Agent based stock and flow model for project planning and control

DYNAMIC ANALYSIS OF CONSTRUCTION PROCESS FOR PROJECT MANAGEMENT

Abstract

This study analyzes the ex post patterns of resources employed in the construction of an industrial
plant. The aim of the study is to understand the rational of planned resource allocation and hence to
extrapolate the actual pattern when conditions change. The study explains how the identified logic
of resource allocation becomes the basis of a Stock and Flow dynamic model applicable to any
construction Work Package (WP). The model takes into account the delay time of resource
allocation, the erection sequence constraints and the feedback control generated by the actual vs.
planned recovery actions.

Keywords

System Dynamics, Agent Based, Stock & Flow, Project Management, Project Control, Construction simulation

Foreword

“Construction is the biggest industry in the world, and yet, is not performing well. The construction ecosystem represents 13 percent of global GDP, but construction has seen a meager productivity growth of 1 percent annually for the past two decades. Time and cost overruns are the norm, and overall earnings before interest and taxes (EBIT) are only around 5 percent despite the presence of significant risk in the industry.” (McKinsey Global Institute, June 4, 2020). A statistic carried out by the Construction Industry Institute (USA) on 975 industrial projects of various sizes, found that only 5.4 percent of these were able to meet the schedules and costs budgeted (PricewaterhouseCoopers LLP report, 2013). As regards in particular the large projects (Megaproject), the statistics confirm what is called the “iron law of Megaproject” or nine out of ten end up late with extra costs that frequently reach up to +50 percent and in some cases even more (Flyvbjerg, B.,2014). Fortunately, project planning and management techniques have been refining more and more during last 30 years and this improvement has reduced the negative impact on performance (PMI’s Pulse of Profession, 2017). However, the results remain poor and seem difficult to improve due to the intrinsic complexity of the projects (Girmsheid, G., Brokmann, C., 2007). Among the main causes, if not the main cause, of these disappointing results is the bad forecast of time and costs (PricewaterhouseCoopers LLP report, 2013). Unfortunately, in several project plans, resources are not even quantified (White, J.C., Sholtes, R.M., 2016) notwithstanding they are the mean that bear the construction effort. Project planned duration is often based on predefined duration of each single WP that is part of the project Work Breakdown Structure (WBS) whose critical path gives the overall project duration. This means that each single WP duration and the overall project duration are mainly based on the experience of the person or team in charge for planning the project more than on a rational evaluation about the amount of resources that have to be mobilized and how effectively they can work.

This fact has the following negative consequences: 

  1. The plan becomes a simple monitoring tool of the project progress that can only provide a “picture” of the project delay and extra effort without any control on it; 
  2. The plan does not consider those constraints that may affect the resource allocation to each single WP of the WBS.

 This may “hide” the real project critical path and therefore the overall project duration. This study aims to provide a contribution to the understanding the rationale of workforces allocation in the construction that is essential to define the WP work duration. To do that, we start analyzing the ex post resources data of the typical Work Packages (WP) of a real infrastructure (a power plant) i.e.: civil works, steel structure erection, piping network erection and cable laying. After the data split and their analysis, each WP has been simplified into its essential terms and its causal relationships. The simplified construction process was used to tune up a dynamic simulation model. The model will reproduce the pattern of workforce resources and the progress of the construction of the plant in the initial planning of the project. The following step was the simulation of the actual WP conditions in order to check the consistencies between the model and the actual data in the new condition. Through the model, we try to highlight the growth limiting factors of the construction.

Leave a Comment

Your email address will not be published. Required fields are marked *