Engineering Conferences International
ECI Digital Archives
Integrated Continuous Biomanufacturing II
Proceedings
Fall 11-2-2015
Optimal control of a continuous bioreactor for
maximized beta-carotene production
Nazmul Karim
Texas A&M University, [email protected]
Jonathan Raftery
Texas A&M University
Xinghua Pan
Texas A&M University
Follow this and additional works at: http://dc.engconfintl.org/biomanufact_ii
Part of the Biomedical Engineering and Bioengineering Commons
Recommended Citation
Nazmul Karim, Jonathan Raftery, and Xinghua Pan, "Optimal control of a continuous bioreactor for maximized beta-carotene
production" in "Integrated Continuous Biomanufacturing II", Chetan Goudar, Amgen Inc. Suzanne Farid, University College London
Christopher Hwang, Genzyme-Sanofi Karol Lacki, Novo Nordisk Eds, ECI Symposium Series, (2015). http://dc.engconfintl.org/
biomanufact_ii/100
This Conference Proceeding is brought to you for free and open access by the Proceedings at ECI Digital Archives. It has been accepted for inclusion in
Integrated Continuous Biomanufacturing II by an authorized administrator of ECI Digital Archives. For more information, please contact
[email protected].
Optimal control of a continuous
bioreactor for maximized carotene
production
Jonathan P. Raftery and Dr. M. Nazmul Karim
Artie McFerrin Dept. of Chemical Engineering, Texas A&M University, College
Station, Texas 77843
𝛽-carotene Market
• An orange pigment produced by diverse organisms such
as plants, fungi, and bacteria
• Used in many industries:
–
–
–
–
–
Food
Animal nutrition
Pharmaceuticals
Cosmetics
Colorant
http://www.bellybytes.com/nourish/images/betacarotene.jpg
• Projected worldwide market value is $1.4 billion in 20191
• Higher antioxidant activity found in naturally produced 𝛽carotene when compared to the synthetic version
1Carotenoids
Market By Type (Astaxanthin, Beta-Carotene, Canthaxanthin, Lutein,
Lycopene, & Zeaxanthin), Source (Synthetic And Natural), Application (Supplements,
Food, Feed, And Cosmetics), & By Region - Global Trends & Forecasts To 2019.
MarketsandMarkets February 2015
𝛽-Carotene Production via
Recombinant Saccharomyces
cerevisiae
Continuous Production of
𝜷-Carotene
Recovered
Media
Pure Carotene Product
Glucose
and
Media
Media
Only
Pump
Media
Recovery
Pump
Media
Ethanol
Acetic Acid
Off Gases
Ethanol
Acetic Acid
Carotene
and
Solvent
Carotene
Purification
Unit
Solvent
Fermenter
Cell Separator
Carotene
Recovery
Unit
Cellular Waste
To Waste Treatment
Oxygen
Cells with
Carotene Product
Cell Disruptor
Fresh Solvent
Recovered Solvent
Liquid Waste
Solvent
Recovery
Outline
• Motivation
• Batch Operation of a Carotene Bioreactor
• Extension of Batch Operation to Continuous
Systems
• Model Predictive Control of a Continuous
Bioreactor to Maximize 𝛽-Carotene Production
• Results and Conclusions
Batch Process Overview
The aim of this study is to:
•
Develop a suitable and reliable kinetic model
for the carotene production in batch cultures
of an engineered Saccharomyces cerevisiae
strain using glucose as the main substrate
•
Apply this model to predict cell growth,
substrate consumption, ethanol and acetic
acid formation and later assimilation.
•
Determine the carotene productivity of the
batch system
Batch Operation of a
𝛽-carotene Bioreactor
Saccharomyces cerevisiae SM 141
Product
Beta-carotene
Reactor Working Volume
3 Liters
Initial Glucose
20 g/L
Air Flow Rate
6 L/min
Run Time
72 hours
Temperature
30℃
(controlled)
pH
4
(controlled)
X, G, P, E,
A
Volume
Air
8
160
7
140
6
120
5
100
4
80
3
60
2
40
1
20
0
0
0
10
20
30
40
Time (h)
50
60
70
80
Glucose (g/L) and Beta-carotene
(mg/L) Concentration
Biomass, Acetic acid, Ethanol
(g/L)
Organism Strain
1Reyes,
L.H., Gomez, J.M., and Kao, K.C.
Improving carotenoids production in yeast via
adaptive laboratory evolution. Metabololic
Engineering. 21 (2014) 26-33.
Unstructured Batch
Models
Growth Rate Models:
Batch Models:
𝜇 = 𝜇𝐺 + 𝜇𝐸 + 𝜇𝐴
dX
= rX = (μG +μE + μA ) X
dt
𝜇𝐺 =
𝜇𝑚𝑎𝑥,𝐺 ⋅ 𝜉𝐸 ⋅ 𝜉𝐴 ⋅ 𝐺
𝐾𝑆𝐺 + 𝐺 + 𝑎𝑔𝑒 𝐸 + 𝑎𝑔𝑎 𝐴
𝜇𝐸 =
𝜇𝑚𝑎𝑥,𝐸 𝐸
𝐾𝑆𝐸 + 𝐸 + 𝑎𝑒𝑔 𝐺 + 𝑎𝑒𝑎 𝐴
dG
μG X
= rG = −
dt
YX G
dE
μE X
= rE = k1 μG X −
dt
YX E
𝜇𝑚𝑎𝑥,𝐴 𝐴
𝜇𝐴 =
𝐾𝑆𝐴 + 𝐴 + 𝑎𝑎𝑔 𝐺 + 𝑎𝑎𝑒 𝐸
dA
μA X
= rA = k 2 μG + k 3 μE X −
dt
YX A
Inhibition Models:
dP
= rP = α1 μG + α2 μE + α3 μA ⋅ X + βX
dt
𝜉𝐸 = 𝑓 𝐸
𝜉𝐴 = 𝑓 𝐴
8
160
7
140
6
120
5
100
4
80
3
60
2
40
1
20
0
0
0
20
40
Time (hr)
60
80
Carotene Productivity
𝐦𝐠
𝐦𝐠
𝐋
= 𝟏. 𝟒𝟔
𝟕𝟐 + 𝐱 𝐡𝐫𝐬
𝐋 ⋅ 𝐡𝐫
𝟏𝟐𝟎
𝐚𝐬𝐬𝐮𝐦𝐢𝐧𝐠 𝐱 = 𝟏𝟎 𝐡𝐨𝐮𝐫𝐬 𝐟𝐨𝐫 𝐟𝐢𝐥𝐥𝐢𝐧𝐠, 𝐜𝐨𝐨𝐥𝐢𝐧𝐠, 𝐚𝐧𝐝 𝐜𝐥𝐞𝐚𝐧𝐢𝐧𝐠
Glucose (g/L) and Carotene (mg/L)
Concentration
Biomass, Ethanol, Acetic Acid
Concentration (g/L)
Estimation Procedure and
Results
Continuous Process
Overview
The aim of this study is to:
•
Propose a novel continuous carotene production
process utilizing a two tank system to allow for the
independent manipulation of the inlet flow rate and
inlet glucose composition
•
Develop continuous models describing the dynamic
nature of this novel fermentation system
•
Utilize dynamic optimization techniques to develop a
model predictive controller capable of maximizing
carotene production to rival that of batch production
processes
Continuous Operation of a
𝜷-carotene Bioreactor
Recovered
Media
Process
Glucose
and
Media
𝑮𝒇
Media
Only
Pump
Off Gases
𝑑𝐺 𝐹2
𝐹𝑜𝑢𝑡
= ⋅ 𝐺𝑓 −
⋅ 𝐺 − 𝑟𝐺
𝑑𝑡
𝑉
𝑉
𝑑𝑋
=−
⋅ 𝑋 + r𝑋
𝑑𝑡
𝑉
Ethanol
𝐅𝟏
𝑑𝑃
𝐹𝑜𝑢𝑡 Acetic Acid
Media
=
−
⋅ 𝑃 + 𝑟𝑃
Ethanol
𝑑𝑡
𝑉
Acetic Acid
𝑑𝐸
𝐹𝑜𝑢𝑡
=−
⋅ 𝐸 + 𝑟𝐸
𝑑𝑡
𝑉
Fermenter
Bioreactor
Cell Separator
𝐅𝐨𝐮𝐭
Oxygen
P
Media
𝐹𝑜𝑢𝑡
Recovery
Pump
𝐅𝟐
Models:
𝑑𝐴
𝐹𝑜𝑢𝑡
=−
⋅ 𝐴 + 𝑟𝐴
𝑑𝑡
𝑉
𝑑𝑉
= 𝐹𝐺+𝑀 + 𝐹𝑀 − 𝐹𝑜𝑢𝑡
𝑑𝑡
Cells with
Carotene Product
𝐹𝑜𝑢𝑡 = 𝑓Cell
𝑉 Disruptor
Ca
P
Carotene
and
Solvent
Carotene
Recovery
Unit
Optimization Algorithm
Starting at the beginning of continuous
operation characterized by the time
𝑡 = 𝑡𝑠𝑤𝑖𝑡𝑐ℎ = 20 hours:
1.
Discretize the process models for a
1
given Δt = hours with 𝑛 ≥ 10
n
2.
Solve the following optimization
problem for the optimum hourly
flowrate from each tank, 𝐹1 and 𝐹2 :
𝐦𝐢𝐧 −𝑷𝒕 + 𝜶
𝑭𝟏 𝑭𝟐
s. t.
3.
𝑽𝒕 − 𝑽𝒓
𝟐
𝒕
Discretized ODEs
0 ≤ 𝐹1,2 ≤ 𝐹𝑚𝑎𝑥
Repeat Step 2 for each hour after
𝑡𝑠𝑤𝑖𝑡𝑐ℎ , using the previous hour’s end
state as the new initial condition, to
determine the optimum flowrate as
the system moves forward in time
Steady State Analysis
Optimal Control Actions
0.18
35
0.16
30
0.14
25
Flowrate (L/hr)
Concentration (β-Carotene = mg/L, all else in g/L)
Concentration Profiles
20
15
10
0.12
0.1
0.08
0.06
0.04
5
0.02
0
0
200
220
240
200
Time (hours)
Glucose
Biomass
Ethanol
Acetic Acid
220
240
Time (hours)
Carotene
𝛽-Carotene
Media Only
Media + Glucose
Batch vs Continuous
Comparison
Batch Operation
𝜷-carotene Concentration
P = 120
Continuous Operation
𝛃-carotene Concentration
mg
L
Fermentation Time
t = 72 hours
P = 28.73
mg
L
Inlet Flowrate (𝐅𝐭𝐨𝐭𝐚𝐥 ) and Volume
Ftotal = 0.169
L
hr
V=3L
Productivity
Productivity
P
mg
= 1.46
t
L ⋅ hr
P⋅F
mg
= 1.62
V
L ⋅ hr
Continuous operation increases process productivity by
10.5% when compared to traditional batch processing.
Conclusions
• Continuous operation has the potential to increase
productivity of bioreactor systems
• A novel two-feed continuous reactor system capable
of independently varying the dilution rate and inlet
glucose concentration was implemented
• A bi-level dynamic optimization methodology was to
determine the maximum productivity of steady-state
continuous 𝛽-carotene production
• Continuous production shows a 10.5% increase in 𝛽carotene productivity compared to a tradition batch
system
Acknowledgements
Texas A&M Institute for Advanced Studies
National Science Foundation
Dwight Look College of Engineering
Graduate Teaching Fellowship
Karim Research Group
(M. Carolina Ordoñez-Franco, Tejasvi Jaladi, Melanie DeSessa)
Kao Research Group (S. cerevisiae SM14)