Teaching · Graduate · 2105623

Optimization of Chemical Processes

Fifteen weeks, weighted toward the part that is actually hard. Weeks 1–5, model construction: turning a plant problem into algebra that says what you meant. Week 6: drafting models with AI assistance, and proving them right or wrong. Weeks 7–13, theory and methods: what solvers do, why they behave as they do, and where data-driven methods beat mathematical programming.

Code
2105623 · 3 (3-0-9) · elective
Term
Semester 1 / 2026
Sessions
Monday · 10 August to 16 November 2026
Taught by
Soorathep Kheawhom
Tools
Excel Solver and OpenSolver in the early weeks, then Python with Pyomo. No commercial solver is required.
Department
Chemical Engineering, Faculty of Engineering, Chulalongkorn University

Slides

These open in the browser — no download, and they work on a phone. Press F for full screen, E for a printable layout, ? for every shortcut.

Weekly schedule

#SessionDateHands-on
1 Introduction, the optimization workflow, and the tool stack
Orientation
10 Aug Excel Solver
2 Modeling I: linear models, product mix, multiperiod planning and inventory
Model construction
17 Aug OpenSolver
3 Modeling II: networks, flows and process superstructures
Model construction
24 Aug OpenSolver
4 Modeling III: blending, pooling and quality specifications
Model construction
31 Aug Excel, then Python
5 Modeling IV: logical and discrete decisions
Model construction
7 Sep OpenSolver and Pyomo
6 AI-assisted modeling and model debugging
Tools and verification
14 Sep Pyomo · workshop
7 Linear programming theory and solver behaviour
Theory and algorithms
21 Sep Python
8 Duality and sensitivity analysis
Theory and algorithms
28 Sep Excel and Pyomo
9 Unconstrained nonlinear programming
Theory and algorithms
5 Oct Python
10 Constrained nonlinear programming
Theory and algorithms
12 Oct Pyomo and Ipopt
11 Integer and mixed-integer programming
Theory and algorithms
19 Oct Pyomo and CBC
12 Optimization under uncertainty
Advanced topics
26 Oct Pyomo
13 Data-driven and AI-based optimization
Advanced topics
2 Nov Python
14 Project presentations, part 1
Synthesis
9 Nov
15 Project presentations, part 2, and wrap-up
Synthesis
16 Nov

Assessment

WeightComponentWhat it asks
35% Final examination Cumulative, weighted to Weeks 9 to 13 — nonlinear and integer programming, uncertainty, data-driven methods
25% Midterm examination Weeks 1 to 8, weighted to model construction. Includes a model-verification question. No nonlinear programming
15% Assignments Four problem sets — model construction, LP and duality, nonlinear programming, integer programming
15% Computational project A real optimization problem of your own, modelled in Pyomo, with a test suite and a presentation
10% Participation and quizzes In-class activities, the weekly notebook, short quizzes
Generative AI. Drafting an optimization model with AI assistance is not merely permitted in this course, it is expected, and Week 6 teaches the workflow. What is graded is verification: every AI-assisted deliverable states which parts were AI-drafted and which tests were run against them. An unverified generated model receives no credit for model correctness, however plausible its numbers look. Generative AI is prohibited in both examinations.