Count attempts, not just the last failure
A failure allowance needs a clear definition. Here, failure rate means the probability that an entire plate attempt fails. Each attempt is assumed independent, and attempts continue until one complete plate succeeds.
Under that model, expected additional attempts equal f ÷ (1 − f), where f is the failure probability. This differs from adding f percent to a successful plate’s cost, because a retry can also fail.
Only repeat the costs a failed attempt consumes
A failure halfway through a job may consume less material, power and machine time than a complete plate. The Failed resources input applies one shared percentage to these categories. Setup labor is repeated in full for each attempt.
Finishing and packaging are applied only to the good parts from the successful plate. The model does not charge packaging for a failed plate that never reaches packing. If recovery creates extra cleanup labor beyond normal setup, add that time to your setup assumption or model it separately.
When this model is a poor fit
Partial successes need a different model. A plate with nine good parts and one rejected part is not the same as losing the entire plate. Persistent mechanical problems also violate independence because one failure can make the next failure more likely.
A 100% failure rate has no successful endpoint and is rejected. Very high failure assumptions can create large expected costs; the result should trigger a process check, not automatically a high customer quote.
Replace examples with production records
Keep a simple record of attempted plates, successful plates, elapsed time at failure and material consumed. Use comparable machines, materials and models when estimating a rate. A tiny sample can be misleading.
Test a lower and higher rate around your estimate to see how sensitive the quote is. Save those scenarios with clear names. The current calculator does not read a printer, verify records, or predict the next print’s outcome.