Overview

The Weibull distribution is widely used for lifetimes and failure times, such as how long a bearing, a battery or a component lasts. Its shape parameter describes whether failures get more or less likely with age.

Enter a time x, the shape k and the scale λ.

Worked examples

InputsResult
x500Shape (k)1.8Scale (λ)650Weibull PDF: 0CDF: 0.46Reliability P(X > x): 0.54
Show 2 more examples
x1Shape (k)1.5Scale (λ)2Weibull PDF: 0.37CDF: 0.3Reliability P(X > x): 0.7
x100Shape (k)0.7Scale (λ)300Weibull PDF: 0CDF: 0.37Reliability P(X > x): 0.63

How it works

The CDF gives the chance of failure by time x, and reliability is the chance of lasting longer. A shape under 1 means early failures dominate, 1 means a constant failure rate, like the exponential distribution, and over 1 means wear-out. The scale λ is the time by which about 63.2% of items have failed.

The formula

P(X ≤ x) = 1 − e^(−(x ÷ λ)^k)
kShape
λScale, the characteristic life

Practical tips

  • With k = 1.8 and λ = 650 hours, about 46% fail by 500 hours, so 54% are still working.
  • PDF values are densities, and for large scales they're tiny, so the default shows 0 after rounding.
  • Fit k and λ from real failure data with a Weibull plot or statistics software.

Frequently asked questions

What does a shape of 1 mean?

A constant failure rate, so age doesn't matter. It's the same as the exponential distribution.

Why 63.2% at the scale value?

At x = λ, the CDF is 1 − e^(−1), which is about 0.632, whatever the shape.

Where is it used besides engineering?

Wind speeds, survival analysis and material strength.