High-concentration biologics are now the norm, and most pump architectures were not designed for them. Push a pump toward a thick, concentrated feed and its behavior changes: the flow it promised on thin buffer sags, the reading wanders, and at some point it simply stops moving fluid. It has stalled, and the reason usually sits on a side of the pump nobody looks at.
Two trends push viscosity up. First, formulations are more concentrated: high-dose subcutaneous antibodies routinely reach protein concentrations where viscosity climbs steeply, and every concentration step in TFF ends thicker than it began. Second, real processes run cold and run crowded, and both temperature and crowding raise viscosity further.
The result is that the hardest fluid a pump has to move is often the most valuable one, at the very end of the process, where there is no margin to lose it.
The key concept: viscosity is not a fixed property of "the product." It rises with concentration and falls with temperature, so the same molecule can be easy to pump at harvest and punishing to pump at final formulation. Size the pump for the hardest point, not the average.
Viscosity raises the resistance to flow everywhere, but the failure almost always begins on the suction side. To deliver flow, a pump first has to fill. A thick fluid resists being drawn in: it cannot flow into the pumping chamber fast enough to keep up, so the chamber runs partially empty. Flow drops below setpoint, and if the pressure in that starved chamber falls far enough, the fluid can cavitate, doing damage as well as losing flow.
Different pump types hit this wall differently:
| Pump type | Behavior on a viscous feed |
|---|---|
| Peristaltic | The tubing has to spring back open to draw fluid in. A thick feed refills the tubing slowly, so delivered flow falls short and gets less accurate as viscosity rises. |
| Centrifugal | Efficiency drops sharply with viscosity; head and flow collapse, and these pumps are poorly suited to thick feeds to begin with. |
| Lifted-suction displacement | Any pump that has to lift fluid to fill will starve first, because viscosity makes that lift the binding constraint. |
| Flooded positive displacement | A positively displaced chamber fed by gravity keeps filling even when the fluid is thick, so flow holds and the stall condition is removed. |
Two design choices decide whether a pump copes with thick feeds. The first is positive displacement: moving a fixed volume per stroke rather than relying on velocity or vane efficiency, so delivered flow does not collapse as resistance climbs. The second, and the one most often missed, is how the pump fills. Give the inlet a flooded, gravity-fed supply and the suction-side starvation that causes viscous stalls never gets started.
Where Alphinity fits: the PIXER® pump is a positive-displacement diaphragm pump with a gravity-flooded, top-fed inlet, which is why it holds flow on demanding feeds and is rated to handle viscosities up to 3,000 cP. The architecture removes the suction-side limit that stalls other pumps on concentrated biologics.
The lesson for process design is to stop treating viscosity as an edge case. High-concentration feeds are now routine, they appear exactly where the product is most valuable, and the pump has to be chosen for that hardest point, not the easy thin-buffer start.
The failure usually starts on the suction side. A thick fluid cannot flow into the pumping chamber fast enough, so the chamber runs partially empty, delivered flow drops below setpoint, and the fluid can cavitate.
Formulations are more concentrated, such as high-dose antibodies and concentrated TFF retentate, and processes often run cold. Viscosity rises with concentration and falls with temperature, so the hardest fluid to pump is often the most valuable one at the end of the process.
Positive-displacement pumps with a flooded, gravity-fed inlet handle viscous feeds best, because they move a fixed volume per stroke and keep filling even when the fluid is thick. Peristaltic and centrifugal pumps sag and stall as viscosity rises.
Size for the hardest point in the process, not the average, and prioritize positive displacement and flooded suction so suction-side starvation never begins.
The suction-side physics behind stalls and cavitation, and the framework to tell each mechanism apart.
Why the industry default struggles with exactly the feeds that matter most.
Pumping a concentrated feed?
Speak to an engineer