CHE 562: Advanced Bioprocess Engineering
Estimated study time: 9 minutes
Table of contents
Sources and References
Primary texts — Doran, P.M., Bioprocess Engineering Principles, 2nd ed., Academic Press, 2013; Villadsen, J., Nielsen, J., and Lidén, G., Bioreaction Engineering Principles, 3rd ed., Springer, 2011.
Supplementary texts — Blanch, H.W. and Clark, D.S., Biochemical Engineering, 2nd ed., CRC Press, 1997; Nienow, A.W., Reactor Engineering in Large-Scale Animal Cell Culture, review series.
Online resources — MIT OCW 10.28 “Chemical-Biological Engineering Laboratory”; NPTEL “Bioprocess Engineering”; ISPE baseline pharmaceutical engineering guides public summaries; ASME BPE standard public documents.
Chapter 1: Large-Scale Fermentation
1.1 Industrial Fermentors
Stirred-tank fermentors from 10 L (pilot) to 500 m\(^3\) (industrial antibiotics). Design comprises vessel (typically SS316L, GMP-finished), agitation (Rushton, pitched-blade, Intermig, Scaba for high viscosity), sparger (ring, porous, micro), baffles (to prevent vortex), cooling coils or jacket, sample ports, sensor probes, sterilization-in-place (SIP) and cleaning-in-place (CIP) infrastructure.
Alternative geometries: bubble column (no agitation, gas-driven mixing), airlift (internal or external loop, low shear), fluidized bed (immobilized cells/enzymes), packed bed, rotating-disc contactor.
1.2 Sterilization
Batch sterilization of media: heat to 121 °C with steam, hold, cool. The Del factor \( \nabla = \ln(N_0/N_f) \) quantifies required decimal reductions. For microbial destruction following first-order kinetics with Arrhenius temperature dependence:
\[ \nabla = \int_0^{t_f} k_d e^{-E_a/RT(t)} dt. \]Continuous sterilization (high-temperature short-time, HTST) minimizes thermal damage to heat-labile nutrients (vitamins, amino acids). Filter sterilization with 0.22 µm cartridges handles non-autoclavable components.
1.3 Media and Feed Design
Defined media (chemically specified) for reproducibility; complex media (yeast extract, peptones, corn steep liquor) for cost and rich supplementation. Carbon source (glucose, glycerol, molasses), nitrogen (NH\(_4\)\(^+\), amino acids), phosphate, sulfate, trace metals, vitamins. Fed-batch feeding strategy (exponential, DO-stat, pH-stat, substrate-limited) controls specific growth rate to maximize product.
Chapter 2: Mass Transfer with Biological Constraints
2.1 Oxygen Transfer
\( OTR = k_L a (C^* - C_L) \), balanced by cellular OUR \( = q_{O_2} X \). Steady-state \( C_L \) solves
\[ k_L a (C^* - C_L) = q_{O_2} X. \]Critical dissolved oxygen concentration \( C_{crit} \) (typically 5–20% saturation) separates aerobic-limited from kinetically-limited regimes.
2.2 \( k_L a \) Correlations
Van’t Riet for Newtonian broth:
\[ k_L a = 0.026 \left(\frac{P_g}{V}\right)^{0.4} v_s^{0.5}. \]Corrections for non-Newtonian rheology (shear-thinning power-law fluids), coalescing/non-coalescing media (salt effects), and surfactants. For mycelial fermentations (penicillin), apparent viscosity scales with biomass and morphology; correlations specific to filamentous cultures apply.
2.3 Shear and Cell Damage
Mammalian cells are shear-sensitive; bubble burst at the interface produces microscale forces that damage membranes. Pluronic F-68 surfactant, low-shear impellers (pitched-blade, Rushton with larger D/T), and bubble size control mitigate damage. Insect cells and hybridomas have similar sensitivities; microbial cells (robust walls) tolerate much higher shear.
2.4 Power Input
Ungassed: \( P_0 = N_p \rho N^3 D^5 \). Gassed: \( P_g/P_0 \) decreases with aeration number \( N_A = Q_g/(ND^3) \), typically 0.4-0.7 at process conditions. Total mechanical energy is the shaft work; isothermal compression of inlet gas adds a second contribution. Total energy demand frames operating cost.
Chapter 3: Heat Transfer in Fermentors
3.1 Heat Generation
Metabolic heat: \( Q_{met} = Y_\Delta X \mu V \) where \( Y_\Delta \) is the heat yield per unit biomass (roughly 12-18 kJ/g for aerobic growth on glucose). Agitation adds shaft-work heating \( P_g \). Total heat load for a 50 m\(^3\) E. coli culture at peak rate: typically 100-300 kW.
3.2 Heat Removal
Jackets limited by vessel surface area (scales as \( V^{2/3} \), so falls short at large scale). Internal coils (greater surface) reduce mixing effectiveness if poorly arranged. External pump-around coolers circulate broth through a plate or shell-and-tube exchanger—preferred at 50 m\(^3\)+.
Heat balance at steady state:
\[ U A \Delta T_{lm} = Q_{met} + P_g. \]3.3 Thermal Control Challenges
Fed-batch overheating scenarios: peak glucose uptake, maximum cell density, minimum margin on cooling water temperature in summer. Validation of thermal design against worst-case plus margin is routine industrial practice.
Chapter 4: Rheology
4.1 Newtonian and Non-Newtonian Broths
Single cells in water form Newtonian broths up to moderate cell density (20-50 g/L DCW). Mycelial (filamentous fungi, streptomyces) and extracellular polysaccharide producers (xanthan, alginate, hyaluronic acid) produce strongly non-Newtonian, shear-thinning broth: power-law (Ostwald-de Waele),
\[ \tau = K \dot\gamma^n, \]with flow behavior index \( n < 1 \) (shear-thinning). Apparent viscosity \( \mu_{app} = K\dot\gamma^{n-1} \); 10-1000-fold variation between quiescent and impeller-tip regions.
4.2 Impact on Transport
Shear-thinning broth exhibits “caverns” near impellers (vigorously mixed) with stagnant regions beyond. Oxygen, carbon, and temperature gradients can appear within a single tank. Specialized impellers (Scaba, Chemineer Maxflo) and axial flow patterns (down-pumping, hydrofoil) improve bulk circulation.
4.3 Rheological Characterization
On-line rheology by impeller torque (Metzner-Otto correlation): \( \dot\gamma_{avg} = k_s N \), \( k_s \approx 10-13 \) for turbulent flow. Off-line by rotational viscometer; care required because mycelial pellets are non-homogeneous.
Chapter 5: Downstream Processing Integration
5.1 Harvest and Clarification
Continuous disc-stack centrifuge (CHO, Pichia), tangential flow microfiltration (E. coli, yeast), depth filtration (polishing). Selection depends on cell type, titer, product location (intracellular vs secreted).
5.2 Capture Chromatography
Protein A affinity for monoclonal antibodies: load crude harvest, wash, elute at low pH; typical yield 90%+ in a single step with 95%+ purity. Ion exchange (S-Sepharose, Q-Sepharose) and mixed-mode resins provide alternatives or polishing steps. Column productivity (g product / L resin / hour) directly drives capital and operating cost.
5.3 Integrated and Continuous Bioprocessing
Perfusion culture feeding continuous capture (MCSGP, periodic countercurrent chromatography) reduces facility footprint, buffer volumes, and tank sizes. Ongoing trend from batch to continuous bioprocess mirrors petrochemical evolution of the mid-20th century.
5.4 Virus Safety
Mammalian cell culture risks adventitious viruses (porcine parvovirus, reoviruses). Orthogonal removal steps: low-pH hold, detergent inactivation, nanofiltration (20 nm), chromatographic removal (multimodal, heparin). Regulatory expectation (ICH Q5A): two orthogonal steps demonstrating > 4 log\(_{10}\) viral clearance.
Chapter 6: Scale-Up and Process Control
6.1 Scale-Up Criteria
No single criterion universal; common choices:
- Constant \( P/V \): preserves \( k_L a \) for aerobic bacterial fermentations.
- Constant tip speed (\( \pi N D \)): limits shear for mammalian cells.
- Constant mixing time (\( t_m \propto 1/N \)): maintains homogeneity.
- Constant impeller Reynolds number: rarely achievable.
Industrial practice prioritizes one criterion while monitoring the others through pilot studies.
6.2 Process Analytical Technology (PAT)
Real-time or near-real-time measurement of critical quality attributes: Raman spectroscopy (nutrients, metabolites, cell density), NIR, dielectric spectroscopy (viable biomass), off-gas mass spec (respiration quotient, biomass-independent metabolism). Soft sensors estimate unmeasured states; model-based control enables tight trajectory tracking of fed-batch profiles.
6.3 Quality by Design
ICH Q8 paradigm: define target product profile, identify critical quality attributes, map to critical process parameters via design of experiments, establish design space, control strategy. Quality is built in, not tested at end.
6.4 Advanced Topics
Intensified processes: high-cell-density perfusion, single-use bioreactors (disposables up to 2000 L), modular facilities. Synthetic biology integration: chassis strains, genome-scale models, optimal pathway engineering. Digital twins: validated models running alongside plants for optimization and anomaly detection.
