What false-reject rate is realistic to demand
There is no universal false-reject number for every contract. Define it as a stop or an ejection that the recheck finds clean, then count it per kilometre on your hardest roll.
There is no single false-reject number that belongs in every contract. First define a false reject: the line stopped or the piece was ejected, and the recheck found no defect. Then take your hardest roll and count how many times that happens per kilometre. Write that definition and the trial number into the contract. Do not write “as few false rejects as possible.”
Why it is judged this way
A demo roll is new and clean, so the false rejects are almost zero. That cannot be a promise. Take the material in your plant that false-rejects most easily, and record the length and the count. Later acceptance uses this definition. Do not change the wording on the spot.
What to compare
| What you see | What to call it | What not to do |
|---|---|---|
| An ejection that the recheck finds clean | Count it as a false reject | Count a real defect as a false reject too |
| A count measured on the hard roll | It can go into the contract | Use the supplier sample roll’s zero |
| No definition | Write the definition first | Write a percentage that looks good |
At job change or when a roll is opened
- Write down the definition of a false reject.
- Run one roll of the hardest material and count.
- The contract states the definition, the type of roll, and the count. It does not state an empty phrase.
