Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Anna Mikola TkT D Sc (Tech)
Modelling and Control of Water andwastewater treatment processes
WAT - E2130
Lecture outline
Course team introductionParticipants’ introductionLearning outcomesContent and assessment of thecourse- Lectures & exercises- Modelling project work- Excursion- Exams
Introduction lecture:Introduction to modellingand simulation
Hands-on SUMOFirst introduction exercise
Lecturer Anna Mikola• Working experience:
• 3 years at Nopon Oy• Researcher at HUT/Aalto• 18 years with a consultant
(Kiuru&Rautiainen Oy,Ramboll Finland)
• Post-doctoral researcher atAalto 2013-2018
• Lived 5 years in Berlin, 4children
• Visiting researcher in INSAToulouse in 2017
• Professor of Practice since2018
14.4.20203
• M.Sc. From HUT Water lab1999
• Exchange year in France atENCR 1994-1995
• D. Sc. (Tech.) Spring 2013Dissertation: The effect offlow equalizationa andprefermentation on BNR
Lecturer for the control partMichela Mulas
Professor at the FederalUniversity of Ceará,Department of TeleinformaticsEngineering in Fortaleza, Brazil
14.4.2020Laitoksen nimi
4
The course teamCourse assistant: Sara Saukkonen
Lab staff (for IT issues): Heikki Särkkä andAntti Louhio
Guest lecturers:Dr. Ewa Zaborowska GTUKristian Sahlstedt PöyryTeemu Koskinen RambollHenri Haimi FCGPasi Puranen Hyxo
4/14/2020Laitoksen nimi
5
Participants’ introduction1 minute each containing e.g.- Background?- Experience with process
modeling and control- Expectations for this
course?
14.4.20206
Learning outcomesUpon completion, the student should be able to:• Understand the overall process train as well as selected process performance, the
definition the main disturbances for the process operation and the identification of theprocess dynamics
• Be able to describe the modelling and control techniques: state-of-the-art models,basic controllers and their practical application to full scale processes
• Recognise the instrumentation available in the plants: actuators, on-linesensors/analyzers, structure of the automation system and their representation on thepiping and instrumentation diagram
• Optimise plant operation in terms of resources consumption and effluent qualityimprovement
• Analyse and understand the on-line and off-line data available at the treatment plants• Design the automation system for the treatment plants by means of simulator software
• Understand the role of a process modeler or control system designer in the treatmentprocess construction or operation teamId
entit
y
Skill
Kno
wle
dge
Why process modelling and controlcourse?• Models are important tools in process development, design,
optimization and operation.• Modelling allows to understand more complex systems and
process trains• Process operation and control are crucial parts of well
performing processes – a well-designed plant can be verybadly operated!!
• Process engineers tell automation engineers what thecontrol system needs to do
14.4.2020Laitoksen nimi
8
Timeline for the course
Lectures Intro * * Excu * * * * * * * *Exercise DLs HW 1 HW 2 HW 3 HW 4
Exams SUMOquiz
Exam 1 Exam 2 Extra 1&2
Modellingproject work
Selecttopics
Questionsto Porvoo
Modellingtask
Controltask
Presentation Clientpresentation
Presentations
Written report 1st draft+feedback
Finalreport
14.4.2020Laitoksen nimi
9
Week 16 17 18 19 20 21 22
HW 1 SUMO quiz HW 2&3 HW 4
Project Topics C&C1 Excu C&C2 OutcomesPres&Rep
Course execution • Two mid-term exams• 60 min at the beginning of two
sessions (8.5. & 22.5.)• Simulation project work
• Individual work with pairdiscussions
• Create your own virtualtreatment plant, HermanninsaariWWTP
• Demonstrate a specific operationand process control study withyour process
• Presentation of the work andresults (Final session on Wed27.5.)
• Meeting with Hermannansaaripeople Friday 24.4.!!!
14.4.2020Laitoksen nimi
10
• Lectures and exercises• Lecture sessions: 3.5 hours
Wednesday afternoon and Fridaymorning
• Each session will be divided intoseveral interactive lectures, andhands-on exercises with SUMO
• Four homework exercises• Additional optional sessions for
homework assignmentsMondays between 10 and 12in Teams!
• Get to konw SUMO quiz – tobe finished in MyCourses,grading pass/failed
Communication§ MyCourses -page§ Lecture material available mostly before the lecture§ Instructions for homework assignments§ Submission of home assignments & grades§ Information and submissions for the modelling project
• Communicating– Whole course: MyCourses & email– [email protected], [email protected],
Course grading• 30 % mid-term exams - 2 exams 20 points each• 40 % individual project (presentation and written report)
- 1/3 from the presentation, 2/3 from the written report- Grading scale 1 – 5- Based on assessment of Michela and Anna
• 30 % homework exercises + activity during the course- 4 exercises, 80 points total- Bonus possibility up to 0.5 grade when attending the lectures- NOTE!! Late submission – 1 weekà 50% off, moreà 100%
off
14.4.2020Laitoksen nimi
12
Grading thresholds:1-40% of total points 2-52% 3-64% 4-76% 5-88%
Workload
Changes from last year to reduce the workload:- Lectures 11à 10- Changes in the exercises, HWs published earlier- Improved assignment documents
Learning activity Workload calculation(hours)
Remarks
Lectures 24Excercises 20Home assignments x4 20 5 hours per homework
assignmentReading materials 30 5-10 pages for most of the
sessions Project work 16 8 hour for computer simulations,
8 h for reporting and preparingthe presentation
Midterm exam (2x) 10 5h preparation for each mid-term exam
Independent reflection 15In total 135
OUR CASE STUDY PLANT –HERMANNINSAARI WWTP IN PORVOO
Project work tasks prepared onthe basis of this plant
Friday 24.4. during the courseteaching session meeting withElina Antola from Porvoo
Special focus on instrumentsat the plant and processcontrol
14.4.2020Laitoksen nimi
14
PEDAGOLOGICAL TRIAL FOR THEPROJECT WORKJIDEA: Include professionalcontext and role games intoteaching
IN PRACTICE:- Playing the role of client and
consultant in pairs- Two C&C meetings during
the project:- 1) Introduction of the topicà
questions for Porvoo- 2) Testing/training for the
final presentation- From C&C meeting 1 prepare
a Teams or Zoom recording(about 10 min per topic)
04/14/2020Laitoksen nimi
15
Course booksRieger L., Gillot S., Langergraber G., Ohtsuki T.,Shaw A.,Takacs I., Winkler S. (2012) Guidelines for UsingActivated Sludge Models
Biological wastewater treatmentAuthor(s) / Editor(s) Henze, Mogens;Loosdrecht, Mark C. M. van; Ekama, George A.;Brdjanovic, Damir Publisher IWA PublishingCopyright Date 2008ISBN 978-1-84339-188-3Electronic ISBN 978-1-68015-582-2
Olsson, G., Nielsen, M., Yuan, Z., Lynggaard-Jensen, A., Steyer, J.P.(2005). Instrumentation, Control andAutomation in Wastewater Systems,IWA Publishing, London.
14.4.2020Laitoksen nimi
17
Important to do after the introductionlecture
Install SUMO to your computer!If you have problems contact Sara or Anna!
Start to think about your project topic!!
You can read chapters 14.1. and 14.2. from the coursebook (Biological wastewater treatment)
14.4.2020Laitoksen nimi
20
Key definitions
A model is a systematicrepresentation of a realsystem
SystemInputs Outputs
Identification
Model
Disturbances
Simulation
What is a model?
The system is a defined partof the world. The interactionswith the environment (the restof the world) are described bymeasurable inputs and outputsignals and the disturbances
Key definitionsModels in general give a formal description of the system and provide arepresentation of what we consider the essential aspects of that systemin a usable form. They can be very different.
Mental models
Give an explanation of someone’s thoughtsabout how something works in the real worldwithout involving any mathematicalrepresentation
Graphical modelsProvide a convenient way to describe theproperties of a process by means of tablesor plots
Mathematicalmodels
Describe the relationships among the systemvariables in terms of difference ordifferential equations
Key definitions
A full understanding of the nature of thesystem can identifying and describing thephysical, chemical and biological laws thatgovern the system.
High interdependence of the state variablesmakes the calibration a difficult task
Structured models
Key definitions
The model development is mainly driven bymeasured data from the actual system. Itsmain advantage is that highly accuratemathematical models without detailedknowledge of the system
The accuracy of the model relies on thequality of the data
Unstructured models
Key definitions
Physical interpretable parameters that arepossible to be estimated by means ofstatistical methods
The potential advances include a reduceddemand of experimental data and morereliable extrapolation
Hybrid models
Uses of wastewater modelsWhat is the “perfect” model?
What are the characteristicsof a perfect models?
The simplest model which does the job
Science-based to the extent possible,with judicious use of empiricalknowledgeInclude only what is essential giventheir intended use
Consistent with existing and evolvingpractice
Adaptable as both knowledge andpractice requirementsevolve
G. T. DaiggerWater Science and Technology 63 (3) 516-526, 2011
Uses of wastewater models
Why models? Mathematical models are useful andnecessary for:
q System understanding: to obtain and enlarge insight in differentphenomena (ranging from physical laws to economic relationships)
q System design: simulation and operator trainingq Process control: to quantify the effect of the manipulated variables
and disturbances on the controlled variablesq Process monitoring: to obtain a template of the process under
normal conditions or to perform root-cause analysisq Soft-sensing: to estimate state variables that cannot be easily
measured in real time on the basis of the available measurementsq Optimization: to identify constraints among decision variables and
make optimal decisions
Model building exerciseSelect the model purpose: required modelaccuracy, model boundaries, …
Determine the system parameters: equipmentsizes, volumes, process topology, …
Develop a set of axioms or descriptionof the process
Evaluate the a priori knowledge of theprocess and postulate themechanisms within the process
Test the constructed model by simulations toanalyze/verify its behavior (at steady-state anddynamically)
Modelling platforms
Programming languages
Commercial software
q Matlab/Simulinkq C++q Fortranq Pythonq ...
q Sumo – Dynamitaq GPS-X – Hydromantisq Simba – Ifak systems GmbHq West – Mike powerd by DHIq BioWin – Envirosimq ...
Modelling platforms
q Interfaces with models of entire wastewater treatment systemsq Able to model entire wastewater treatment plantq Incorporates wide range of unit processesq Flexible and adaptable
§ Accommodates range of unit process models§ Easily modified units to evaluate potential innovations§ New unit processes easily formulated and incorporated
q Does not require extended learning to at least perform basiccalculations
q Interfaces easily with other engineering automation toolsq Accommodates unit processes with diverse characteristics
Characteristics
G. T. DaiggerWater Science and Technology 63 (3) 516-526, 2011