Process Simulation Cup (PSC2018) Increase the Profit of a ... · especially the metabolic reaction...

PSC2018_2_Details ©2018 by Chemstations Europe GmbH. All Rights Reserved.

www.chemstations.eu

Chemstations Europe GmbH

Process Simulation Cup (PSC2018)Increase the Profit of a Bacteria-Based PDO Production Facility

Part 2: Detailed Model Description


Please be aware of our copyright

1

• Ownership of copyright: The copyright for this website and the material on this website (including without limitation the text, computer code, flow sheets, artwork, photographs, images, music, audio material, video material and audio-visual material) is owned by Chemstations Europe GmbH or its affiliates (" Chemstations Europe") and its licensors.

• Copyright license: Chemstations Europe grants to you a worldwide non-exclusive royalty-free revocable license to:

– view this website and the material on this website on a computer or mobile device via a web browser;

– copy and store this website and the material on this website in your web browser cache memory; and

– copy and store the flow sheets on this website on a computer.

• Chemstations Europe does not grant you any other rights in relation to this website or the material on this website. In other words, all other rights are reserved.

• For the avoidance of doubt, you must not adapt, edit, change, transform, publish, republish, distribute, redistribute, broadcast, rebroadcast or show or play in public this website or the material on this website (in any form or media) without Chemstations Europe’s prior written permission.


Is this document for me?

• This document is for you, if you want to

– fully understand what was modeled in CHEMCAD

– learn how biotechnology processes can be modeled in a flow sheet simulator, especially the metabolic reaction network, the cross-flow filtration unit op, and the thermodynamic models

• This document

– contains all the main assumptions of the PSC2018 flow sheets

– shows the complete list of 15 design variables of the PSC2018

➢ some design variables may not yet be open for input by the PSC2018 participants

2


Outline

• Motivation

• Process Overview

• Cost Analysis

• Reactor Model

• Cross-Flow Filtration Model

• Thermodynamic Models

3


In Bio-Refineries bulk products for the chemical industry are produced from a renewable feedstock instead of using petroleum.

Motivation

4

©funfunphoto/Fotolia


1,3-Propanediol – A Versatile Molecule

• 1,3-Propanediol is an organic compound, a three-carbon diol, abbreviated to PDO

• It is part of several industrial products like

– Composites

– Adhesives

– Laminates

– Moldings

– Aliphatic polyesters

– Co-polyesters

5

• Or it can be used directly as

– Solvent (e.g. in wood paint) and

– Antifreeze


1,3-Propanediol – Usage as a Co-Polyester

• PDO can be synthesized by fermentation:

6

Terephthalic acid

Harvest Cane juice

Fermenter brothPDOPolymer (Product)

©MB/Fotolia ©utah778/Fotolia

©Hort/Fotolia©sumire8/Fotolia

©BEAUTYofLIFE/Fotolia


1,3-Propanediol – A Versatile BIO Molecule

• Some other PDO products:

7

Composite materials

Wood paint

Adhesive tapes

©Scisetti Alfio/Fotolia

©Saklakova/Fotolia

©benschonewille/Fotolia


Process Overview

8

©sittinan/Fotolia


Aerobic Fermentation of Cane Juice with E. Coli

9

1

1

Air

4

Off-Gas

3

Overflow

2

21

Cane Juice

22

Ammonia28

23

AceticAcid

2

E.Coli

©Maxim Suvorov/Fotolia


Metabolic Reaction Network - Overview

10

S + 4.5 O2 + NH3 + 2 A → 5 X + 6.5 H2O + 5 CO2

5 S + 4 O2 → 6 P + A + 4 H2O + 10 CO2

S + 6 O2 → 6 H2O + 6 CO2

X → DX

Growth

Production

Maintenance

DeathS: Substrate

A: Acetic acid

P: PDO

X: Biomass

DX: Dead biomass


Fed-Batch Fermentation

11

S

P


Fed-Batch Fermentation

12

SP

X

A

DX

N


• A battery of 3 bio-reactors is operated in parallel, each having a volume of 400 m³

• The cycle time is 30 hours

• Every 10 hours the content of one bio-reactor is drained to the intermediate storage tank

• From the intermediate storage tank a continuous stream is fed to the downstream process

Fermentation Section

13


Fermentation Section– Design Variables

14

Biomass growth and PDO productioncan be controlled by adjusting thefeed rates of

F01: Air

F02: Cane Juice (Substrate)

F03: Ammonia

F04: Acetic Acid

All of these feeds also represent costs!However, only reduce them with care,as you might run into a limitation.

And keep an eye on the liquid level ofthe reactor, as the overflow is purged!



15

Access designvariables



16


Downstream Process - Overview

17

In order to achieve a PDO purityof > 99.9 wt. % the downstreamprocess is separated into threesections:

1. Filtration (biomass removal)

2. Ion Exchange (salt removal)

3. Distillation (PDO purification)


Downstream Process - Filtration

18

Three cross-flow filtration units are used to remove biomass and cell debris

- Micro Filtration 1: Bulk removal of biomass

- Micro Filtration 2: Recovery of concentrate overflow from MF1

- Ultra Filtration 1: Removal of cell debris

MF1 MF2

UF1


Downstream Process - Filtration

19

Power consumption and separation efficiency depend on the transmembrane flux(TMF), the concentration of biomass X and cell debris DX, and on the feedtemperature. The design variables in this section are:

D01: Flowrate of dilution water (Stream 14)

D02: Split of dilution water to MF1 (Stream 15)

D03: Split of dilution water to MF2 (Stream 17)

D04: Pre-heating of dilution water to MF1 (Stream 16)

MF1 MF2

UF1


Downstream Process – Ion Exchange

20

The ion exchanger unit is used to removesalts and proteins. Only ammonia andacetic acid are included in the metabolicreactor model and thus they are the onlyspecies to form salts in the simulation.This is a necessary over-simplification.

The ion exchanger is modelled as areactor which is neutralizing these twodiluted electrolytes with caustic sodaand sulfuric acid.

Water is recycled from the distillationsection. The amount of this recyclestream 41 is the only design variable inthis section:

D05: Flowrate of recycled head productto the ion exchanger section (Stream 41)


Downstream Process - Distillation

21

The PDO is purified by oneevaporation and twodistillation steps.

The pressure is reduced andheat is added in order toremove surplus water in theevaporator EVAP.

In the first distillationcolumn DC1 PDO and waterare separated over headfrom rests of substrate andcontaminants.

In the second distillationcolumn DC2 the finalproduct purity is achieved.

EVAP

DC1

DC2


Downstream Process - Distillation

22

The design variables in thissection are

D06: EVAP pressure

D07: EVAP vapor fraction

D08: Condenser pressure

D09: Condenser temperature

D10: DC1 pressure

D11: DC1 reflux ratio

D12: DC1 bottom temp.

D13: DC2 pressure

D14: DC2 reflux ratio

D15: DC2 bottom temp.

EVAP

DC1

DC2


Downstream Process – Summary of Design Variables

23

Fifteen design variables areavailable in the downstreamprocess in order to maximizethe profit of the PDOproduction.


Downstream Process – Summary of Design Variables

24

The design variables can be set from the Excel table via Data Map:


Cost Analysis

25

©Dmitry Sunagatov/Fotolia


The production costs of the PDO process are split into different parts:

• Variable Costs

– Utilities

• Electricity

• Steam

• Cooling water

– Effluents

• Waste water

– Raw materials

• Feed (cane juice, initial reactor charge)

• Chemicals (acetic acid, ammonia, caustic soda, sulfuric acid)

• Fixed Costs

– Capital costs

– Labor

– Maintenance

– Marketing transport and sales

Cost Analysis

26


Cost Analysis

27

Electricity

Steam


Cost Analysis

28

Cooling water

Waste water


Cost Analysis

29

Raw materials

Fixed costs


Cost Analysis

30

The desired quality of PDO is 99.9 wt. %. A higher quality does not increase the salesprices. However, the PDO can be sold with a lower quality at a reduced price.

The following curve is used for calculation of the sales price:

Purity

Price


Cost Analysis

31

With the specific prices this finally gives the profit of the plant


Cost Analysis

32

The detailed cost analysis and the resulting profit can be found in the Excel table:


Reactor Model

33

©helenedeyun/Fotolia


Simulation Models - Metabolic Reaction Network

34

S + 4.5 O2 + NH3 + 2 A → 5 X + 6.5 H2O + 5 CO2

ሶ𝑟 = 𝜇𝐺,𝑚𝑎𝑥

𝑆

𝐾𝐺 + 𝑆𝑋 ⋅ 𝐿

YX/S = 5 mol/mol = 0.69 kg/kg

YX/A = 2.5 mol/mol = 1.04 kg/kg

YX/N = 5 mol/mol = 7.35 kg/kg

YX/O2 = 1.11 mol/mol = 0.87 kg/kg

S: Substrate

A: Acetic acid

P: PDO

O: dissolved Oxygen

N: dissolved Ammonia

L: Rate limiting factor

X: Biomass

DX: Dead biomass

𝐿 =𝑂

𝐾𝐿 + 𝑂⋅

𝑁

𝐾𝐿 +𝑁⋅

𝐴

𝐾𝐿 + 𝐴

Biomass growth:



35

5 S + 4 O2 → 6 P + A + 4 H2O + 10 CO2

ሶ𝑟 = 𝜇𝑃,𝑚𝑎𝑥

𝑆

𝐾𝑃 + 𝑆𝑋 ⋅ 𝐿

𝐿 =𝑂

𝐾𝐿 + 𝑂

S: Substrate

A: Acetic acid

P: PDO

O: dissolved Oxygen



X: Biomass

DX: Dead biomass

PDO production:

YP/S = 1.2 mol/mol = 0.51 kg/kg

YP/S = 1.5 mol/mol = 3.56 kg/kg

YA/Sof = 0.2 mol/mol = 0.07 kg/kg



36

S + 6 O2 → 6 H2O + 6 CO2

ሶ𝑟 = 𝑘𝑀 𝑋

S: Substrate

A: Acetic acid

P: PDO

O: dissolved Oxygen



X: Biomass

DX: Dead biomass

Biomass maintenance:



37

X → DX

ሶ𝑟 = 𝑘𝐷 𝑋

S: Substrate

A: Acetic acid

P: PDO

O: dissolved Oxygen



X: Biomass

DX: Dead biomass

Biomass decay:


A cross-flow filtration unit operation is mainly consuming electricity. This is used fordriving the pumps to produce the constant transmembrane flux. Commonly cross-flowfiltration units are operated continuous with scheduled cleaning intervals.

Simulation Models – Cross-Flow Filtration

39


In order to derive the electricity consumption a simple model was developed based on a characteristic Transmembrane Flux (TMF) over Transmembrane Pressure (TMP) curve:

This curve depends on specific membrane parameters, but also on the fluidtemperature, the particle concentration, and the particle size distribution of theretained solids.


40

Transmembrane Pressure

Tran

smem

bra

ne

Flu

x

No particles

Low particle concentration

High particle concentration


The concentration of the two solids present (X and DX) is calculated by using the Screen unit operation with the separation strength formula available in CHEMCAD. The particle size distributions of X and DX coming from the fermenter are fixed:


41


The flowsheet simulation is performed in flow-driven mode. The TMF is given and the TMP must be calculated. Thus the TMF-TMP curve must be inverted:


This curve shape allows multiple solutions for agiven TMF. In order to avoid this, the curve is cut:

TMP = 200 bar

42

Transmembrane Pressure

Tran

smem

bra

ne

Flu

x

Transm

emb

rane P

ressure

Transmembrane Flux


The following equations are used to calculate the TMP:

𝑇𝑀𝑃 = −𝑡𝑓

𝑡𝑐ln 1 −

𝑇𝑀𝐹

𝑇𝑀𝐹𝑚𝑎𝑥

𝑡𝑐 = 1 − exp −𝑇

𝑇𝑚𝑖𝑛

𝑇𝑀𝐹𝑚𝑎𝑥 = 𝑇𝑀𝐹𝑚𝑎𝑥0 − 𝑠𝑓𝑥 ⋅ 𝐶𝑥,𝐹 + 𝑠𝑓𝐷𝑥 ⋅ 𝐶𝐷𝑥,𝐹


43

Membrane parameters:𝑡𝑓 Temperature factor

𝑇𝑚𝑖𝑛 Recommended min. temperature

𝑇𝑀𝐹𝑚𝑎𝑥0 TMF with clean

membrane𝑠𝑓𝑥 Sludge factor for X𝑠𝑓𝐷𝑥 Sludge factor for DX

or TMP = 200 bar

Process parameters:𝑇 Feed temperature𝐶𝑥,𝐹 Concentration of

biomass in feed𝐶𝐷𝑥,𝐹 Concentration of

dead biomass in feed


𝑃𝐶𝐹𝐹𝐶 =𝑐𝑟 ⋅ ሶ𝑉𝐹𝑒𝑒𝑑 ⋅ 𝑇𝑀𝑃

𝜂𝑃𝑢𝑚𝑝


44

The power consumption is calculated with the equation

Cross-flow filtration parameters:

𝐴 Active membrane area𝑐𝑟 Cycle rate𝜂𝑃𝑢𝑚𝑝 Cycle pump efficiency



45

The parameters are specified in a specific screen and the model equations are evaluated in an Excel Sheet by using the CHEMCAD Excel Unit Operation.

Vapor fractions above 0 and TMF larger than TMFmax lead to a TMP of 200 bar!



46

The membrane performance curve can be evaluated in the same Excel table for different concentration factors:



47

The related Excel tables can be accessed via the DataMap Explorer:


Simulation Models – Distillation

49

The Modified UNIFAC model is used to calculate the vapor-liquid equilibria in thedownstream flowsheet. The following Group Assignment is used:

Latent heat is used as model for the Enthalpy. The coefficients of the related DIPRRequations are taken from the CHEMCAD databank.

The solubility of Ammonia, Carbon Dioxide, Nitrogen and Oxygen in water iscalculated with Henry’s Law.

Biomass, Acetic Acid, Caustic Soda and the resulting salts are removed before the distillation section and thus do not matter for the VLE calculations.



50

Relevant phase diagrams: SLE of Glucose – PDO (ideal)



51

Relevant phase diagrams: VLE of PDO - Glucose



52

Relevant phase diagrams: VLE of Water - PDO



53

Relevant phase diagrams: Residue curves of Water – PDO – Glucose



54

The distillation columns are modelled as equilibrium stage models with 8 stagesincluding condenser and reboiler. The feed is introduced in the middle of the columns.

The reflux ratio is used as condenser spec

The bottom temperature is used asreboiler spec

Process Simulation Cup (PSC2018) Increase the Profit of a ... · especially the metabolic reaction...

Documents

Transcript of Process Simulation Cup (PSC2018) Increase the Profit of a ... · especially the metabolic reaction...