Single-use bag bioreactors, 50 L to 2000 L

The disposable stirred-tank bag is the workhorse of mammalian cell manufacturing between pilot and commercial scale, and it stops at roughly 2000 litres. That ceiling is not a market preference; three independent limits converge there.

Why the ceiling sits where it does#

The first limit is mechanical. Hydrostatic pressure rises with liquid height, so the load on the film and, more importantly, on its welded seams grows as the bag gets taller. A polymer laminate a few hundred micrometres thick can carry that load only so far, and the failure mode is not gradual — a seam either holds or opens, taking the batch and the sterile boundary with it. The second is simply mass: two cubic metres of medium weighs about two tonnes, which the bag cannot support at all. It is carried by a rigid steel or polymer holder that the bag lines, so the supposed advantage of dispensing with a vessel is only ever partial, and the holder must be engineered for the full load.

The third is fluid mechanics. A bag is a smooth cylinder without baffles, usually with one bottom-mounted impeller driven through a magnetic or bearing coupling that limits torque. Without baffles, faster stirring converts into solid-body rotation and a surface vortex rather than into axial circulation, so power input cannot be raised indefinitely and the top-to-bottom mixing time lengthens with height. Against the framework in high cell density culture systems, the consequence is a lower achievable kLa than a comparable baffled steel tank, which is tolerable for cell culture and not for high-demand microbial work.

The qualification burden is chemical#

The real recurring cost of these systems is extractables and leachables work, described generally in single-use bioprocessing. What makes it a bioreactor problem rather than a documentation problem is that the culture is a living assay far more sensitive than most analytical methods.

The documented case is bis(2,4-di-tert-butylphenyl) phosphate, usually written bDtBPP. It is not an ingredient of the film; it is a degradation product of the phosphite antioxidant added to stabilise polyethylene, formed during gamma irradiation and storage. Leached into medium it inhibits the growth of Chinese hamster ovary cells at concentrations below one milligram per litre, and it was identified only after unexplained growth failures were traced back to particular film lots. The failure mode is the instructive part: the bag was sterile, intact and within specification, the medium was correct, and the process simply did not grow. Nothing on the control system showed anything.

That is why film qualification is lot-based rather than product-based, why users run cell-growth-based assays on new film lots rather than relying on chemical characterisation alone, and why a change of resin, additive package or irradiation dose by the supplier is a change control event for the drug maker. The single-use bag moved a variable that used to be inside the plant — the surface the culture touches — into a supply chain the plant does not control.