Abstract
Time and cost estimation and the accuracy of it, are central to engineering design and forms the basis for economic analysis of projects. There are several factors that could result in cost or schedule overruns ranging from unplanned non-productive time, inefficiencies, changes in macro indices etcetera; however, one often overlooked factor is deficient time and cost estimation. Hence in addition to factoring prevailing market and contract rates for materials and services, it is important to critically analyze and benchmark plans against known performance for higher accuracy around estimations.
Well project time and cost models generally consist of estimating in modules or sub-phases and aggregating these modules to makeup the total. This can either result in a single discrete estimate or in ranges based on probability and statistical performance - inherently implying that some form of historic performance is crucial to the estimation accuracy. This paper describes a structured approach to developing a probabilistic estimation tool by analyzing past performance data at a phase or subphase level. This tool can be domiciled on a range of computation platforms using similar methodology, which comprises data collection from execution reports, data cleanup and organization to harmonize terminologies and group operation types, and finally statistical and mathematical data analysis. Statistical analysis develops probabilistic relationships in the dataset and correlation between performance variables such as depth and time; while mathematical analysis incorporates numerical correlations and multiple variables to generate estimates in modules and finally aggregates the discrete phase estimates.
The estimation has two major components – Time and Cost.
The analysis of time component considers the productive and non-productive time by phase; determines depth dependent operations and their correlation to time and assigns a mathematical function to each phase
The cost component is broken down into two sub-components – recurrent cost which is highly time-dependent and non-recurrent (material and services) cost which are usually based on pre-defined contractual rates
An additional end function is benchmarking, for comparison between estimates and historic performance
The aggregate of the probabilistic estimate of each module gives the total estimate of time and cost for a given well construction or intervention scope, with the overall objective of improving and maintaining estimation accuracy to avoid overruns and over-estimation of drilling, completions, workover and intervention projects.
