Wider column. Same travel time.
Change the diameter of a packed column, then disturb its packing. See which quantities scale together and which do not.
Enlarge for bigger labels, then swipe sideways →
21.2 mm diameter gives 1.00 times the area. At fixed liquid speed, flow is 1.00× and load is 1.00×. Bed depth is 1.0×. The illustrative packing penalty gives 1.00× relative resolution; it does not predict a real column’s performance.
What this model assumes
Reference: 21.2 mm diameter, 20 mL/min and 500 µL injection. Liquid linear velocity, particle size and chemistry stay fixed. Flow scales with diameter squared. Load and solvent per run scale with bed volume in this illustrative translation; taller beds also increase retention time and pressure drop. Ideal resolution scales with the square root of plate count. Packing quality applies an explicitly illustrative quality-squared plate penalty; it is not a fitted correlation or a validated operating limit. Band drawings and Gaussian chromatogram peaks are qualitative and normalized to show retention and broadening, not detector response or recovered mass. The travel slider is a common normalized clock, not measured retention data. Solvent per unit load stays unchanged in the ideal translation; widening alone does not lower process mass intensity.
More width buys capacity. Packing still matters.
At fixed bed height and linear liquid velocity, flow scales with cross-sectional area. Double the diameter and the area, flow and proportional sample load become four times larger. The ideal travel time and pressure drop stay unchanged.
Gilson’s worked scale-up moves from 21.2 mm to 50 mm: a 5.56-fold increase in flow and injection volume. The published chromatogram comparison compensates for a tubing delay. The diagram here shows an idealized separation, not those measured chromatograms.
Lengthening the bed is a different change. At the same velocity it raises travel time and pressure drop. The model shows ideal square-root improvement in resolution with bed length; poor packing reduces that benefit.
The packing slider applies an illustrative plate-count penalty. It is not a correlation for any real resin, distributor or column. Wider equipment needs engineering of liquid distribution, wall effects and packing. Increasing capacity alone does not reduce solvent per unit of product.
The video’s 13,000 kg figure is an average total material input per kilogram of peptide, including water. About half belongs to purification and isolation. It is not a waste quantity, a universal design factor or an oligonucleotide value.
Double the diameter at the same bed height and linear velocity. What changes?
Taking a real process
beyond the lab?
Jon and I are building RARE Labs toward an autonomous pilot plant builder using AI and robotics. Join the waitlist for updates, and tell us what you’re trying to scale.
Go a little deeper.
- Gilson Applications LaboratoriesTN231: Prep HPLC Scale Up ↗
Flow and injection-volume scaling; 21.2 mm to 50 mm worked example, with tubing-delay correction in the measured overlay.
- CytivaColumn scale-up and packing ↗
Packing, wall support and distribution context. The interactive penalty is illustrative, not a fit to this note.
- Kekessie et al. (2024)Process mass intensity for peptide manufacturing ↗
Average inputs around 13,000 kg per kg peptide; purification and isolation contribute about half.