A formula approved in a one kilogram beaker can behave differently in a five hundred kilogram vessel. The chemist sees it first as a small disappointment: the texture is slightly thinner, the colour a shade warmer, the foam a little slower to build. Nobody made an obvious mistake. The process simply changed the environment in which the formula has to work.
Scale-up is the discipline of finding those differences before they reach a filling line & a launch date. It sits between development & production, & it is where most schedule slips actually originate, because the gap between a laboratory batch & a commercial batch is discovered late rather than early. This piece explains where the drift comes from, how to size a pilot run, & what to measure before committing to volume.
Why a lab batch is not a small production batch
Laboratory equipment delivers energy differently from plant equipment. A small overhead stirrer produces high shear in a small volume; a large vessel with a different impeller produces a different shear profile even when the recipe is identical. Heating & cooling follow different curves because surface area to volume ratios change, & a vessel that takes forty minutes to cool will hold a heat-sensitive ingredient at temperature far longer than the beaker did.
Time matters as much as equipment. A lab batch may be mixed, cooled & filled inside an hour. A production batch may sit through transfer, holding & quality checks across a shift. Anything that is sensitive to time at temperature, to shear history or to order of addition will notice the difference.
Where the drift usually appears
Texture is the most common surprise. Emulsions thicken differently when the shear profile changes, & rheology modifiers hydrate at different rates depending on how the water phase is handled. A formula that sat at the right viscosity in the lab can arrive thin, or grainy, or with a slightly different yield value that shows up as poor suspension.
Colour & odour follow. Longer heating shifts some botanical extracts, & a warmer hold can deepen a tint or flatten a top note. Neither is a defect in itself, but both change what the customer receives relative to the approved sample.
Order of addition & hydration time
The written method is part of the formula, not an optional instruction. If a gum is added to the water phase before the humectant in the lab & after it in the plant, hydration can be incomplete & the batch will thicken slowly over the following days. That late thickening is often read as instability when it is really a procedural difference.
Write the method as a sequence with times & temperatures, & require the plant to follow it during the pilot. Where plant constraints force a change, record it, because an unrecorded change is the hardest kind of problem to trace later.
Heat, cooling & shear history
Heat-sensitive materials should be added at the temperature the development work proved, & the cooling curve should be part of the specification rather than a matter of convenience. A fast cool can trap structure; a slow cool can let a wax system reorganise into a different crystal form, which changes feel rather than composition.
Shear history deserves the same attention. Homogeniser passes, recirculation time & pump type all leave a signature. When the pilot result differs from the lab, matching the shear history is usually the first adjustment worth trying.
Sizing a pilot run
A pilot run should be large enough to use the same equipment class as production & small enough to be affordable if it fails. In practice that often means a batch that fills the intended vessel to a level where the impeller & the heat transfer behave as they will at full scale, rather than a token quantity run through plant equipment.
Ask the factory what the vessel's working range is, because mixing below a minimum level behaves like a different process entirely. A pilot that runs at twenty per cent of vessel capacity will not predict the full batch, however carefully it is measured.
What to measure during the pilot
Measure the things that will decide acceptance at volume: viscosity at a stated temperature, appearance against the approved standard, odour against the retained sample, pH where relevant, fill weight across the run, & any property specific to the format such as foam height for a cleanser or payoff for a stick.
Record measurements at defined points rather than once at the end. Start-up, middle & end-of-run readings show whether the batch is stable through the run or drifting, & drifting matters because the first & last units are the ones most likely to differ from the approved sample.
Retention samples & the observation window
Keep retention samples from the pilot & hold them under the same conditions as production retention samples. They are the reference if the market later reports a difference, & they make a technical conversation concrete rather than speculative.
Set an observation window before the launch, not during it. A few weeks of controlled observation on the pilot batch will reveal texture movement, separation or odour drift while there is still time to adjust the process.
Fill behaviour at line speed
Formulas that pour beautifully at the bench can behave badly at line speed. Thicker systems trap air, foaming systems flood the nozzle, & shear-thinning systems recover at a rate that affects how the product settles in the pack. All of these show up as fill weight variance rather than as a laboratory failure.
Run the pilot on the intended filler, or on a filler of the same class, & weigh units from the start, middle & end. Fill weight data from a pilot is the cheapest predictor of complaints about underfilled packs.
Packaging interaction at scale
Contact time & surface area change at volume, & so does the opportunity for interaction. A formula that was stable in a glass beaker can pick up a different odour profile from a closure liner, or lose a little water through a closure that was fine in a short laboratory test.
Include the final pack in the pilot. Filling pilot units into the commercial container, closing with the commercial closure & holding them through the observation window is the only way to see the product as the customer will.
Acceptance criteria before the main run
Acceptance criteria turn opinion into a decision. Write them as ranges: viscosity between two values at a stated temperature, appearance matching the approved standard within an agreed tolerance, odour comparable to the retained sample, fill weight within the average-content rule, no separation after the observation window.
Agree who signs them & what happens when a criterion is missed. If the pilot fails a criterion, the options are adjusting the process, adjusting the formula, or adjusting the criterion with a documented reason. All three are legitimate; only the undocumented version is dangerous.
Documentation that should come out of scale-up
The output of scale-up is a manufacturing instruction, not just an approved batch. It should record vessel class, impeller type, order of addition, times & temperatures, homogeniser settings, cooling profile, filler class, target fill weight & the acceptance ranges. That document is what makes the second batch resemble the first.
It should also record what was tried & rejected. Negative results are cheap to write down & expensive to rediscover, particularly when a different team runs the next campaign.
Planning the schedule around scale-up
Scale-up needs calendar room. A pilot batch, an observation window, an adjustment & a confirmation batch cannot be compressed into the week before a launch date without removing one of them, & the one removed is usually the observation window.
Book the pilot slot when the formula is close to approval rather than after it. Most factories can prepare vessel time & filler access well in advance, & waiting for final sign-off is the most common avoidable delay in this stage.
When to go back to development
Some results are not process problems. If the pilot cannot hit the target texture without a change that damages the sensory profile, or if the formula only works within a processing window the plant cannot hold reliably, the answer is a formulation change rather than a heroic process fix.
That is a normal outcome & it is cheaper than launching a product that depends on conditions nobody can repeat every time.
| Pilot checkpoint | What to record | When measured | Typical follow-up if missed |
|---|---|---|---|
| Viscosity | Value at stated temperature with method | Start, middle, end of run | Adjust shear or hydration time |
| Appearance | Comparison against approved standard | After cooling & after 24 hours | Review order of addition |
| Odour | Comparison against retained sample | After cooling & end of window | Review heat exposure & raw material lot |
| Fill weight | Individual weights across the run | Start, middle, end | Adjust filler settings or de-aeration |
| Pack interaction | Closure fit, liner contact, seal integrity | End of observation window | Change closure or liner specification |
| Stability signal | Separation, texture movement, colour shift | Weekly through window | Return to formulation review |
| Documentation | Manufacturing instruction completeness | Before main run booking | Complete draft & re-approve |
What to put in the brief
- Vessel & impeller class intended for production, plus the working range the pilot must reach.
- Written manufacturing method with order of addition, times, temperatures & shear settings.
- Measurements to be taken in the pilot, with the method & the point in the run for each.
- Retention sample quantity, storage condition & the length of the observation window.
- Filler class & target fill weight, with units weighed at start, middle & end.
- Commercial pack & closure to be used for pilot units, not laboratory containers.
- Acceptance ranges for each measured property & who signs them.
- Schedule allowance for an adjustment batch & a confirmation batch after the pilot.
Where our capability fits
We regard scale-up as a scheduled stage with its own slot rather than a step squeezed between approval & filling. Our team runs the pilot in production-class equipment, records measurements at defined points through the run, fills pilot units into the commercial pack, & returns a manufacturing instruction with the acceptance ranges attached. Where a pilot misses a criterion, we document what changed & why before moving to the confirmation batch.
That service covers private label (your brand on a proven base), OEM/ODM contract manufacturing for brands arriving with a finished specification, & full custom formula development for products built from the brief onward. Alongside this piece you may want to review argan oil shampoo refill blocks, softgel capsule blister sample packs & black perfume sampler box sets. For context on process & approvals, see our notes on cosmetic manufacturing quality systems & formula development from brief to production.
Planning a first production run? Share the formula stage you are at through our inquiry page, or write to capsulebiotech@163.com. Calls are welcome on WhatsApp at +86 158-7655-5319 (Capsule Series), +86 158-7655-5319 (Cosmetics Series).
This article offers general manufacturing guidance for brand owners & is not legal or regulatory advice. Processing requirements, testing expectations & market rules vary by destination & change over time; please confirm current requirements against official texts & with your own professional advisers before relying on them.