1. Three Manufacturing Modes
Every bioprocess operates in one of three fundamental modes: batch, fed-batch, or continuous (perfusion). Each represents a different trade-off between simplicity, productivity, and cost. The right choice depends on your product, your scale, and your organization’s manufacturing capabilities.
Understanding the economic and operational differences between these manufacturing approaches is essential for making informed process design decisions at any stage of development.
The biopharmaceutical industry has historically gravitated toward fed-batch as the standard manufacturing mode for monoclonal antibodies, but steady-state perfusion is gaining momentum as companies seek higher productivity from smaller facilities. Meanwhile, the traditional single-charge approach remains the simplest option and is still widely used for microbial fermentations, enzymes, and early-phase clinical manufacturing.
Batch = simplest, lowest risk, lowest productivity. Fed-batch = moderate complexity, highest titer, industry standard. Perfusion = highest complexity, highest volumetric productivity, smallest facility footprint. Each step up the complexity ladder offers higher output per unit volume but demands more process understanding and operational expertise.
Grouped bar chart with three categories on the X axis: Batch, Fed-Batch, and Perfusion. Two bars per category show bioreactor volume in liters and annual media cost in dollars. Batch requires 10,000 L and $800,000 media cost. Fed-Batch requires 2,000 L and $1,200,000 media cost. Perfusion requires 500 L and $2,500,000 media cost. Perfusion has the smallest bioreactor but the highest media cost.
2. Batch Mode
In this traditional mode, all nutrients are charged to the bioreactor at the start of the run. No additional substrates are added during the process (except acid/base for pH control and antifoam as needed). The culture grows until one or more nutrients are exhausted or metabolic by-products accumulate to inhibitory levels, at which point the entire contents are harvested.
Key Characteristics
- Cycle time: 3–14 days depending on organism (3–5 days for E. coli, 7–14 days for CHO)
- Typical titers: 1–3 g/L for mAbs; 5–20 g/L for microbial products
- Simplicity: No feeding strategy, no complex controls beyond pH, DO, and temperature
- Scale-up: Straightforward—same medium, same initial conditions, only vessel geometry changes
Advantages
- Lowest operational complexity and operator training requirements
- Easiest to characterize and validate (fewest process parameters)
- Lowest risk of contamination (no additions during the run beyond sterile gases)
- Well-suited to regulatory frameworks (the discrete run definition is unambiguous)
Limitations
- Low titers due to nutrient depletion and metabolite accumulation (ammonia, lactate)
- Low volumetric productivity: long turnaround times between runs (CIP/SIP) reduce annual output
- Substrate inhibition at high initial concentrations (e.g., glucose >10 g/L triggers overflow metabolism in CHO and E. coli)
Best for: enzymes and small molecules with high initial substrate tolerance, early-phase clinical manufacturing where speed to first production run matters more than yield, and processes where the organism naturally achieves adequate titers without feeding.
3. Fed-Batch Mode
Fed-batch extends the single-charge concept by adding concentrated nutrient feeds during the culture, avoiding both substrate depletion and substrate inhibition. This allows cultures to grow to much higher cell densities and produce for longer periods, dramatically increasing final titers. This feeding-based approach is the industry workhorse for monoclonal antibody manufacturing.
Key Characteristics
- Cycle time: 10–21 days for mAbs (14 days is typical); 1–5 days for microbial processes
- Typical titers: 5–10 g/L for mAbs (up to 15 g/L in best cases); 20–100+ g/L for microbial products
- Feed strategies: Bolus, continuous, exponential, pH-stat, or DO-stat feeding
- Volume increase: Working volume typically increases 20–40% during the run due to feed additions
Advantages
- Dramatically higher titers than the single-charge approach (3–5x for CHO, 5–20x for E. coli)
- Controlled nutrient environment avoids overflow metabolism (e.g., acetate in E. coli, lactate in CHO)
- Well-established regulatory framework—majority of approved biologics use this mode
- Moderate complexity increase over the conventional approach—mostly requires feeding pumps and a feed strategy
Limitations
- Feed strategy requires development and optimization (weeks to months of process development)
- Still a discrete process—requires full turnaround (CIP/SIP) between production runs
- Product quality can vary with cell age: late-harvest material may have different glycosylation or charge variant profiles
The industry has roughly doubled mAb fed-batch titers every 5–7 years: from <0.5 g/L in the late 1990s to 1–3 g/L by 2005, to 5–8 g/L by 2015, to 8–15 g/L in leading processes today. This titer improvement has been the single largest driver of COGS reduction over the past two decades, as detailed in our COGS guide.
Best for: monoclonal antibodies, Fc-fusion proteins, recombinant proteins for commercial manufacturing, high-cell-density microbial fermentations.
4. Continuous / Perfusion Mode
In perfusion mode, fresh medium is steadily added to the bioreactor while spent medium (containing the product) is steadily removed. A cell retention device—typically alternating tangential flow filtration (ATF) or tangential flow filtration (TFF)—keeps cells inside the bioreactor while allowing the product-containing permeate to pass through.
Key Characteristics
- Run duration: 30–90 days is typical; some processes run for 6+ months
- Perfusion rate: 1–3 vessel volumes per day (VVD)
- Cell density: 50–100+ million cells/mL (vs. 10–25 million for fed-batch)
- Instantaneous titer: 1–3 g/L in permeate (lower than fed-batch final titer)
- Volumetric productivity: 1–3 g/L/day (vs. 0.3–0.7 g/L/day for fed-batch)
Why 5–10× Higher Volumetric Productivity?
VPFB = Titer / Cycle Time = 8 g/L / 14 days = 0.57 g/L/day
Perfusion volumetric productivity:
VPperf = Titer × Perfusion Rate = 2 g/L × 1.5 VVD = 3.0 g/L/day
Ratio: 3.0 / 0.57 = 5.3× higher productivity per unit bioreactor volume
This productivity advantage translates directly to smaller bioreactors for the same annual output. A 500 L perfusion bioreactor can match the annual output of a 2,000–5,000 L fed-batch reactor, reducing capital cost and enabling single-use implementations at commercial scale.
Advantages
- 5–10x higher volumetric productivity than the fed-batch approach
- Smaller bioreactor for the same output—enables single-use at commercial scale
- Consistent product quality—steady-state operation produces uniform material throughout the run
- Gentle on the product—ongoing harvest means the product spends less time in the bioreactor, reducing degradation. Critical for unstable molecules (enzymes, Factor VIII, certain bispecifics)
- Lower capital cost—smaller facility footprint
Limitations
- High operational complexity—requires cell retention devices (ATF/TFF), cell bleed management, and ongoing harvest processing
- High media consumption—1–3 VVD means consuming 500–1,500 L/day of fresh medium for a 500 L reactor. Media cost per gram of product can be 2–3x higher than the fed-batch approach.
- Contamination risk—longer runs (30–90 days) increase the probability of contamination events. A single contamination can destroy an entire month’s production.
- Regulatory complexity—defining “a batch” requires careful thought (time-based pooling, marker-based pooling)
- Process development burden—more parameters to optimize (perfusion rate, cell bleed rate, cell retention)
A 500 L perfusion bioreactor at 2 VVD consumes 1,000 L/day of medium. Over a 60-day run, that is 60,000 L of medium. At $10/L, the media cost alone is $600,000 per run. Compare this to a single 2,000 L fed-batch run using ~2,800 L of medium at $28,000. The steady-state approach only wins economically when the productivity gain (g/L/day) more than offsets the higher media consumption.
5. Cost Comparison Table
The following table compares key parameters across all three manufacturing modes for a monoclonal antibody process targeting 100 kg/year annual output.
| Parameter | Batch | Fed-Batch | Continuous (Perfusion) |
|---|---|---|---|
| Titer | 1–3 g/L | 5–10 g/L | 1–3 g/L (in permeate) |
| Volumetric productivity | 0.1–0.3 g/L/day | 0.3–0.7 g/L/day | 1–3 g/L/day |
| Bioreactor size for 100 kg/yr | 10,000–15,000 L | 2,000–5,000 L | 200–500 L |
| Run duration | 3–14 days | 10–21 days | 30–90 days |
| Media cost per kg product | Low ($5–15K/kg) | Medium ($10–25K/kg) | High ($20–60K/kg) |
| Capital cost (facility) | High (large vessels) | Medium | Low (small vessels, SU) |
| Operational complexity | Low | Medium | High |
| Contamination risk | Low (short runs) | Low–Medium | Medium–High (long runs) |
| Product quality consistency | Variable batch-to-batch | Variable with culture age | Consistent (steady state) |
| Regulatory familiarity | High | Very high | Growing (multiple approvals) |
Worked Example: 100 kg/year mAb Production
6. When to Choose Each
A growing trend is “intensified fed-batch”—a hybrid that uses perfusion during the seed train (N-1 perfusion) to achieve very high inoculation densities (10–20 million cells/mL at inoculation vs. 0.3–0.5 million in standard processes). This shortens the production bioreactor run by 3–5 days, increases titer by 20–40%, and improves facility throughput without the complexity of full perfusion production. Many large pharma companies are adopting this as a near-term step toward continuous manufacturing.
7. The Industry Trend
The biopharmaceutical industry is moving toward intensified and steady-state processes, driven by several converging forces:
- Biosimilar pressure: With multiple blockbuster mAbs losing patent protection, biosimilar manufacturers need the lowest possible COGS. Perfusion-based processing with smaller facilities offers a route to sub-$50/g COGS.
- Facility flexibility: Single-use, smaller-footprint perfusion manufacturing facilities cost $30–$80M vs. $300–$500M for traditional stainless steel plants. This enables distributed manufacturing and multi-product facilities.
- Regulatory support: FDA and EMA have explicitly encouraged adoption of steady-state manufacturing. Several perfusion-produced biologics have been approved (e.g., ReFacto/Xyntha, Kogenate FS, and more recently, certain biosimilars).
- Technology maturation: ATF devices, high-density cell banking, and integrated downstream processing (multi-column chromatography) have matured from experimental to commercially validated.
However, the transition is gradual. Most companies are adopting a stepwise approach:
- Near-term: Intensified fed-batch with N-1 perfusion seed train (already being adopted widely)
- Mid-term: Full upstream perfusion with discrete or semi-integrated downstream
- Long-term: Fully integrated upstream and downstream processing in a single workflow
Model Your Process Economics
Compare the economics of all three manufacturing modes for your specific product using our free calculators.
For further reading:
- Fed-Batch Feeding Strategies — Optimize your fed-batch process with exponential, linear, and pH-stat feeding approaches.
- How to Estimate COGS per Gram — Detailed breakdown of all cost components in biopharmaceutical manufacturing.
References
- Konstantinov, K.B. & Cooney, C.L. (2015). “White paper on continuous bioprocessing. May 20–21, 2014. Continuous manufacturing symposium.” Journal of Pharmaceutical Sciences, 104(3), 813–820. doi:10.1002/jps.24268
- Walther, J. et al. (2015). “The business impact of an integrated continuous biomanufacturing platform for recombinant protein production.” Journal of Biotechnology, 213, 3–12. doi:10.1016/j.jbiotec.2015.05.010
- Bielser, J.-M. et al. (2018). “Perfusion mammalian cell culture for recombinant protein manufacturing — A critical review.” Biotechnology Advances, 36(4), 1328–1340. doi:10.1016/j.biotechadv.2018.01.011
- Kelley, B. (2009). “Industrialization of mAb production technology: The bioprocessing industry at a crossroads.” mAbs, 1(5), 443–452. doi:10.4161/mabs.1.5.9448