Oil industry is very strategic for any country. Not only for their market share and their budgets, but also for the critical infrastructure interdependences that creates. Considering statistics and fractal geometry as a support for analysis, interpretation of data and as an aid in taking decisions, this paper deals with data obtained in 1994 and 2004 from the inspection of the onshore pipelines in Mexico. Making use of software Benoit and ITSM-2000, the time series were found to have a fractal behavior. Further analysis and contrast with a previous research base on artificial neural networks was also accomplished. As a result it was obtained a forecast for 2014, based on the sum of the time series of data.