Engineering Guide · Vendor-Neutral

Continuous vs Batch Chromatography: Which Should You Pick?

Published June 23, 2026 · 14 min read
Continuous multi-column vs batch single-column chromatography side-by-side BATCH Single-column · sequential Harvest Protein A mAb Load → Wash → Elute → Regen Resin idle 40-60% of cycle ~50% utilisation VS CONTINUOUS Multi-column · simultaneous Feed LOAD LOAD 2 WASH ELUTE mAb Columns rotate · always producing Breakthrough captured downstream ~85% utilisation
Figure 1: Batch chromatography cycles one column through load-wash-elute-regenerate, leaving the resin idle most of the time. Continuous chromatography keeps 3-8 columns in different cycle positions simultaneously, so the resin is always producing.
Quick Verdict

Continuous chromatography (PCC, SMB, MCSGP) delivers 2-4x higher resin productivity and 30-60% less resin volume than batch single-column, but costs 2-3x more in capex and demands a stable feedstock. Pick continuous for dedicated high-titre mAb processes above 2 g/L on a 2-4 year horizon; pick batch for clinical-scale, multi-product CMO, vector products, and any process where the feedstock varies cycle-to-cycle. ICH Q13 (2022) makes both regulatory-acceptable.

Key differences at a glance

Side-by-side comparison

Factor Batch (single-column) Continuous (multi-column)
Operating principle Load-wash-elute-regenerate one column in series 3-8 columns in parallel positions, rotating cyclically
Resin utilisation 40-60% 75-90%
Productivity (g/L resin/h) 5-15 20-50
Resin volume for equivalent throughput Baseline 30-60% less
Buffer consumption Baseline (high wash + equilibration overhead) 30-50% less per gram product
Capital cost (process scale) 150-400k USD 500k-1.5M USD
Validation complexity Mature, batch-defined Higher, requires time/volume-based batch definition + PAT
Feedstock flexibility High, multi-product friendly Low, needs stable characterised feed
Footprint per kg/year output Baseline 40-60% smaller

Values reflect typical published specifications and pricing for Protein A mAb capture at 200-2000 L harvest scale. Your vendor's current quote takes precedence.

Batch chromatography in detail

Batch chromatography is the conventional single-column workflow that has dominated biopharma downstream processing since the first recombinant proteins reached the clinic in the 1980s. One column receives the entire harvest load, runs through wash, elute, and regenerate cycles, and is then ready for the next batch. Every commercial mAb approved before 2018 was made with batch chromatography, and most contract manufacturers still default to it.

How it works

A batch capture cycle proceeds in series: equilibrate the column with running buffer, load harvest broth until 1% breakthrough (a safety margin below the dynamic binding capacity), wash away non-bound impurities, elute the product with a pH or salt gradient, sanitise/strip with caustic, then re-equilibrate. The cycle takes 4-12 hours depending on column volume and load. Throughout, the resin is only productively binding product during the load step, perhaps 30-40% of the cycle, and idle the rest of the time. The standard hardware is an AKTA Avant 150 or AKTA ready-class skid driving a single column of 30-200 L resin, controlled by Unicorn software. Common platforms include the Cytiva AKTA process platform, Sartorius Resolute, and Pall's Cadence batch skids. For an entry into the discipline, our chromatography calculator sizes a single-column batch run from titre, harvest volume, and DBC.

When batch wins

Batch chromatography wins whenever flexibility beats efficiency. A CMO running 12 different molecules a year cannot afford to validate continuous skids per product, the changeover validation alone wipes out the productivity gain. Clinical-stage manufacturing under 100 L harvest scale rarely produces enough material for the continuous capex to pay back. AAV and lentivirus products are dose-defined per batch, so regulators expect a discrete batch identity that batch chromatography delivers natively. And when titre is below 1 g/L, the resin reduction simply does not move the needle on cost of goods.

Continuous chromatography in detail

Continuous chromatography is an umbrella term for multi-column approaches where the resin is always in use. The most common biopharma variant is Periodic Counter-Current (PCC) chromatography, 3-4 columns where the lead column loads past its 1% breakthrough point and the breakthrough is captured by the next column in series. Simulated Moving Bed (SMB) chromatography uses 4-8 columns in zones to simulate counter-current movement of the stationary phase. Multicolumn Countercurrent Solvent Gradient Purification (MCSGP) extends the concept to polishing steps with gradient elution. Sanofi's Cablivi (caplacizumab), Janssen's Darzalex Faspro, and several other approved biologics now use continuous capture commercially. Our in-depth continuous chromatography guide covers the SMB, PCC, CaptureSMB, and MCSGP variants in full.

How it works

In a 3-column PCC capture set-up, columns are arranged so that at any moment one column is loading, one is being washed, and one is being eluted or regenerated. As soon as the loading column reaches 1% breakthrough, the column at the wash position takes over the load role; the breakthrough that batch chromatography would discard as a safety margin is now caught by the column immediately downstream. The columns physically stay in place, what rotates is the buffer and feed routing through programmable valves. Commercial implementations include the Cytiva AKTA pcc (75 mm development and 6.0 mm process scale), the Sartorius BioSMB (process scale up to 8 columns), and the YMC Contichrom CUBE 30 (focused on MCSGP polishing). Aumann and Morbidelli (2007) demonstrated MCSGP achieving near-100% yield for calcitonin polishing where a single-column gradient capped at 66%, the foundational result that triggered industrial adoption.

When continuous wins

Continuous chromatography wins on three economic levers: resin saving (40-60% volume reduction at high titre), buffer reduction (30-50% lower buffer consumption per gram product), and facility productivity (40-60% smaller footprint per kg/year output). Pollock et al. (2013) modelled a dedicated 1000 L bioreactor commercial mAb process and found that semi-continuous Protein A capture broke even on capex against batch within 18-30 months at titres above 5 g/L. Mahajan et al. (2012) showed equivalent purity and HCP clearance with 30-40% less Protein A resin. The case sharpens further when the upstream is perfusion: continuous capture eliminates harvest hold tanks, surge volume, and batch-by-batch reconciliation. Repligen, Sartorius, and Cytiva all now market fully continuous upstream-plus-downstream architectures combining ATF or TFF cell retention with PCC capture.

Pros and cons

Batch chromatography

Advantages

  • Lowest capex per skid (150-400k USD process scale)
  • Mature validation pathway, batch identity is regulatory shorthand
  • Flexible across multiple molecules, changeover is hours not weeks
  • Lower PAT and process control demands
  • Ubiquitous operator skill base, no specialised training

Disadvantages

  • Resin utilisation 40-60%, most of the resin is idle most of the time
  • Productivity capped at 5-15 g/L resin/h
  • Buffer consumption 30-50% higher per gram product
  • Loading limited to 1% breakthrough, wasted resin safety margin
  • Larger column footprint per kg/year output

Continuous chromatography

Advantages

  • 2-4x higher productivity (20-50 g/L resin/h)
  • 30-60% less resin volume, Protein A savings 400-600k USD per resin cycle at 100 L scale
  • Loads past 1% breakthrough, closer to dynamic binding capacity
  • 30-50% lower buffer consumption
  • Natural pairing with perfusion, eliminates harvest hold tanks

Disadvantages

  • Skid capex 2-3x higher (500k-1.5M USD)
  • Requires stable characterised feedstock, multi-product changeover is costly
  • Heavier validation burden, time- or volume-based batch definition + PAT
  • Higher operator training cost, lower skill-base availability
  • More complex troubleshooting, column-to-column variability matters

Which should you choose?

Pick based on the dominant constraint in your process. Most chromatography decisions come down to four: titre and modality, facility model (dedicated vs CMO), production schedule (clinical vs commercial), and upstream architecture (fed-batch vs perfusion).

Clinical-stage mAb, single product

Phase 1/2 manufacturing under 200 L harvest scale. Resin saving is too small to repay 1M USD of capex; validation cycles are too short.

Choose Batch

Commercial high-titre mAb, dedicated line

Titre 4-8 g/L, 1000-2000 L bioreactor, dedicated facility, multi-year campaigns. Resin savings of 400-800k USD per cycle plus buffer reduction repay capex in 18-36 months.

Choose Continuous

Multi-product CMO facility

12+ molecules per year, variable titre, variable harvest volume. Changeover validation per product erases the continuous productivity gain.

Choose Batch

Perfusion-fed dedicated process

ATF or TFF cell retention producing 1-2 reactor volumes/day. Continuous capture eliminates surge tanks, fits the steady feed pattern, doubles facility productivity.

Choose Continuous

Real-world use cases

Typical setups where bioprocess teams have converged on one choice or the other.

CHO mAb, 2000 L fed-batch, 5 g/L
3-column AKTA pcc Protein A capture

40% less MabSelect SuRe resin (60 L vs 100 L batch), 35% less wash buffer, payback 22 months. Common in dedicated mAb sites at Sanofi, Roche, and Lilly.

E. coli microbial fed-batch
Single-column batch IEX + HIC

Inclusion-body refolding pools are too variable in titre and impurity profile to justify continuous. Batch single-column dominates microbial DSP today.

AAV viral vector, HEK293 transient
Single-column AEX polishing

Per-batch dose definition plus small lot sizes (1-100 L harvest) keep batch as the default. CaptureSMB pilots exist but not yet commercial.

Perfusion mAb, 200 L bioreactor
PCC Protein A + flow-through AEX

ATF cell retention feeds the AKTA pcc directly. No harvest tank. 60% smaller chromatography footprint. Used in intensified platforms at CMC Biologics and AGC Biologics.

Need to size a chromatography column before you decide?

Our chromatography calculator sizes a single-column batch run from titre, harvest volume, dynamic binding capacity, and column geometry. It gives you the baseline resin volume that continuous reduces by 30-60%.

Open the Chromatography Calculator

Cost and lifecycle considerations

Three components drive chromatography TCO

Capex (skid + columns) + resin (10-15k USD/L for Protein A, 2-4k USD/L for IEX and mixed-mode, replaced every 100-300 cycles) + opex (buffers, labour, validation, facility allocation). Batch wins on capex; continuous wins on resin and opex. The crossover depends on titre, scale, and how many years you operate before the resin needs replacing.

For a representative 1000 L fed-batch CHO mAb process at 4 g/L titre running 50 batches per year, the published economic models (Pollock et al. 2013; Steinebach et al. 2016) converge on a 2-3 year payback for continuous capture over batch. The dominant saving is Protein A resin volume: batch needs ~120 L of MabSelect SuRe per cycle pool (~1.4M USD), continuous needs 70-80 L (~0.9M USD). With resin replaced every 200 cycles, that saving recurs every 4 years.

At lower titre (1-2 g/L) or smaller scale (200 L harvest), the resin saving shrinks below the capex premium and batch chromatography stays economical. At higher titre (>5 g/L) or larger scale (>2000 L), continuous starts to pay back within the first year of operation. The crossover table below uses public list prices and typical industry assumptions; your vendor's quote will vary.

Cost component (1000 L, 4 g/L mAb) Batch (single-column) Continuous (3-column PCC)
Skid capex250k USD800k USD
Protein A resin (year 1, 120 L vs 75 L)1.44M USD0.90M USD
Buffer consumption / year280k USD180k USD
Labour + validation / year120k USD180k USD
3-year TCO estimate2.85M USD2.34M USD

Vendor landscape

Major vendors in each camp, with one-line positioning notes.

Batch chromatography platforms

Continuous chromatography platforms

Frequently asked questions

What is the difference between continuous and batch chromatography?
Batch (single-column) chromatography runs one column through a load-wash-elute-regenerate cycle, then repeats, the column sits idle during the non-loading phases, so average resin utilisation is 40-60%. Continuous chromatography uses 3-8 interconnected columns that rotate through the cycle positions in sequence, with breakthrough from the loading column captured by the next downstream column. This drives resin utilisation up to 75-90%, productivity 2-4x higher (g product per L resin per hour), and total resin volume down 30-60% for equivalent throughput. The most common variants are Periodic Counter-Current (PCC) and CaptureSMB for capture steps, and Multicolumn Countercurrent Solvent Gradient Purification (MCSGP) for polishing.
Is continuous chromatography always better than batch?
No. Continuous wins on resin utilisation, productivity, and buffer consumption, but it costs more upfront (multi-pump, multi-valve skids run 500k-1.5M USD vs 150-400k USD for a comparable batch skid), demands a stable feedstock to keep the column rotation in balance, and requires more sophisticated PAT and control. Batch chromatography is still the right answer for clinical-scale (single batches under 100 L harvest), multi-product CMO facilities with variable feedstocks, low-titre molecules where resin reduction does not pay back the skid cost, and vector products where batch-defined dosing is regulatory shorthand. The break-even is usually 2-4 years for dedicated high-titre mAb processes.
How much resin does continuous chromatography save?
Mahajan et al. (2012) reported 30-40% reduction in Protein A resin volume when moving a 2 g/L mAb capture from single-column batch to a 3-column semi-continuous system at the same throughput. Pollock et al. (2013) showed 40-60% reduction at 5 g/L titres, where the resin saving is large enough to offset the higher capex within 18-30 months. Protein A resin lists at 10-15k USD per litre, so a 100 L production-scale column carries 1-1.5M USD of resin, and a 40% reduction is 400-600k USD recovered per cycle of resin replacement. The saving is smaller for ion-exchange and mixed-mode resins (typically 2-4k USD/L), so the economic case for continuous is strongest on Protein A capture.
What is PCC chromatography and how does it differ from SMB?
Periodic Counter-Current (PCC) chromatography uses 3-4 columns that switch position on a periodic time-based schedule. Load proceeds on the lead column past its 1% breakthrough point; the breakthrough is captured by the next column in series, eliminating the resin-wasting safety margin of batch operation. Simulated Moving Bed (SMB) chromatography typically uses 4-8 columns arranged in zones (adsorption, desorption, separation) with continuous feed and continuous product withdrawal, simulating a counter-current movement of the stationary phase. PCC is simpler to control and dominant for biopharma capture; SMB is more efficient theoretically but mostly used in petrochemical and small-molecule separations. CaptureSMB is a 2-3 column hybrid variant optimised for biopharma capture.
Does FDA accept continuous chromatography for GMP manufacturing?
Yes. ICH Q13 (Continuous Manufacturing of Drug Substances and Drug Products, finalised November 2022) gives the global regulatory framework, and FDA, EMA, and PMDA all signed off. Continuous chromatography is in commercial use for several approved biologics, Sanofi's Cablivi process and Janssen's Darzalex Faspro Protein A step are public examples. The validation work is heavier than batch because the definition of a "batch" must be set up front (typically time-based, e.g., 24 h or 7 days of continuous operation), and steady-state monitoring requires PAT (UV at multiple positions, conductivity, pH, sometimes Raman). Once validated, the regulatory bar is not higher than batch.
How does continuous chromatography fit with perfusion bioreactors?
It is a natural pairing. Perfusion produces a steady, low-volume harvest stream (typically 1-2 reactor volumes per day at 30-100 million cells/mL), which is exactly the feed pattern continuous capture is designed for. A perfusion bioreactor feeding a PCC Protein A skid eliminates the need for hold tanks, surge volume, and batch-by-batch reconciliation. ATF or TFF cell retention plus PCC is the canonical "fully continuous" upstream-plus-downstream architecture pursued by Sartorius, Cytiva, and Repligen. Fed-batch processes can also feed continuous chromatography, but require a surge tank between the harvest centrifuge or depth filter and the continuous skid to smooth the volumetric pulse.
What does a continuous chromatography skid cost?
Process-scale continuous chromatography systems list between 500k and 1.5M USD depending on column count, scale, and automation tier. Cytiva AKTA pcc 75 (development/pilot) is in the 350-500k USD bracket; AKTA pcc 6.0 mm process-scale runs 600k-900k USD. Sartorius BioSMB Process targets the 700k-1.2M USD range. YMC Contichrom CUBE 30 (development scale, MCSGP focus) is 300-450k USD. A comparable single-column AKTA Avant or BioProcess skid lists at 150-400k USD, so the capex premium for continuous is roughly 2-3x. The premium pays back through resin savings, buffer reduction, smaller footprint, and higher facility productivity, typically within 2-4 years on dedicated high-titre processes.
When should you not use continuous chromatography?
Avoid continuous chromatography when (1) the feedstock is variable or multi-product, campaign changeovers wipe out the productivity benefit; (2) the product is a vector or other batch-defined dose where regulators expect a discrete "batch" identity; (3) total annual demand is low (under 10 kg per year), so the capex premium never pays back; (4) the molecule has low titre (under 1 g/L), so the resin saving is small; (5) downstream PAT and process control infrastructure is immature, continuous fails without real-time breakthrough monitoring. Batch chromatography is still the dominant choice for clinical-stage manufacturing, contract manufacturers (CMOs) running multi-product campaigns, and AAV/lentivirus production.

Resources and references