A bench formulation that works is a good day. Turning it into a production run at a cost that makes commercial sense is a different exercise, and it is where a lot of promising products stall.
The problems are rarely aromatic. They are arithmetic, procurement and process, and they mostly become visible only once volume goes up.
Start by knowing your real loading
Before you scale anything, confirm what your bench batch actually contained rather than what the recipe said.
This matters because bench work is where measurement error is largest. Small volumes, syringes, fast iterations and open containers all push actual loading away from intended loading. If your bench sample was formulated at a nominal seven per cent but the terpenes were dispensed by volume without a density conversion, the real figure was closer to six.
Scale that up faithfully and you have scaled the error. Scale up the corrected number and the product will not smell like your sample.
So: recalculate. Weight of terpenes divided by total product weight, using actual measured weights, not recipe figures. If you did not record actual weights at bench scale, that is the first process change to make.
The scaling maths
The good news is that scaling is linear. Concentration by weight stays constant regardless of batch size, so you multiply proportionally.
For a target of five to ten per cent:
- One ounce of base material (28.35 g) needs roughly 1.4 to 2.8 grams of terpenes
- One pound (453.6 g) needs roughly 22.7 to 45.4 grams
- Ten pounds needs roughly 227 to 454 grams
Convert to volume at the end, using the batch density from the certificate of analysis, not before. Doing the conversion early and then multiplying compounds any rounding you introduced.
The other thing that scales linearly is error. A one per cent measurement error on a 5 mL bench batch costs you almost nothing. The same proportional error on a fifty-litre production batch is expensive, and it is why weighing rather than measuring by volume becomes non-negotiable at volume rather than merely advisable.
Modelling cost per unit
This is the calculation that decides whether a product line works, and it needs more inputs than people usually include.
Terpene cost per unit. Take your loading percentage, apply it to fill weight, convert to volume using density, and multiply by your per-millilitre cost at the volume tier you will actually be buying at.
The volume tier point is important. Terpene pricing typically drops substantially with quantity, and the per-millilitre cost at sample size bears no relationship to the per-millilitre cost at bulk. Modelling your product’s margin using the 5 mL sample price will make the product look unviable when it may not be. Modelling it at bulk pricing you cannot yet commit to will make it look better than it is.
Model both. Know your break-even volume.
Waste and yield loss. Real processes lose material. Transfer losses, residue in vessels, evaporation during handling, rejected units. A yield assumption of one hundred per cent is a fiction that shows up as a margin shortfall later. Build in a realistic loss figure based on your actual process, and revisit it after your first few production runs.
Testing costs. Batch testing is a per-batch cost, which means it is a larger per-unit cost at small batch sizes. This is one of the strongest arguments for consolidating production runs rather than making frequent small batches.
Labour. Mixing, homogenisation, quality checks and documentation all take time, and that time does not scale linearly downwards. A small batch costs nearly as much labour as a medium one.
Carrying cost. Terpenes bought at bulk pricing sit in inventory. Volatile compounds degrade, and capital tied up in ingredients is capital not doing something else. Bulk pricing is only cheaper if you use the material within its useful life.
Add those together and the terpene ingredient cost is often a smaller share of true per-unit cost than people assume, which changes procurement decisions. Chasing a marginally lower per-millilitre price at the expense of documentation quality or batch consistency is usually a bad trade once you account for the cost of a failed run.
What breaks at volume
Five things that work at bench scale and stop working at production scale.
Homogenisation. Stirring a small vessel by hand distributes terpenes adequately. Stirring a large vessel by hand does not. Concentration gradients in large batches are common, and they mean units from the same batch test differently. Scale your mixing method, not just your quantities, and verify by sampling from multiple points and depths.
Evaporation exposure. A large batch takes longer to prepare, which means longer open-container time, which means more loss of light monoterpenes. Profiles drift towards their heavier components. This is a real and often unexplained reason production batches smell different from samples.
Temperature control. Small volumes equilibrate to room temperature quickly. Large volumes do not, and mixing warm terpenes into cool base material or vice versa produces uneven results.
Measurement tooling. The scale that handled bench work may not have the capacity for production quantities, and switching instruments mid-process introduces inconsistency. Plan the tooling for the batch size you intend to reach.
Supply continuity. A bench formulation locked to a specific supplier batch is not reproducible if the next batch differs. This is where analytical documentation and supplier consistency stop being a procurement nicety and become a production dependency.
Keeping cost modelling honest
A few habits that separate useful models from optimistic ones.
Build the model in a way that lets you change the loading percentage, fill weight, batch size and per-millilitre price independently, then look at what actually drives the number. Usually it is fill weight and loading percentage, not price per millilitre.
Test the sensitivity. If a half per cent change in loading moves your margin materially, your process needs tighter measurement control before you scale.
Model at more than one volume tier so you know where the economics change.
Recalculate after every production run using actual consumption rather than theoretical consumption. The gap between the two is your real waste figure, and it is almost always larger than the assumption.
Anyone comparing tools for this will find that a true terpenes calculator style utility that outputs both required quantity and cost per unit across a loading range is more useful than one that only handles the mixing ratio, because the loading decision and the margin decision are the same decision.
The sequence that works
- Confirm actual bench loading from measured weights
- Recalculate scaled quantities by weight, converting to volume only at the end
- Model cost per unit at realistic volume tiers with a genuine waste allowance
- Run a mid-size pilot batch before full production
- Sample the pilot from multiple points to verify homogenisation
- Evaluate the pilot at intervals, not just on day one
- Reconcile actual material consumption against theoretical and adjust the model
The pilot batch is the step people skip under time pressure, and it is the one that catches homogenisation and evaporation problems while they are still cheap to fix.
The bottom line
Scaling terpene formulations is linear on paper and non-linear in practice, because mixing, evaporation and temperature all behave differently at volume.
Confirm your real loading, calculate by weight, model cost with honest waste assumptions across the volume tiers you can actually access, and pilot before you commit. That sequence turns a formulation that smells right into a product that ships consistently, which is a considerably harder achievement and the one that actually pays.
