A decision -support model for selection of a trenchless construction method

The increasing demand for trenchless services over the past fifteen years has resulted in the development of numerous methods for installing and replacing buried pipelines and conduits. Each of these methods has its advantages and limitations, and each is best suited for a particular combination of design requirements, project environment, and subsurface conditions. Previous research and experience has shown that evaluation of various competing construction methods should not be limited to direct costs, but must also account for social and indirect costs, environmental impact as well as the inherent risks associated with underground construction. Accounting for these factors requires a systematic evaluation methodology of the unique characteristics of each method in an unbiased manner. An automated decision-support system can assist the decision-maker in performing this immense task by breaking it into manageable subtasks that can then be automated.

This work describes a framework for a decision-support model for the selection of a construction method for the installation or replacement of buried pipelines and conduits. Review of the literature revealed that a number of well-established decision-making methodologies suffer from some inherent shortcomings when employed in the development of construction method evaluation models. A new algorithm named I.M.P.E.C.T (Innovative Modular Procedure for Evaluation of Construction Technologies) was developed to overcorne some of these shortcomings. I.M.P.E.C.T combines well-established concepts in decision-making theory together with innovative algorithms developed during the course of this study to provide a comprehensive decision support environment for the decision-maker.

In contrast to ‘black-box’ decision support models, when fully implemented, I.M.P.E.C.T will provide the decision-maker with the opportunity to gain an insight into the potential difficulties associated with the project and the quality of the available information. Additionally, the user is required to identify his or her priorities and constraints. Selected aspects of I.M.P.E.C.T were implemented using MS Excel and MATLAB Version 5.3. Initial verification of the model demonstrates the ease of automation, the relative simplicity of user input and the straight forwards, but yet informative, nature of the output. Continuing research in this area will focuses on completing the automation of the framework, further population of the database and a rigorous validation process via the modeling of a series of case histories.