Linear time invariant model of anaerobic digestion of waste activated sludge

Transcription

1 Engineering Conferences International ECI Digital Archives Wastewater and Biosolids Treatment and Reuse: Bridging Modeling and Experimental Studies Proceedings Spring Linear time invariant model of anaerobic digestion of waste activated sludge Antonios Sendjaja Nanyang Technological University, Singapore Follow this and additional works at: Part of the Environmental Engineering Commons Recommended Citation Antonios Sendjaja, "Linear time invariant model of anaerobic digestion of waste activated sludge" in "Wastewater and Biosolids Treatment and Reuse: Bridging Modeling and Experimental Studies", Dr. Domenico Santoro, Trojan Technologies and Western University Eds, ECI Symposium Series, (2014). This Conference Proceeding is brought to you for free and open access by the Proceedings at ECI Digital Archives. It has been accepted for inclusion in Wastewater and Biosolids Treatment and Reuse: Bridging Modeling and Experimental Studies by an authorized administrator of ECI Digital Archives. For more information, please contact

2 Linear time invariant model of anaerobic digestion of waste activated sludge Speaker: Dr. Antonius Yudi Sendjaja AEBC NEWRI, Singapore IWA Conference on Wastewater and Biosolids Treatment and Use Otranto, Italy:

3 NTU-NEWRI Advanced Environmental Technology Centre 2/12

4 Activated Sludge Plant tation-tools/wastewatertreatment/hardware/semi-centralisedwastewater-treatments-3 Anaerobic digestion: Waste to energy Mathematical model for analysis LTI model: Stability analysis Parameter sensitivity Kinetic estimation Controller design 3/12

5 Anaerobic Digestion No. 1 Particulate matters Particulate inert Soluble inert Carbohydrates Proteins Lipids Sugars Amino acids Fatty acids Propionate Butyrate Valerate Acetate Hydrogen Methane 4/12

6 Simpler version for WAS Activated sludge Acidogens Methanogens Particulate Substrate VFAs Methane ADM1 Simpler version 24 soluble + 6 ions + 3 gases 7 major components 35 ODEs 7 ODEs 25 kinetic constants 8 kinetic constants 5/12

7 Linear Time Invariant Linear: Input n Output n Time invariant: Same model applies, now and then Mass balance x = Input Output + Generation - Consumption Differential equation Nonlinearity: Biomass growth, input - output Linearization via Taylor Series LTI model Analysis 6/12

8 Linearization Biomass growth rate Input and output 7/12

9 Model Validation A multivariable 6 6 model of anaerobic digestion of activated sludge Verification against original model 8/12

10 Kinetic Constant Estimation Open/closed loop step/impulse test for model identification Hydrolysis and biomass decay: 1 st order Kinetic parameter is straightforward Biomass growth: Monod Need steady state concentration + least square 9/12

11 System Behavior Recommended control structure Independent individual controller preferable Substrate through biomass, biomass via substrate Stability Original process is stable Imbalance in acidogens growth and decay rate causes instability Sensitivity Biomass is more sensitive to influent condition compared to substrate Methanogens is the most sensitive parameter 10/12

12 Acknowledgement Energy Market Authority, Singapore Sembcorp industries, Ltd., Singapore Dr. Jerry Liu, Dr. Shaun Pan, Mr. Prannoy Chowdhury NEWRI Team, NTU, Singapore A/Prof. Tan Soon Keat, Mr. Teng Wang, Mr. Harish Venkatakrishnan, Mr. Dongzhe Li 11/12

13 Linear time invariant model of anaerobic digestion of waste activated sludge THE POWER OF WATER The Flow of Hope Antonius Yudi Sendjaja, Youming Tan, Santosh Pathak, Yan Zhou, Maszenan bin Abdul Majid, Wun Jern Ng Thank You

14 Industrial Process Control Biogas OLR 1 2 Which one is better? HOWEVER Model based control requires good mathematical model

15 Controller design Feedback Influent COD increase Adjust OLR Effluent ok! Feedforward Temperature decrease Adjust OLR Effluent ok! Mathematical model: 1.Black box 2.White box 3.Grey box