This concluding part of an article exploring a model for reducing statistical uncertainty when pooling pipeline data applies Part 1’s methodology to actual pipeline-failure data compiled by the US ...