Quick answer: Match the machine to your two-year volume forecast, sized to run 50–70% of stated capacity for repeatable quality. Burner output drives true capacity: plan 11,500 kJ/hr per kg of green coffee. Verify domestic service support and get a written delivery contract with a penalty clause before you wire a deposit.

The most common roastery buying mistake I see isn't overspending on the machine. It's undersizing it. Buy for where you'll be in 24 months, not where you are today. I've watched a dozen first-year operations end up running six-day weeks and turning away wholesale accounts because they bought to current volume. The math on a second machine almost always lands higher than the cost of stepping up once.

What does a commercial coffee roaster actually cost?

The machine price isn't the budget. For a 6–15 kg roaster installed in the US, your actual spend includes chimney ductwork, pollution-control equipment (an afterburner or thermal oxidizer — many regional air quality districts now require one), your initial green-coffee inventory, roast-logging software, cupping supplies, and QC gear. Then add installation labor, architectural drawings, and permits on top of that.

The permit side alone takes time most buyers don't budget for. A new roastery in most US municipalities needs a change-of-use permit plus a separate air quality permit from the regional air district. Neither moves quickly.

Specific dollar figures from earlier versions of this article were written in 2020 and aren't reliable today. Before you lock in a number, call your local contractors and your regional air quality management district for real line-item quotes. Those two conversations will calibrate your budget faster than any guide can.

What hasn't changed: I've never talked to a roaster who regretted buying slightly more capacity than they needed. The ones who regret it went the other direction.

How do you size a coffee roaster for your operation?

Start with a two-year volume projection. Estimate your weekly green-coffee volume 24 months out, then choose a machine that handles it in 25 operating hours or fewer per week. Work backwards from that number to the batch size and throughput you need.

Two facts the spec sheet won't surface:

• Green coffee loses 14–20% of its weight during roasting. Third-wave light roasts shed roughly 14%; darker commercial roasts can shed 18–20% or more.
• Real throughput for quality output is 50–70% of stated capacity at 3–3.5 batches per hour — not the number on the brochure.

A concrete example: the Diedrich IR-12 is rated at 12 kg. Run it at 50% load (6 kg green per batch), 3.5 batches per hour, with 15% weight loss, and you're producing about 17.9 kg of roasted coffee per hour (3.5 × 6 × 0.85). The sales rep will tell you 42 kg/hr at full load. Both figures are technically accurate. Only one of them reflects what you'll actually produce with repeatable, quality roasts.

The real constraint isn't drum size. It's burner output. A quality roast needs approximately 11,500 kJ/hr (5,000 BTU/hr) of burner output per kg of green coffee. An undersized burner will roast slowly and unevenly at half load, regardless of drum capacity. Note that this formula doesn't apply to recirculation machines, which recover heat from exhaust and get higher effective capacity per unit of burner output.

How do you evaluate a roaster's reliability before buying?

Talk to owners, not sales reps. Sales reps are paid to close orders. They won't tell you about the control board that fails after 200 hours or the gas valve that sticks in cold weather. Owners will — especially if you ask directly.

Simpler machines tend to fail less. The UG-series Probats are the classic example. Roasteries that bought them in the 1970s and 80s are still running them. Heavy cast-iron construction, minimal electronics, very little to go wrong. The tradeoff is that they demand skilled manual operation and don't log data the way modern machines do. Whether that's acceptable depends on your staff and your operation.

Modern, more automated machines produce consistent results with less operator skill. But when the control board dies at 6 a.m. on a Monday with 200 kg of orders to fill, what you actually need is a US-based technician who can overnight a replacement part. Find out if that exists before you wire any deposit.

Seek out roasters who've owned your target machine for at least three years. Ask what broke, how fast it got fixed, and what it cost. Roasters are unusually candid about equipment failures.

Why does post-sale service matter more than the sales pitch?

The sales experience is almost useless as a signal. Every manufacturer has a good sales rep. What matters is what happens 18 months in, when something breaks and your roastery goes quiet.

Before you sign anything, get answers to these questions:

• Is there a service technician based in the US, or does every repair mean flying someone in from overseas?
• What's the typical lead time on replacement parts? Are common wear parts stocked domestically?
• What does the warranty cover, specifically?
• Can you speak with three owners who've needed service in the past two years?

That last question is the one most buyers skip. Three owners who've actually needed post-sale service will tell you more than any number of spec sheets or trade-show conversations. Manufacturer-chosen sales references are useless for this. Find owners yourself through industry forums and regional SCA chapter networks.

Does user interface actually affect roast quality?

More than most buyers expect. At 20–40 production hours per week, a clunky control surface compounds into real quality variance over time.

Here's the specific failure mode: machines that require multiple button taps to adjust gas pressure respond slower than machines with an analog dial or a smooth touchscreen. When you need a 2% gas adjustment during the development phase after first crack, slow controls mean you're always reacting late. The logged curve doesn't match your intended profile. Tuesday's roast doesn't replicate Sunday's.

Look for high-resolution gas control (a dial or smooth digital response, not button-tap increments), large temperature and time displays you can read without turning away from the drum, and a digital air-pressure manometer at the drum-exhaust duct. The manometer is the spec most entry-level machines skip. Without it, you can't confirm you're running the same airflow batch to batch as weather and duct fouling shift the system's resistance.

What are the four main roaster configurations?

Four architectures cover the commercial market. Each has a distinct thermal profile and different operational demands.

Classic drum roasters

A drum rotates above a gas flame while a fan draws hot air through and out the exhaust. Most commercial machines under $50,000 use this design. They're thermally stable, but they respond slowly to gas changes. Drum mass absorbs and releases heat between your input and the bean's response, so there's always a lag.

The critical spec is drum construction. A thin single-walled drum transfers heat conductively to the bean surface, which produces harsh cups and increases scorching risk. A double-walled drum runs cooler at the surface and transfers heat primarily through convection. If you're buying a classic-drum machine, insist on double-walled construction and confirm the burner output is adequate for your target capacity. Fans, valves, and software can all be upgraded post-delivery. The drum and burner rating usually can't.

Indirectly heated drum roasters

The burner chamber sits separated from the drum. Hot combustion air flows through the drum rather than a direct flame sitting beneath it. The drum surface runs cooler, which reduces tipping risk. The tradeoff is that airflow management becomes more involved. Classic-drum machines rarely need mid-roast airflow adjustment; indirectly heated machines require it as standard operating procedure.

Recirculation roasters

A portion of exhaust air recirculates back through the burner and drum. Energy-efficient by design. The risk is smoke taint: recirculated air that isn't brought back up to afterburner-equivalent temperatures before re-entering the drum will contaminate the batch. Well-designed recirculation systems produce clean cups. Systems that cut corners on combustion temperature don't.

Fluid-bed roasters

Rising hot air heats and agitates the bean mass simultaneously. No drum contact means no conductive heat transfer and no scorching risk. The historical limitation has been control-system simplicity: most commercially available fluid-bed machines haven't offered the independent airflow and heat controls needed for nuanced profile work. That's changing as roast-control software matures. Fluid-bed designs are becoming genuinely competitive for specialty production in 2026.

Which roaster features are actually worth paying for?

None of these are required to produce good coffee. All of them make precision and repeatability meaningfully easier.

Double-walled drum and adequate burner output. These are the two specs you can't change after delivery. Everything else is upgradeable.

Variable-speed-drive (VSD) fan. A fixed-speed fan runs at one speed. But weather changes, duct fouling, and seasonal pressure variations all shift actual airflow through the system even when the dial reads the same number. A VSD fan paired with a digital air-pressure manometer gives you repeatable airflow batch to batch. Without it, you're chasing variables you can't see.

Digital air-pressure manometer. Measures pressure in the duct between drum and exhaust fan. Pressure correlates with airflow and gives you a repeatable reference number that a fixed fan setting alone can't provide. Clean the chimney regularly — duct fouling shifts the pressure-to-flow relationship over time, and that's the only fix.

High-resolution digital gas manometer. Replace the stock analog gauge with a high-resolution digital one. Analog gauges are imprecise at the low-pressure ranges that matter during development. Digital gauges let you record and reproduce genuinely precise gas settings.

Properly positioned bean and environmental probes. You need both, each 2.5–4 mm in diameter. The bean probe tip should sit 3–5 cm from the inside of the faceplate and 3–5 cm from the inner drum edge, positioned in the center of the bean mass. For machines under 1 kg capacity, 2 cm from the drum edge is acceptable. If the probe tip sits above the bean mass during a small batch, you're reading air temperature, not bean temperature — and you can't replicate that curve.

A written delivery contract with a late-penalty clause. Not a machine feature, but it belongs in any honest buyer's guide. Manufacturers routinely promise three-month delivery, take a 50% deposit, and ship six to nine months later. The buyer sits on an empty roastery paying rent while the manufacturer faces zero financial consequence. Get a written sales contract with a guaranteed delivery date and a penalty clause for late delivery before you wire anything. A manufacturer that won't sign that is showing you exactly how every future dispute will go.

What accessories does a new roastery actually need?

For a 6–15 kg operation, have these on hand before your first production roast:

• Floor scale, resolution 0.005 kg (0.01 lbs) or better; capacity should exceed your largest batch weight plus the bucket
• Fire suppression: a hard-plumbed water line into the faceplate and chaff collector with a spray head and easy-access valve — the most effective protection against a roaster fire
• Full-spectrum spotlight mounted above the bean trier
• Wet/dry vacuum for chaff collector cleaning
• Chimney brush
• Cupping supplies
• Green coffee storage buckets with labels
• Roasted coffee buckets (larger than green; beans expand during roasting)
• Heat sealer, bags, boxes, packing tape, label printer
• Food-grade high-temperature silicone for pipe resealing after cleaning
• Bearing grease specified by your machine's manufacturer

A commercial dishwasher is worth adding if the budget allows. You'll clean a lot of cupping equipment.

For the full range of commercial and prosumer machines stocked at CoffeeRoast Co., browse the home and commercial roasters collection. If you're still evaluating smaller-batch home setups, the companion guide to best home coffee roasters covers machines under 500 g batch capacity in detail.

Frequently Asked Questions

How much does a commercial coffee roaster cost?

Commercial drum roasters in the 6–15 kg class run roughly $15,000 to $80,000+ depending on brand, drum construction, and automation level — and that's only the machine. Chimney ductwork, pollution-control equipment, permits, installation, and initial green-coffee inventory all add to the total. First-time buyers consistently underestimate total startup cost by 30–50%. Get real line-item quotes from local contractors and your regional air quality district before you lock in any budget.

What capacity roaster should I buy for a startup roastery?

Project your weekly green-coffee volume two years out, then choose a machine that handles it in 25 or fewer operating hours per week at 50–70% of stated capacity. If you expect to roast 150 kg of green coffee per week in year two, a 6 kg machine running 3.5 batches/hour at 50% load produces roughly 53 kg of roasted coffee per 8-hour day. You'd cover that volume in about three production days. Never plan around 100% stated capacity.

Classic drum vs. fluid-bed: which should I buy?

For most commercial buyers, a classic double-walled drum with adequate burner output is the lower-risk starting point. The technology is proven, service infrastructure is established, and the operational learning curve is well-documented across thousands of installations. Fluid-bed machines have genuine advantages — cleaner cup profile, faster development, no scorching risk — but most commercially available models still give you less profile-control precision than the best drum machines. That gap is closing in 2026 as software improves.

How do I verify a manufacturer's service quality before buying?

Ask to speak with three owners who've needed warranty or post-warranty service in the past two years. Specifically ask: how long did the repair take, did the technician come on-site or did you ship the machine, what did parts cost out of warranty, and was there a US-based technician or did you wait on international support? Manufacturer-chosen references tell you nothing. Find owners yourself through industry forums and regional SCA chapter networks.

What is the 50% capacity rule for coffee roasters?

Run 50–70% of stated capacity per batch for quality output. At 100% load, most machines lack sufficient burner output to maintain the rate-of-rise curve you need. Roast times extend, development gets compressed, and batch-to-batch consistency drops. A Diedrich IR-12 produces more consistent output at 6–8 kg batches than at the nominal 12 kg maximum. The throughput formula: batches/hr × batch size (kg green) × (1 minus weight loss %) = kg roasted/hr.

Why do coffee roasters take longer to deliver than promised?

Manufacturers typically require a 50% deposit at order time, then ship on their own production schedule. Once your deposit lands, the incentive to hit the quoted date weakens. Delays of 3–6 months beyond the quoted lead time are common; delays of 6–9 months happen on roughly half of new-buyer orders. A written contract with a guaranteed delivery date and a financial penalty clause for late delivery is the only structural fix. If a manufacturer won't sign it, treat that as a signal.

What probe specifications should I require on a commercial roaster?

Require both a bean probe and an environmental probe, each 2.5–4 mm in diameter. The bean probe tip should sit 3–5 cm from the inside of the faceplate and 3–5 cm from the inner drum edge, positioned in the heart of the bean mass for batches as small as 20% of stated capacity. A probe tip sitting above the bean mass on a small batch reads air temperature, not bean temperature. A curve built on air-temperature readings can't be replicated.

What roast-logging software do commercial roasters use?

Cropster and Artisan are the two dominant options in 2026. Cropster is a subscription SaaS platform with green-coffee inventory management, production scheduling, and QC tracking built in — well-suited for operations above 300 kg/week. Artisan is open-source, free, and paired with most serious prosumer and small-commercial setups because it runs on a laptop and integrates directly with most machine control hardware. Both log bean temperature, rate of rise, and development time ratio per batch, which is the minimum you need for repeatable production.

Key takeaways:

• Size to your two-year volume forecast. Run 50–70% of stated capacity at 3–3.5 batches/hour for quality output.
• Burner output, not drum size, determines true throughput. Target 11,500 kJ/hr (5,000 BTU/hr) per kg of green coffee.
• Ask three owners who've needed post-sale service before committing. No sales rep can substitute for that.
• Insist on a written delivery contract with a late-penalty clause before wiring any deposit.
• Drum construction and burner output are the two specs you can't easily change after delivery. Everything else is upgradeable.