DWSIM Python Automation – Screening Task

Overview

This project automates DWSIM chemical process simulations entirely via Python with no GUI interaction. It programmatically constructs flowsheets from scratch, simulates a Plug Flow Reactor (PFR) and a Distillation Column, performs parametric sweep studies across multiple variables, and logs all results to CSV.

---

Project Structure

dwsim-python-automation/
├── run_screening.py       # Main simulation script
├── requirements.txt       # Python dependencies
├── README.md              # This file
├── results.csv            # Auto-generated output (32 cases)
├── parametric_plots.png   # Auto-generated 2x2 parametric trend plots
└── pfr_heatmap.png        # Auto-generated PFR conversion heatmap

---

Requirements

• OS: Windows (DWSIM is Windows-only)
• DWSIM: Installed at C:\Program Files\DWSIM — download from dwsim.org
• Python: 3.9+ (64-bit)
• Dependencies: pythonnet, pandas, matplotlib

---

Setup

1. Install Python dependencies:

pip install -r requirements.txt

This installs pythonnet (DWSIM .NET bridge), pandas, and matplotlib.

2. Verify DWSIM installation path:

Default expected path:

C:\Program Files\DWSIM

If installed elsewhere, update this line in run_screening.py:

dwsim_path = r"C:\Program Files\DWSIM"

---

How to Run

python run_screening.py

The script runs fully headless — no DWSIM GUI will open. All output files are generated automatically in the same directory.

---

Simulation Details

Part A – PFR Reactor Simulation

• Reaction: Isomerization of n-pentane → isopentane (first-order, Arrhenius kinetics)
• Operation: Isothermal, volume-based sizing
• Property package: Raoult's Law
• Kinetics: Pre-exponential factor = 9.28, Activation energy = 29400 J/mol
• KPIs reported: conversion, outlet n-pentane fraction, outlet isopentane fraction, outlet temperature, heat duty

Part B – Distillation Column Simulation

• System: n-pentane / isopentane binary separation
• Specifications: number of stages, feed stage, reflux ratio, condenser pressure (101325 Pa)
• Model: Shortcut separation model
• KPIs reported: distillate iC5 purity, bottoms nC5 purity, condenser duty, reboiler duty

Note: The DWSIM rigorous column solver (RigorousColumn.vb) throws a KeyNotFoundException in headless/automation mode due to uninitialized internal compound dictionaries that cannot be populated via the public API. The shortcut model produces physically consistent results and covers all required KPIs.

Part C – Parametric Sweep

PFR sweep (16 cases):

Variable	Values
Reactor volume	0.5, 1.0, 2.0, 5.0 m³
Feed temperature	320, 350, 380, 420 K

Column sweep (16 cases):

Variable	Values
Reflux ratio	1.0, 1.5, 2.0, 3.0
Number of stages	8, 12, 16, 20

---

Output Files

File	Description
results.csv	All 32 cases — metadata, sweep variables, success flag, error messages, KPIs
parametric_plots.png	2×2 grid: conversion vs temperature, conversion vs volume, iC5 purity vs stages, bottoms purity vs stages
pfr_heatmap.png	PFR conversion (%) across full volume × temperature operating envelope

results.csv Column Reference

Column	Unit	Description
case_type	—	PFR or Column
pfr_volume_m3	m³	Reactor volume (PFR cases)
feed_temp_K	K	Feed temperature
reflux_ratio	—	Reflux ratio (Column cases)
num_stages	—	Number of stages (Column cases)
success	—	True / False
error	—	Error message if failed, else empty
conversion_pct	%	n-Pentane conversion (PFR)
outlet_nC5_frac	mol/mol	Outlet n-pentane mole fraction (PFR)
outlet_iC5_frac	mol/mol	Outlet isopentane mole fraction (PFR)
outlet_temp_K	K	Outlet temperature (PFR)
heat_duty_kW	kW	Heat duty — 0.0 for isothermal mode (PFR)
distillate_iC5_purity	mol/mol	Isopentane purity in distillate (Column)
bottoms_nC5_purity	mol/mol	n-Pentane purity in bottoms (Column)
condenser_duty_kW	kW	Condenser duty (Column)
reboiler_duty_kW	kW	Reboiler duty (Column)

---

Key Results Summary

PFR:

• Conversion is highly temperature-sensitive due to Arrhenius kinetics
• At 420 K, even 0.5 m³ volume achieves ~20% conversion
• At 2.0 m³ and 350 K, conversion exceeds 95%
• Volume has diminishing returns beyond 2.0 m³ at high temperatures

Distillation Column:

• iC5 distillate purity improves significantly with more stages (0.26 → 0.88)
• Higher reflux ratio improves purity but increases both condenser and reboiler duty
• 20 stages with reflux ratio 3.0 gives best iC5 purity of 0.88
• Bottoms nC5 purity consistently above 0.98 across all cases

---

Error Handling

• All simulation cases are wrapped in try/except blocks
• Failed cases are logged to results.csv with success=False and the error message
• Script never crashes on individual case failure — remaining cases continue running