Chapter 18: The Batch-to-Batch Bottleneck — Reclaiming the Dead Time Between Drops
In the previous chapter, we used OEE to expose where your machine's hours are leaking, and for the vast majority of roasteries, the largest single loss lives in Availability — the difference between the time the drum is scheduled to run and the time it's actually roasting. This chapter is about hunting that loss down to its source, which is almost always the same place: the turnaround between batches.
A roaster's drum only earns money while it is charged and roasting. Every second it spins hot and empty is fuel spent, time lost, and capacity quietly evaporating. Yet this dead time is so routine that most operators stop seeing it. It hides in plain sight, batch after batch, day after day.
Anatomy of the Dead Time
The gap between dropping one batch and charging the next is rarely one single delay. It's a stack of small ones, and they compound. Three culprits dominate:
- The Cooling Tray That Hasn't Cleared: This is the most common bottleneck of all. If your cooling tray is still working the previous batch when the next one is ready to drop, the entire line stalls. The roaster cannot drop into an occupied tray, so it holds — hot, full, and overdeveloping the coffee inside while it waits. The cooling tray, not the drum, is dictating your pace.
- Manual Destoning and Handling: If beans are being hand-scooped, lifted, or shuttled into a destoner or bin between stages, you've inserted a slow, human, error-prone step directly into your critical path. The drum waits on a person carrying a bucket.
- Green Staging at the Last Minute: If the next charge of green isn't weighed, positioned, and ready to drop the instant the drum is clear, you've built a pause into every single cycle. Multiply that pause by your daily batch count and the lost hours are staggering.
The Lean Angle: Attack the Changeover, Not the Machine
In Lean Manufacturing, this is a classic changeover problem, and the discipline that solves it is SMED — Single-Minute Exchange of Die. The core principle is deceptively powerful: separate the tasks that must happen while the machine is stopped (internal setup) from the tasks that can be done while it's still running (external setup) — then move as much as possible from internal to external.
Applied to a roastery, this discipline takes a concrete form I call the Between Batch Protocol (BBP) — a strict, repeatable routine that governs exactly what happens in the window between one batch dropping and the next one charging. The BBP is where SMED stops being theory and becomes a drill your team runs without thinking. The question it forces is relentless: what are you currently doing during the dead time that could have been done before the batch dropped? Weighing the next green charge, positioning the bin, clearing the tray, prepping the destoner — every one of these can often be shifted to run in parallel with the active roast, so that the moment the drum is clear, the next charge goes straight in. The drum never waits.
This is the heart of why a well-drilled 22kg roaster can out-produce a clumsy 60kg one. The bigger machine isn't roasting any faster — it's just losing the same dead time in larger, more expensive chunks.
The Specialist Technician's Action Plan
To reclaim the dead time between drops, work through these steps:
- Time the Gap, Not the Roast: Most operators obsess over roast time and ignore turnaround. For one day, time only the dead interval — from the second the batch drops to the second the next one charges. That number is your real target, and it's usually far larger than anyone guesses.
- Map Internal vs. External Tasks: List everything that happens in that gap. Mark each task as one that genuinely requires the drum to be empty, or one that could be prepped in advance during the previous roast. Be ruthless — most "internal" tasks are external in disguise.
- Build a Written Between Batch Protocol (BBP): Turn that mapping into a fixed, written routine. Weigh and position the next green charge before the current batch finishes. Ensure the cooling tray is empty and ready. Have the destoner and collection bins set before you need them. A BBP that lives in someone's head isn't a protocol — it's a hope. Write it down, post it at the machine, and drill it until it's automatic.
The Tech Note: "I once watched a team run a beautiful, fast roast profile — and then lose four minutes after every single drop because the next batch of green was still sitting in the sack. They were convinced they needed a second roaster. We wrote a simple Between Batch Protocol, pre-staged the green, cleared the tray on a timer, and found nearly two hours of extra roasting in their existing day. No new machine. No new loan. Just a drum that stopped waiting around." — JG