Part III. Lesson 10. Rate and Bottlenecks.
An engineer runs a rate calculator, gets a clean number, writes it on the drawing, and moves on. Then the first real wave releases, and the line that was going to do twenty cartons a minute is jamming at fifteen. Nothing broke. The math was correct.
The number was still wrong for the building.
Every rate calculator produces a theoretical result on ideal conditions: perfect surface contact, commanded belt speed, consistent gaps, no variation. Real systems don't run that way. The biggest source of the gap is slippage: the belt moves faster than the package, so the space that opens between cartons comes out smaller than the formula predicted.
The response is margin. You never design right to the minimum. In solutioning, apply plus or minus ten percent and confirm the flow can carry the rate, not set a specification.
The formulas do not account for slippage. That is the most important thing to understand about gap calculations. When you run the calculator and it gives you a minimum required gap, that is the minimum in a perfect world. In the real world, you need to add buffer to that number. But never design right to the minimum. The gap that forms in practice will be smaller than the gap the formula assumes.
The gap between cartons after a speed change. Accelerate and it grows; decelerate and it shrinks.
The maximum cartons per minute at a given speed and gap. Design to the worst case.
Center-to-center distance between cartons. Use it to sanity-check the rate.
The Calc Logic Guide is the formula authority. Run each at plus or minus ten percent and read it as directional, not final.
If the Product Spec Calc hands you a Gap Produced or a Theoretical Rate, then build margin into it before you use it in any decision: design above the minimum gap and below the max theoretical rate. Tradeoff: you give up a little headroom on paper. Verify: the gap that forms in practice is smaller than the formula's, because the belt slips and the formula doesn't know that. A system calculated to just barely meet the minimum won't meet it in the field.
The volume design target has to be set explicitly and written down. Design for full peak, or for a percentage of peak the customer agrees to supplement with temporary labor. Both are valid; leaving it unresolved isn't a choice, it's a gap.
"Design rate is one of the first decisions made on any project, and it has to be made deliberately... Spikes are not exceptions in distribution. If the design rate was not explicitly agreed with the customer early in the process, the system will eventually be asked to do something it was never sized to handle."
MICHAEL COLLINSDesigning for average volume and calling it a complete design. It runs smooth in the demo and every calm afternoon. Then a wave releases, or volume grows, or a station backs up, and the system that was sized for the average discovers it was never designed for the day it will actually have. Set the design target above average, deliberately, and write it down.
Take your flow and run the wave-release question out loud with someone. Describe what happens the second a wave drops and every zone fills at once. If they ask you something you can't answer, you didn't find a hole in your explanation, you found a hole in the flow. Find it now. The stress test takes twenty minutes; a redesign takes weeks.
| Section | Share | Required rate |
|---|---|---|
| Door 1, Carrier A | 55% | 11 CPM |
| Door 2, Carrier B | 35% | 7 CPM |
| Door 3, overflow | 10% | 2 CPM |
| Combined at the merge | 100% | 20 CPM |
Set the design rate explicitly at 20 CPM, not the 18 CPM current peak. Sanity-check with the standard case, apply margin, then run the five-question stress test against the flow.
Next, Lesson 11: How does a conveyor actually work, and what in it decides whether it lasts?