Sustainable asset and project management

Abstract

There is a growing demand for companies to report and demonstrate their environmental credentials and corporate responsibility, which presents an opportunity for them to differentiate and gain a competitive advantage in the marketplace. They are now recognizing the limited capacity of the environment to endure the current level of development and economic growth, depletion of natural resources, increasing problem of waste, worrying carbon dioxide emissions, and other environmental impacts. In fact, for asset and project managers, financial criteria are no longer the sole considerations to achieve success and shareholder value. Therefore, environment is being considered as a future source of risk or opportunity. The present research proposes methodology and mathematical models for a sustainable asset and project management, with the focus on the environmental aspect of sustainable development and more specifically the issue of greenhouse gas (GHG) emissions. The following models have been presented in this dissertation. First, a mathematical fleet optimization model is developed, which incorporates the environmental impacts of a fleet of assets over a finite horizon, in addition to its total cost of ownership. As a unique feature of the model, it allows the assets to be kept in storage over any time period, in which such assets do not deteriorate as in-use assets do. The mathematical model optimizes the number of new, in-use, in-storage, and salvaged assets in each time period, so that the total economic costs and environmental impacts are minimized. The application of this work is illustrated in a fleet of excavators. Second, a hybrid Bayesian network (BN) is proposed for fleet availability analysis, focusing on the uncertainty of assets failure and repair rates. We model the common causes to individual rates, as well as the common causes that affect both failure and repair rates at the same time. The proposed model explicitly quantifies uncertainty in repair and failure rates of a fleet of assets and provides an appropriate method for modeling complex dependencies and factors affecting reliability, maintainability, or both, by considering influencing factors, either technical (such as working temperature, environment, quality, stress, etc.) or organizational (such as staff quality, management policies, etc.). We will then extend the model to consider extremely rare and/or previously unobserved risks (e.g. heavy storms, droughts, floods, etc.) that can significantly weaken reliability or maintainability levels. Third, a deterministic model for equipment repair-replacement (R/R) decision with both economic and environmental considerations is formulated. We converted the model into an algorithm and an automatic R/R Calculator. A probabilistic version of this model is then developed to factor in the quality of preventive maintenance, repair perfection, and risk events. We also model the causal relationship between equipment reliability and its GHG emissions during the operation phase. A plastic shredder case study was used to present the models’ results. Fourth, we aim to capture the uncertainty of carbon price in the Western Climate Initiative (WCI) market, by determining the causality between carbon price and its driving forces. A probabilistic model is developed using BNs to infer the possible ranges of each driving force that could have an escalation/depreciation effect on price as well as the magnitude of this effect. The model is developed and run based on a database of historical and projection on the selected driving factors in all the jurisdictions of the WCI market, providing the most probable price(s) over the next ten years. Finally, we developed two models to estimate and control project GHG emissions. The first model is developed based the earned value management (EVM) technique, a common practice in project cost and schedule performance measurement. The proposed model provides project managers with metrics to measure project GHG performance at any point in time over the life of a project and forecast the final emissions. In addition, we proposed a probabilistic model to quantify the uncertainty of project GHG emissions using Monte Carlo Simulation and BN techniques. The model provides a quantitative risk analysis mechanism to estimate the total emissions of the project as well as prediction of final emissions during the implementation process. The proposed models are applied to a work package of a real construction project.