What the Spec Sheet Won't Tell You About Choosing an Extraction System

Every extraction system pitch looks the same on paper: throughput per run, solvent-to-biomass ratio, footprint, price. Operators build spreadsheets comparing those four numbers across three vendors and pick the best row. Then the system arrives, gets commissioned, and six months later the "best row" is the one gathering dust because a seal needs a part that's back-ordered eleven weeks.

The spec sheet isn't wrong. It's just answering a much smaller question than the one that actually decides whether the purchase pays for itself.

Throughput is a lab number. Uptime is a business number.

A system rated for 20 lbs/run in a vendor's demo facility, run by a vendor technician, on biomass the vendor hand-picked, tells you almost nothing about what that same system will do in your facility, run by your team, on whatever the cultivation side hands off that week. The gap between rated and realized throughput is where most ROI projections quietly fall apart.

What closes that gap isn't a better spec sheet. It's serviceability: how fast a tech can get a bad valve open, whether your team can diagnose a pressure fault without a support call, and whether "support call" means a same-day callback or a ticket that sits for a week while a solvent recovery run is stalled mid-cycle.

The questions that actually predict five-year cost

None of these show up on a comparison spreadsheet, and all of them matter more than the sticker price:

• Parts lead time, not parts price. Ask what's in stock domestically versus built to order overseas. A $400 part that takes eight weeks to land costs you a lot more than $400.
• Who answers the phone at 11pm during a run. Not whether support exists — whether it's a person who's actually run the system, or a ticket queue.
• Verified solvent recovery numbers, not vendor-claimed ones. Ask for recovery data from an operating facility, not the spec sheet. Recovery efficiency compounds into real solvent cost over a year.
• What the utility and permitting footprint actually requires. Electrical service, C1D1/C1D2 classification if solvent-based, make-up air — these decide whether the system fits the building you have or the one you'd need to build.
• Whether your team can be trained to run it safely without the vendor standing over their shoulder. A system that requires a specialist on every run isn't scalable past one shift.

The first-in-market tax

I've commissioned systems that were among the first of their kind approved in a licensed market — new equipment, no local precedent, no other operator's documentation to borrow from. That work is real, and it's worth doing when the technology justifies it. But it comes with a cost most operators underprice: you become the vendor's beta tester and the regulator's first case study, simultaneously, on the same timeline you promised your investors a launch date.

If you're going to pay that tax, pay it deliberately — go in knowing the documentation burden is yours to build, not the vendor's, and budget the extra months it takes to get a novel system through state approval. If you don't need to be first, being second with a system that has two years of other operators' failure modes already worked out is usually the more profitable choice, even if it feels less impressive.

That's not a preference. It's what the ROI math actually says once uptime is in the equation.