an increase in efficiency, ranging from 17.4% to 24.4%, using as a reference an open loop configuration. Finally, the operational robustness of the MPC is compared with simulations performed by traditional alternatives used in industry, the proportional-integral-derivative (PID) controllers, where some simple operational scenarios to manage for an MPC are longer sufficient to disrupt a normal operation in a PID controller. For this controller, the simulation shows an error close to the 100% of the reference value.https://www.mdpi.com/1996-1073/15/22/8594anaerobic digestionasymptotic observerhomogeneous reaction systemskinetic parameter observermodel predictive controlstep-ahead |
| spellingShingle | Luis G. Cortés J. Barbancho D. F. Larios J. D. Marin-Batista A. F. Mohedano C. Portilla M. A. de la Rubia Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy Energies anaerobic digestion asymptotic observer homogeneous reaction systems kinetic parameter observer model predictive control step-ahead |
| title | Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy |
| title_full | Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy |
| title_fullStr | Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy |
| title_full_unstemmed | Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy |
| title_short | Full-Scale Digesters: Model Predictive Control with Online Kinetic Parameter Identification Strategy |
| title_sort | full scale digesters model predictive control with online kinetic parameter identification strategy |
| topic | anaerobic digestion asymptotic observer homogeneous reaction systems kinetic parameter observer model predictive control step-ahead |
| url | https://www.mdpi.com/1996-1073/15/22/8594 |
| work_keys_str_mv | AT luisgcortes fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT jbarbancho fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT dflarios fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT jdmarinbatista fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT afmohedano fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT cportilla fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy AT madelarubia fullscaledigestersmodelpredictivecontrolwithonlinekineticparameteridentificationstrategy |