Abstract

Oil and gas pipelines are the main means of transporting fossil fuels from the wellheads and processing facilities to the distribution centers. The 2013 US infrastructure report card assigned a grade of D+ to energy pipelines signifying they are in a poor condition. More than 10,000 incidents were reported on oil and gas pipelines during the last two decades, most of which resulted in considerable consequences. Recent failures and ruptures have raised concerns over the risk of failure of such pipes in Canada. The main objective of this research is to develop a risk-based maintenance planning model for oil and gas pipelines. The research develops a probability of failure (POF) and a consequence of failure (COF) prediction model and establishes a risk-based inspection and simulation-based rehabilitation planning models.
The POF model develops a comprehensive index by applying the granular theory of uncertainty and the principles of probability theory to forecast the POF of oil and gas pipes. The neuro-fuzzy technique is employed to develop a model that forecasts the financial consequences of the potential failures of such pipes. An integrated fuzzy risk evaluation model is developed with 25 fuzzy rules to assess a pipeline’s risk index. A fuzzy expert system is developed to select the inspection tools and determine their run-frequency according to the failure risk of a pipeline. Regression analysis is applied to develop a risk growth profile to forecast the maximum failure risk of various inspection scenarios. Scenarios are ranked based on their risk-cost index, which integrates two main indices: 1) maximum risk of failure, and 2) life cycle cost of scenarios, computed by applying the Monte-Carlo simulation. Finally, a comprehensive maintenance model proposes the optimum maintenance plans with the lowest LCC, developing a third-degree risk-based deterioration profile of the pipelines.
The POF model’s sensitivity results highlight that cathodic protection effectiveness and soil resistivity are the leading causes of external corrosion failures, while the depth of cover is an important factor of mechanical damages. The COF model attests that diameter, as well as the location properties are important factors for estimating the financial consequences. The developed risk assessment model is validated using a test dataset that proved the models are accurate with about 80% validity. The developed models are applied on a case study of a 24-inch pipe. The POF and COF of the pipe are computed, and the results suggest that the pipe’s risk index is above medium with an average index of 3.5. The study proposes the application of an inspection tool, which decreases the risk growth by 50% during the service life of the pipeline. The application of the maintenance planning model proposes a combination of recoat, repair, and replacement with a medium size of rehabilitation. The net present value of the proposed scenario of maintenance is estimated to cost around 1.7 million dollars over the life cycle of the pipeline, compared to the last-ranked alternative that costs over three million dollars. This research offers a framework to develop a comprehensive index to predict the failure risk of pipes using historical data that can be extended to the other infrastructure types. It develops a model to plan for the optimal pipeline maintenance, and provides an overall image of its service life. The developed models will help the operators predict the risk of failure and plan appropriately for the life cycle of their oil and gas pipelines.