Skip to content

Ethanol Plant (Fermentation)

In this tutorial you will model an industrial-scale ethanol fermentation process in DWSIM's Classic UI: glucose is fermented to ethanol and CO2 by yeast, then the products are separated.

What you will learn

  • How to model a fermentation reaction with a Conversion Reactor
  • How to handle gas-liquid product separation after the reaction
  • How to verify fermentation yield against theoretical maximum

Prerequisites

Process Overview

Yeast (Saccharomyces cerevisiae) converts glucose anaerobically:

C6H12O6 → 2 C2H5OH + 2 CO2

Gay-Lussac stoichiometric yield: 51.1% ethanol by mass. Real fermentations achieve 90-95% of this.

We model the bioreactor as a Conversion Reactor with 95% conversion, then separate CO2 from the ethanol/water beer.

Process Flow Diagram

graph LR
    F["Fermenter Feed<br/>20% glucose<br/>305 K"] --> H["H-1<br/>Heater<br/>308 K"]
    H --> R["R-1<br/>Conv. Reactor<br/>95% conv"]
    R --> CL["CL-1<br/>Cooler<br/>290 K"]
    CL --> SEP["SEP-1<br/>Separator"]
    SEP -->|Gas| CO2["CO2 Vent"]
    SEP -->|Liquid| BR["Beer<br/>(EtOH + H2O)"]

Key Design Parameters

Parameter Value
Compounds Water, Glucose, Ethanol, Carbon dioxide
Property Package NRTL
Feed 1 kg/s, 305 K, 20% glucose / 80% water (mass)
Reactor T 308 K (35 °C), isothermal
Glucose conversion 95%
Cooler outlet 290 K

Step-by-Step in the Classic UI

1. Set up

File > New Chemical Process Model:

  • Compounds: Water, Glucose, Ethanol, Carbon dioxide
  • Property Package: NRTL

Why NRTL for fermentation?

Water-ethanol forms a strongly non-ideal liquid mixture with hydrogen bonding and an azeotrope at 95.6 wt% ethanol. Activity-coefficient models like NRTL handle this; cubic EOS such as PR or SRK fail to predict the azeotrope correctly.

Glucose in the database

The DWSIM database includes Glucose. If your database is missing it, use File > New Compound Creator Study to create it from molecular structure or import from ChemSep.

2. Define the fermentation reaction

Edit > Simulation Settings > Reactions → Add new Conversion reaction:

  • Name: Fermentation
  • Stoichiometry: Glucose = -1, Ethanol = +2, Carbon dioxide = +2
  • Base compound: Glucose
  • Conversion: 95 %

Create Reaction Set FermSet containing this reaction.

Why a Conversion Reactor (not Equilibrium)?

Fermentation kinetics are slow and never reach equilibrium; conversion is determined by yeast viability, residence time, and inhibitor concentrations, all empirical rather than thermodynamic. You specify the conversion (95%) measured experimentally instead of letting Keq decide.

3. Build the flowsheet

Drag and configure:

  1. Material Stream Feed: T=305 K, P=1 atm, mass flow=1 kg/s, mass fractions: Water=0.80, Glucose=0.20
  2. Heater H-1 (outlet T=308 K)
  3. Material Stream Heated (empty)
  4. Conversion Reactor R-1: Isothermal mode, Reaction Set = FermSet, ΔP=0
  5. Material Stream Reactor-Out (empty)
  6. Cooler CL-1 (outlet T=290 K)
  7. Material Stream Cooled (empty)
  8. Separator Vessel SEP-1
  9. Material Stream CO2-Vent (empty)
  10. Material Stream Beer (empty)

Wire connections appropriately (Heater → Reactor → Cooler → Separator), with energy streams on H-1, R-1, and CL-1.

4. Solve

F6 ON → Solve.

5. Inspect yield

Open the Beer stream Results tab:

  • Mass flow: ~0.7-0.8 kg/s
  • Ethanol mass fraction: ~0.10 (10% beer, typical industrial value)

Open CO2-Vent Results: mass flow ~0.1 kg/s, mostly pure CO2.

Use Flowsheet Analysis > Mass and Energy Balance Summary to check:

  • Glucose in feed = 0.2 kg/s
  • Ethanol out (beer) ≈ 0.097 kg/s (95% × 51.1% × 0.2 = 0.097 kg/s)
  • Yield = 48.5% (matches Gay-Lussac × 95% conversion)

Fermentation yield from balance summary

Results and Validation

Variable Expected
Beer ethanol mass fraction 0.08 - 0.12
CO2 vent flow ~85% of theoretical CO2
Glucose conversion 95% (specified)
Yield (EtOH/glucose) ~48-49 g/g × 100

Expected results

Beer at 9-10% ethanol by mass (typical industrial fermentation produces 10-15% ABV before distillation). CO2 separates almost completely as gas. Mass yield ~ 49% (95% × 51.1%).

Understanding the Results

Gay-Lussac stoichiometry: 1 mol glucose (180 g) → 2 mol ethanol (92 g) + 2 mol CO2 (88 g). Maximum yield 51.1% by mass.

In practice, beer must be distilled to concentrate ethanol above the azeotrope (or use molecular sieves) for fuel-grade applications.

Automating This Tutorial

Files in this repository

See examples/advanced/05_ethanol_plant.py in the DWSIM.Tutorials repository.

dwsim.reaction.define_conversion, then dwsim.unitop.add for Heater, ConversionReactor, Cooler, Separator.

Output may vary

Results depend on the LLM's reasoning quality and tool-use accuracy. Always verify the simulation matches your intent before relying on the numbers.

Use DWSIM (via the MCP server) to build the following simulation:

- Create a flowsheet called "EthanolPlant"
- Add Water, Glucose, Ethanol and Carbon dioxide as compounds; set the
  property package to "NRTL"
- Define a conversion reaction "Fermentation" with stoichiometry
  Glucose = -1, Ethanol = +2, Carbon dioxide = +2; base compound =
  Glucose, conversion = 95%; add it to a new reaction set "FermSet"
- Add a material stream "Feed" at 305 K and 1 atm with mass flow
  = 1 kg/s and mass fractions Water = 0.80, Glucose = 0.20
- Add a Heater "H-1" with outlet T = 308 K
- Add a Conversion Reactor "R-1" in isothermal mode using reaction set
  FermSet, ΔP = 0
- Add a Cooler "CL-1" with outlet T = 290 K
- Add a Separator Vessel "SEP-1" with vapor outlet "CO2-Vent" and
  liquid outlet "Beer"
- Connect: Feed → H-1 → R-1 → CL-1 → SEP-1
- Solve the flowsheet
- Report the mass flow and ethanol mass fraction of the Beer stream,
  the mass flow of CO2-Vent, and the overall mass yield (kg ethanol
  per kg glucose fed)

Exercises

  1. Reduce conversion to 80%. Beer ethanol concentration?
  2. Increase glucose feed to 30 wt%. (Real yeast has ethanol toxicity limits, but the model doesn't.)
  3. Add a Distillation Column after the separator to concentrate beer to 80% ethanol.

Further Reading

Selected references from the DWSIM technical bibliography. Click the DOI link to access each paper.

  • Jacques Monod. (1942). Recherches sur la Croissance des Cultures Bactériennes. Hermann et Cie
  • John F. Andrews. (1968). A Mathematical Model for the Continuous Culture of Microorganisms Utilizing Inhibitory Substrates. Biotechnology and Bioengineering. doi:10.1002/bit.260100602
  • Pauline M. Doran. (2013). Bioprocess Engineering Principles. Academic Press
  • James E. Bailey & David F. Ollis. (1986). Biochemical Engineering Fundamentals. McGraw-Hill
  • H. Renon & J. M. Prausnitz. (1968). Local Compositions in Thermodynamic Excess Functions for Liquid Mixtures. AICHE Journal

Next Steps

In Reverse Osmosis, you will model membrane desalination.