Weibull calculator: fit your failure data in a minute
A free Weibull distribution calculator: paste times to failure, with suspensions if you have them, and get the shape and scale parameters, four plots and 95% confidence bounds, fitted by maximum likelihood or rank regression. Runs in your browser; no sign-up.
Weibull calculator
Runs in your browser. Nothing you enter is sent to a server or stored.
How to use it
Enter one record per line as time,status: the operating time in one unit throughout (hours, cycles or kilometres), then 1 for a failure or 0 for a suspension, a unit removed or still running without having failed. Choose the fit method: MLE (maximum likelihood) suits small samples and many suspensions; RRX and RRY are rank regression on X or on Y, the methods behind a hand-drawn Weibull plot. Five failures is the practical minimum for a fit you can act on.
1200,1 1850,1 2300,0 2650,1 3100,1 3400,0 4050,1
How to read the result
The shape parameter, beta, says how the failure mode behaves. Below 1, failures fall with age: early-life failures, often from installation or quality, where replacing on a schedule makes things worse. Close to 1, failures are random in time and a fixed replacement interval does nothing: inspect or run to failure. Above 1, failures rise with age, and a planned replacement interval can pay.
The scale parameter, eta, is the characteristic life: the age by which 63.2% of the population will have failed, in the unit you entered.
The probability plot shows the fit. Points along the line support a single failure mode; a dog-leg or an S-shape means mixed failure modes or system-level data, and the fit should not be used as it stands.
The 95% confidence bounds show how sure the fit is. Wide bounds mean more failures are needed before the numbers set an interval.
B10 life
B10 life is the age by which 10% of the population is expected to fail, the usual basis for bearing ratings and warranty terms. From the calculator’s parameters, B10 = eta x 0.10536^(1/beta), where 0.10536 is -ln(0.9). For beta = 2.5 and eta = 10,000 hours, B10 = 10,000 x 0.10536^0.4 = 4,065 hours. Any other life works the same way: Bx = eta x (-ln(1 - x/100))^(1/beta).
What to do with the numbers
Set the interval
Beta and eta give the replacement interval with the lowest cost per hour.
Size the spares
The same parameters give the spares to hold for a target service level.
Why a Weibull analysis gives a poor fit
Mixed failure modes and system-level data, and how to spot them in the plot.
Perform a Weibull analysis in Excel
The same fit by hand, step by step.
A free Weibull calculator for Excel
A one-button Excel version that handles suspensions.
Reliability engineering
The whole analysis done on your CMMS data.
Questions people ask
Is the calculator free to use at work?
Yes, for any use, including commercial work. Cite it as the Pardus Consulting Weibull calculator, with the date you used it.
What data does it need?
Times to failure in one consistent unit (hours, cycles or kilometres), at least five failures, and the times of any units removed or still running without failing. Each line is time,status.
Does it handle suspensions?
Yes: enter a suspended unit with status 0. Leaving suspensions out makes the life look shorter than it is.
Can I cite the result in a report?
Yes. State the fit method (MLE, RRX or RRY), the number of failures and suspensions, and the date of the analysis, so the result can be reproduced.
Who built it
The tool was built by Pardus Consulting and uses the same methods the firm applies on client work. When the question is bigger than a calculator, the analysis is run on your own CMMS data and signed by a registered professional engineer.