Statistical analysis to forecast the probability to reach the cost and manhours estimated for a project


Diéguez Ramírez, Brian

The aim of this study is to perform statistical analyses with different tools to obtain statistical models in order to estimate the forecasted manhours that will be spent in future projects. These manhours are presented at proposal phase. The used data source has been engineering projects already performed in Technip Energies. The study has been based on relating the manhours carried out in the projects analysed with the different Key Performance Indicators of the different departments. First, the Key Parameters of the international and big projects have been selected and classified by importance. Then, a preliminary study has been conducted with R Studio trying to reduce the dimension of the available data. The next step has been the statistical analysis of the data to obtain regression curves that would allow estimating the number of manhours of the projects from the values of the characteristic parameters. Finally, the curves obtained have been analysed with the Monte Carlo method, using a tool called Safran Risk, to find the probability of obtaining certain manhours for the project. With these studies it is possible to obtain regression curves and results that, together with subsequent more detailed studies, will help during the estimation phase of manhours required for new projects.



Sempere Cebriàn, Julià
Gumà Gras, Josep


IQS SE - Master’s Degree in Chemical Engineering