Modelling and optimization of energy systems

ME-454

Media

ME-454 Modelling and optimization of energy systems

DOF process units and flowsheets

23.11.2023, 15:40

IPESE course - DOF process units and flowsheets October 16th 2023, 8:23:48 am

23.11.2023, 15:40

IPESE course - heat exchanger network design November 20th 2023, 2:15:48 pm

20.11.2023, 15:49

IPESE course - heat exchanger network design November 20th 2023, 2:15:48 pm

20.11.2023, 15:49

IPESE course - wind and hydro November 20th 2023, 11:16:08 am

20.11.2023, 12:53

IPESE course - wind and hydro November 20th 2023, 11:16:08 am

20.11.2023, 12:53

IPESE course - measurement reconcialiatoin and parameter identification November 20th 2023, 8:20:12 am

20.11.2023, 09:42

IPESE course - measurement reconcialiatoin and parameter identification November 20th 2023, 8:20:12 am

20.11.2023, 09:42

IPESE course -defijnig hot and cold streams November 13th 2023, 3:17:49 pm

13.11.2023, 16:05

IPESE course -defijnig hot and cold streams November 13th 2023, 3:17:49 pm

13.11.2023, 16:05

IPESE course - constitutive equations and resolution sequence November 13th 2023, 8:23:14 am

13.11.2023, 13:01

IPESE course - constitutive equations and resolution sequence November 13th 2023, 8:23:14 am

13.11.2023, 13:01

IPESE course energy conversion Solar Energy - November 13th 2023, 11:17:16 am

13.11.2023, 13:01

IPESE course energy conversion Solar Energy - November 13th 2023, 11:17:16 am

13.11.2023, 13:01

Solving non-linear optimization problems

27.10.2023, 17:49

IPESE optimisation - October 23rd 2023, 8:20:25 am

27.10.2023, 17:49

Solving linear optimization problem

27.10.2023, 17:45

IPESE optimisation - October 23rd 2023, 8:20:25 am

27.10.2023, 17:45

Solving strategies for an optimization problem

27.10.2023, 17:38

IPESE optimisation - October 23rd 2023, 8:20:25 am

27.10.2023, 17:38

KPIs

20.10.2023, 16:20

IPESE course - October 9th 2023, 8:21:13 am

20.10.2023, 16:20

Investment estimation

20.10.2023, 16:16

IPESE course - October 9th 2023, 9:26:39 am

20.10.2023, 16:16

IPESE course - October 9th 2023, 9:26:39 am

09.10.2023, 10:06

IPESE course - October 9th 2023, 9:26:39 am

09.10.2023, 10:06

KPIs and clustering

09.10.2023, 09:13

IPESE course - October 9th 2023, 8:21:13 am

09.10.2023, 09:13

Introduction to course and project MOES

06.10.2023, 14:44

Introduction and course concept

Modeling and optimisation of energy systems introduction

06.10.2023, 14:44

Introduction and course concept

Solving non linear equations

06.10.2023, 14:08

buildings and solving non linear equations

06.10.2023, 14:08

building demands

06.10.2023, 12:13

buildings and solving non linear equations

06.10.2023, 12:13

Modeling and optimisation of energy systems introduction

02.10.2023, 09:52

Introduction and course concept

Modeling and optimisation of energy systems introduction

02.10.2023, 09:52

Introduction and course concept

Miro 3 - Flipped class

13.12.2021, 13:03

Flipped class about T9-T12

MOES Project 2021

29.11.2021, 08:31

MOES Lecture 2021

29.11.2021, 08:28

MOES Miro 2021

29.11.2021, 08:29

Miro Session 1

23.11.2021, 11:17

Miro 2 - Flipped class

23.11.2021, 11:15

Lecture 2021 week 2

04.10.2021, 11:35

2021_Week14: Semester and Master projects @ IPESE

26.05.2021, 14:32

Short explanation of the projects available for next semester @IPESE.

2021_Week13: Oral exam procedure and Theory discussion

19.05.2021, 16:16

|1.I can explain what a state variable is and derive the degrees of freedom of a flow in a process | Jacopo Saracco, Justine Brun | https://miro.com/app/board/o9J_lGZDmw8=/

|2.I can explain what a constitutive equation is and how a thermodynamic model works | Marc Tognola de Quintana, Claire Marie Isabelle Bernier | https://miro.com/app/board/o9J_lFK97xs=/

|3.I can explain what the equations of a process unit are and I can formulate the equations for units like a compressor, a heat exchanger or a cogeneration unit and explain what the definition of the parameters in the model is. | Nicolas Jean Rospars, Olivier Jean Mathias Laferrère | https://miro.com/app/board/o9J_lDi44fs=/

|4.I can explain the flowsheet of a heat recovery system | Paul Arthur Guy de Durand,Raphaël Briguet | https://miro.com/app/board/o9J_lEMNI6E=/

|5.I can realize a degree of freedom analysis, define the specifications and explain what dependent and independent variables of a model are. | Samuel Martin Peter Meyer | https://miro.com/app/board/o9J_lDvaL1I=/

|6.I can explain how to solve a unit model using a sequentialapproach and I can explain the pro and cons of a sequentialapproach. | Sven Luca Menge, Vincent Paul Python | https://miro.com/app/board/o9J_lEZwrl8=/

|7.I can explain the solving methods that can be used in a sequential solving approach. | Wei-Sheng Hung, Xinghai Wang | https://miro.com/app/board/o9J_lECYIOU=/

|8.I can explain how to solve a unit model using a simultaneous approach and I can explain the pros and cons of this approach. I can give an example from my project. | Yexinlei Yang, Yuanjun Feng | https://miro.com/app/board/o9J_lEUhjBM=/

|9.I can explain what the relevant conditions for using a simultaneous approach are and how this approach compares with the sequential one. | Romain Lilian Douat, Timothée Benjamin Antoine Jaubert | https://miro.com/app/board/o9J_lDxHqeE=/

|10.I can explain how to analyze the degrees of freedom of a flowsheet and what is needed to reach zero degrees of freedom.I would be ready to apply it on a given energy conversion flowsheet. | Louis-Nicolas Paul Raymond Durand | https://miro.com/app/board/o9J_lDjmj_0=/

|11.I can apply the Motard method to define a sequence to solve a sequential modular simulation problem. I can explain what the difficulties of the sequential approach are. | Marguerite Lou Lavarini, Anaël Perruchoud | https://miro.com/app/board/o9J_lDu-5-Q=/

2021_Week_10: Theory discussion

28.04.2021, 19:34

|1.I can explain the possible type of usage of optimisation in energy system modeling | Dominik Blaser,Mouhannad Abou Daher | https://miro.com/app/board/o9J_lHHhFnY=/

|2.I can explain the problem of parameter estimation and explain how the problem is stated and solved. I can give an example on how it has been applied in my project. | Guillaume Pierre Nicolas Senentz,Pierre Thomas Robin Marie Bouquet | https://miro.com/app/board/o9J_lHglht8=/

|3.I can explain different strategies to state and solve an optimization problem. I can explain differences between black box, simultaneous, hybrid methods and explain pros and cons. | Antoine Ravetta,Olivia Marie Lucie Julia | https://miro.com/app/board/o9J_lIdJDBQ=/

|4.I can explain the different methods to solve optimization problems and explain what are pros and cons of each of them. | Charles Marc Célin Vuichard,Pål Forr Austnes | https://miro.com/app/board/o9J_lILgnNQ=/

|5.I can explain how to solve an unconstrained optimization problem. | Camila José Morales Undurraga, Matthieu Benoit C. Jacobs | https://miro.com/app/board/o9J_lIbDXTA=/

|6.I can explain how to solve a multi-variable unconstrained optimization problem. | Chiara Ongaro, Clara Emmanuelle Gualtieri Huguenin | https://miro.com/app/board/o9J_lHmWvV4=/

|7.I can explain what a Mixed Integer Linear Programming optimization problem is and how it has been used in the energy system optimisation project. | Daniel Gutierrez Navarro, Edouard Cattin | https://miro.com/app/board/o9J_lHl4Bh0=/

|8.I can explain what are the principles of a heuristic method for solving an optimisation problem and I can explain for which strategy it is used for and I can comment onthe associated pros and cons. | Gabriel Bessette, Hugo Pierre André Casagrande | https://miro.com/app/board/o9J_lIcYf-Y=/

2021_Week_9: part 4 updates (FM, JS)

21.04.2021, 12:32

2021_Week_8: part 3 and part 4 (FM, JG, JS)

14.04.2021, 10:55

Part 4 Project description - Project tasks

12.04.2021, 11:50

Part 4 tasks

Simulation / Flow sheeting
Measurements
Reconciliation
Performance analysis
Fluid selection

Part 4 Project description - Case study

12.04.2021, 11:49

Case study of P2G plant Solothurn

Simulation
Measurements
Reconciliation
Validation

Part 4 Project description

12.04.2021, 11:48

Project overview
Flowsheeting

2021_Week_6: Theory discussion

02.04.2021, 10:16

1.I can explain how to state an energy system model using

mixed integer linear programming : what are the constraints,

what is the objective function, what are the inequality

constraints https://miro.com/app/board/o9J_lNcqQl0=/

2.I can explain the simplifying assumptions made to model the

system with a linear programming problem. https://miro.com/app/board/o9J_lMgMz7c=/

3.I can explain how to model a heat pump with a simplified

model based on temperatures https://miro.com/app/board/o9J_lNJAahw=/

4.I can explain how to calculate the operating cost of an

energy system and explain how it has been done in the project https://miro.com/app/board/o9J_lNUFd-M=/

5.I can explain how to estimate the investment of an energy

conversion system. https://miro.com/welcomeonboard/TS4BQF5czyruVB8N4u920lPaMhgrIzMNVkL9MaxICyaMInYoRbuJAQSOXbWokhqc

6.I can explain how to calculate the key performance indicators of

an energy system. I can explain how to calculate the

thermo-economic performance, and how to consider

the environmental and sustainability aspects. https://miro.com/app/board/o9J_lN5i340=/

VALI tutorial HP

30.03.2021, 10:22

Defining streams
- Mechanical
- Material
Defining units
- Tags
- Compressor
Heat pump first stage: compression

VALI tutorial

30.03.2021, 10:19

Opening VALI
Creating a first PFD
Defining thermods

2021_Week4_VM_AMPL by XL

18.03.2021, 18:01

2021_week3: Theory question by FM

10.03.2021, 13:38

2021_Week2 - Presentation of project part 1 (FM, DL)

03.03.2021, 10:25

2021_week1: course introduction by prof

24.02.2021, 16:36

Introduction of course by Prof

2021_week1: Presentation of tools and projects (FM, JG, XL, JS)

24.02.2021, 16:28

Francois, Julia, Xiang, Jonas

Examination procedure, MOO Introduction and Case Study Part 5 by JH & FM

13.05.2020, 17:00

MILP model complete by FM

01.05.2020, 16:47

Energy technologies - part 1 by FM

01.05.2020, 16:44

Introduction by FM

01.05.2020, 05:47

Thermo- economic analysis by FM

01.05.2020, 05:43

Degrees of freedom and process unit models by FM

30.04.2020, 15:50

Key performance indicators and multi objective optimisation by FM

30.04.2020, 13:49

Optimisation - heuristic methods by FM

30.04.2020, 13:48

Data reconciliation and parameter identification by FM

30.04.2020, 13:46

Defining a sequence application to the two stage heat pump by FM

30.04.2020, 13:44

Defining a resolution sequence by FM

30.04.2020, 13:43

Constitutive equations by FM

30.04.2020, 13:34

Solving an optimisation problem by FM

30.04.2020, 13:25

Solving strategies for optimisation problems by FM

30.04.2020, 13:16

Energy technologies - part 2 by FM

30.04.2020, 13:11

MILP model and typical days by FM

30.04.2020, 12:43

Solving equations by FM

30.04.2020, 12:20

Building model by FM

30.04.2020, 12:17

Presentation of the EPFL project by FM

30.04.2020, 12:15

"Where are we and what is next?" by FM

30.04.2020, 11:44

Data reconciliation Case Study by JG & FM

30.04.2020, 11:43

NLP Case Study by RA

30.04.2020, 11:41

MILP Case Study by XL

30.04.2020, 11:40

NLP example by FM

30.04.2020, 11:38

MODELLING AND OPTIMIZATION OF ENERGY SYSTEMS
Session (Mon 08:15-12:00)
Instructor: Prof. François Maréchal

Assistants: Sai Sudharshan Ravi, Dr Luc Girardin, Arthur Waeber, Michele Poli

Overview:

The objective of this course is to teach you the basics of computer aided process system engineering techniques for modelling and optimization of energy systems. The students will learn how to build energy systems models and how to carry out thermo-economic optimization for optimal system's design. Both linear and non-linear problems will be modelled and solved using proper tools. Multi objective optimizations techniques will be used to consider both economic and environmental targets. The application follows the principles of computational thinking, helping students to:

A: Analyse: Activate your knowledge to state the problem to be solved and collect the necessary information for it, from the engineering knowledge or from observations
G: Generate: Apply calculation methods to generate the numerical results that characterise the problem solving results
I : Interpret: Transform the numerical results into an engineering problem solution that makes sense for an engineer
R: Report: Report in the appropriate format the useful engineering results generated as a problem solution to be proposed to colleagues or peers.

The main concepts will applied in a group project.

Slides for the introduction

Class planning:

Each Monday, there is a course meeting at EPFL. The class meeting consists of different elements:

08.30-10:00 : lectures on theory by Professor Maréchal (note that lecture will start at 08:30 due to CFF time table Sion-Ecublens), breaks are depending on the lecture content.

10:15-12:00 : project/exercice by assistants. Assistant will introduce projects parts and support for the realisation.

Open office hour: by the assistant (one hour of zoom presence to solve IT or implementation problems based on the forum unanswered questions).

Recorded classes and presentations from previous years can be found here.

Examination method is described here

Resources:

Lecture slides as well as relevant videos are uploaded on Moodle.
For questions regarding the theory, please use the forum "theory questions"
For project-related questions that are relevant for the whole group, please use the forum "Project related questions".
If you have questions concerning the organisation of the course, please use the the forum "course organisation questions".
If you need to refresh your knowledge in Thermodynamics and engineering, please use this free course.
The VMs are accessible using VMware Horizon of EPFL in the server https://vdi.epfl.ch under the name STI-FM-cours-2025

Problems in accessing the materials ?

1. If you cannot access the virtual machine (VM) (VPN needed outside EPFL campus), please contact 1234@epfl.ch ;
2. if you have problems on VM, e.g. blackscreen, frozen screen etc, please contact 1234@epfl.ch ;

Group selection and learning quarto

For the first wekk you will have to create groups and learn the quarto computational thinking tool that will be applied in the class.

There will be two assignment : the group formation and the generation of your contribution to the trombinoscope of the course. Who are those students who are willing to learn Modeling and Optimisation of Energy Systems.

Form your group (Deadline of group formation: 23:59 on Sept 14, 2025) (Group choice)

Project Overview/Storyline

The EPFL energy system is facing the problem of the the Quagga mussels that requires the identification of solutions to solve the problem of pipes being blocked by the invasion of the mussels.

Among the solutions proposed, EPFL would like to investigate an new decentralised heating strategy and to compare it with the actual one using the centralised heat pump.

The idea is based on the concept of anergie in which the heat from the environment like a lake or geothermal wells is distributed to buildings, instead of being pre-conditioned in a centralised district heat pump and distributed to the building. The advantage of the anergie concept is that each building has its own heat pump, and can establish its own energy efficiency and harvesting strategy and could even organise its own energy storage. It also provides at the same time heating and cooling allowing heat recovery when heating and cooling is needed simultaneously in different buildings.

The anergie concept can be implemented by using a water loop that would exchange with the water from the lake which would allow to established a centralised treatment of the mussel invasion and could profit from synergie between different sources, in particular geothermal sources.

Researchers of EPFL have in particular proposed and tested in EPFL-Sion an innovative way of implementing the anergie concept using CO2 as a heat transfer fluid. The system is now commercialised by an EPFL start-up exergo.

Weber and Favrat 2010
Henchoz et al. 2011
Application to a city : Suciu et al. 2018

In addition, based on the preliminary results of the heating bits project, it appears as well that heat recovery from data production could become more and more important as EPFL has invested in the LLM open access apertus project.

Base on the recent promising results of multiple studies in EPFL in the development of PV solar cells (CSEM Neuchatel, Nazeerudin) , EPFL is questioning why those cells are not already demonstrated on its campus and would like to know what would be the sustainability cost of those solutions when compared with the state of the art market based solution of the today.

EPFL direction is also aware of the numerous projects on smart building operation and sensors from the urban twins project and would like to investigate what will be the sustainability impact of renewing its energy system integrating innovation.

As a group of engineer you are therefore asked to assess the sustainability impact of the integration of decentralised solutions to renovate the EPFL campus.

The direction would like to receive a comparison of the options identified considering the sustainability aspects : i.e. economic, environmental impact, and impact on the EPFL operation, it particular it would like to know what will be the investments associated.

EPFL project slides presentation

EPFL district heating and coolling system presentation

Course evaluation and Grading

Due to the large number of students, this course will be evaluated in a hybrid mode, combining continuous evaluation, group project report evaluation and a written exam.

Here is the grading strategy:

50% : for the group report based on the resolution of the 3 parts of the project as a whole. The report grade will be "weighted" by an anonymous peer-self-evaluation (each member will declare how much they contributed to the report). Note that you are supposed to know what is in the report as you have signed it.
5%: your results to the quizzes during the semester. Quizzes are set-up to
15%: your peer review of the report of an other group. At the end of the semester you will receive one report from another group to be reviewed. As an assignment to be delivered at the start of the exam period you will have to deliver a review of the report
30%: your answer to the review. For the written exam, you will receive a 4 reviews concerning your report: one from an assistant and 3 from your peers. The written exam will consist in commenting the reviews and explaining what you will do to answer the remarks and suggestions of the reviewers.

Course schedule overview

The lectures and project work take place in classroom GCC330 (sorry for the double deck noise perturbations !!!)

Week	Date	Theory covered	Project goals		Submissions
1	08.09.2025	computational thinking AGIR	Task1 : building energy demand (heating and cooling) and energy bill, impact of climate change
2	15.09.2025	T 1: Energy demand analysis: building modeling T 2: Equations and solving methods	Task 1 : prospectives : Comfort T, CO2 control, adding insulation, CO2 impact
3	22.09.2025	Holiday	Task 1 : prospectives : Comfort T, CO2 control, adding insulation, CO2 impact
4	29.09.2025	T 3: Key Performance Indicators T 4: Estimating the investment	Task 1 : prospectives : Comfort T, CO2 control, adding insulation, CO2 impact	Quiz on Task 1
5	06.10.2025	T 5.1: Flowsheet simulation and DOF T 7: Constitutive equations	Task2 : simulation energy system : heat recovery on the ventilation system
6	13.10.2025	T 8: Resolution sequence definition	Task 2: simulation energy system, heat pump integration, data center integration, heat recovery from electrical appliances, adding insulation, using lake, geothermal or Air heat pump
7	20.10.2025	Break	Task 2
8	27.10.2025	T 5.2: Stating optimisation problem and solving strategies	Task 2: simulation energy system, heat pump integration, data center integration, heat recovery from electrical appliances, adding insulation, using lake, geothermal or Air heat pump	Quiz on Task 2
9	03.11.2025	Optimisation problem types T 6: Solving optimization problem	Task 3 : optimizing investment (NLP) + Fluid selection
10	10.11.2025	T 9: Data reconciliation and parameter identification	Task 3
11	17.11.2025		Task 3	Quiz on Task 3
12	24.11.2025	T 10: MILP optimization methodology	Task 4: system integration : choosing among options
13	01.12.2025	T 11: Techno-economic analysis by MILP optimization T13: Multi-objective optimization by heuristic methods	Task 4
14	08.12.2025		Task 4	Quiz on Task 4
15	15.12.2025	"How to conduct a review?"- Prof. Françios Maréchal	Task 4

Task 1: Building energy demand

In this part of the project you will model the energy demand the EPFL Campus, based on past measurements of the EPFL buildings. Expected understandings from this part would include, understanding: What are the ambient and behavioural parameters that could affect the building energy demand, e.g. what is the impact of the comfort temperature set point in the building, what is the impact of the presence of the people and what is the impact of the solar gains or the electricity consumption in the buildings or what is the impact of renovation of the buildings.

Course materials

Task 2 : Modeling and assessing energy systems

This part focuses on nonlinear optimization (NLP) to evaluate heat recovery solutions for EPFL. Students model and optimize different configurations (e.g., data center recovery, air ventilation recovery, and heat pump integration) to assess their technical and economic benefits.

Course materials

Task 3: Optimising energy system configurations

The objective is to model, calibrate, and analyze a two-stage heat pump using Belsim VALI. Students perform data reconciliation, compare working fluids (e.g., R-290, ammonia), and derive performance and cost models for use in later optimization stages.

Course materials

Task 4: Choosing among options for integrated energy systems

The final part integrates all previous results into a Mixed-Integer Linear Programming (MILP) model to optimise the entire EPFL energy system. The goal is to identify cost-effective, low-emission technology mixes under various scenarios and perform multi-objective and sensitivity analyses to inform decision-making.

Course materials

Guide - AMPL, Quarto, GIT, Vali, NOTO

Project material

During the semester, each group will receive access to a dedicated GitLab repository. This repository contains everything you need for the project: the required data, a Quarto (

.qmd

) template that provides a suggested structure for your report, and a tutorial to help you get started. The Quarto format is designed to integrate both calculations and reporting in one environment, making it easier to present your work in a clear and reproducible way.

To explore the tutorial, download the

.zip

file from the repository, open the

_book

folder, and double-click any of the

.html

files to open them in your browser.

That said, you are not required to use Quarto. You are free to perform your calculations using Jupyter Notebooks (e.g., via noto.epfl.ch) or any other computational environment. You can also use other tools like Overleaf for writing your report, as long as the final output is clear and well-structured.

At the end of the project, each group must submit:

Your working repository (with code, data, and any documentation)
A PDF report of your results

You will be evaluated based on the quality of your scientific reporting, not the specific tools you choose to use.

Project Description (Folder)

Discussion Forums

During this course, we will be using the discussion forums to communicate. In addition, you can contact one of the TAs to organise an office hour for other clarifications.

Lectures Notes, Slides and Videos

The lecture notes and FAQ website is available in the "Lecture_notes" link below.

Report & Review Submission

Submission Section

Final Report – Due: 19 December 2025

Submit one report per group:

Include the signed declaration of authorship with each member’s contribution.

Review Report – Due: 12 January 2026

Your review must fit on one single page (aside from the cover page with just your name + SCIPER).

Because space is minimal, the review should be short, focused, and essential.

Your one-page review should include only:

1. A brief overview (2–3 sentences)

What the project did and what it concluded. Nothing more.

2. A short paragraph of general feedback

A few sentences noting:

clarity of the goals
quality of the storytelling
structure/flow
usefulness of the figures

This should stay high-level and concise.

3. A short paragraph of detailed feedback

Select only the most important comments with the things that genuinely matter for understanding, correctness, or clarity. This is not the place for long lists. Stick to 2–4 key points, phrased as short sentences.

3. One question per part of the project

This is the key requirement:

➡️ For each part of their report, write one thoughtful, relevant question that probes the authors’ understanding or assumptions.

Examples:

“How sensitive is your heating demand estimation to the assumed cut-off temperature?”
“Why did you choose this clustering method over alternatives?”
“What is the physical intuition behind your scenario results?”

These questions count as part of the one page, so they should be short but meaningful. Some more general guidelines on conducting a good peer review can be found here.

TAs remain available for questions until the final submission, but please plan ahead. Availability is limited during the holidays.

Good luck!

Conclusions and summary of the peer review and the written exam (File)