Tarek Zayed

Hierarchical Fuzzy Expert System for Risk of Failure of Water Mains

In Canada and the United States, there have been 700 water main breaks per day costing more than CAD 6 billion since 2000. Risk of failure is defined as the combination of probability and impact severity of a particular circumstance that negatively impacts the ability of infrastructure assets to meet municipal objectives. The presented research in this paper assists in designing a framework to evaluate the risk of water main failure using hierarchical fuzzy expert system (HFES). This system considers 16 risk-of-failure factors within four main categories representing both probability and negative consequences of failure. Results show that pipe age confers a strong impact on risk of failure followed by pipe material and breakage rate. They also show that damage to surroundings has the most negative consequence of a failure event. A set of municipal water network data are collected and used to examine the developed HFES. According to the proposed scale of risk of failure, about 8.4% (13 km) of the network’s...

Fuzzy-Based Life-Cycle Cost Model for Decision Making under Subjectivity

AbstractDecision support models are needed to facilitate long-term planning and priority setting among competing alternatives. Life-cycle cost is the most frequently used economic model that considers all cost elements and related factors throughout the service life of the alternatives being considered. These cost elements and related factors are usually associated with uncertainty and subjectivity. As such, it is important to model the uncertainty arising from the assumed data over the service life of competing alternatives. Probabilistic techniques, such as Monte Carlo simulation, are commonly used to deal with such uncertainty or vagueness. However, they have been criticized for their complexity and amount of data required. This paper presents a fuzzy-based life-cycle cost model that accounts for uncertainty in a manner that disadvantages commonly encountered in probabilistic models are alleviated. The developed model utilized fuzzy set theory and interval mathematics to model vague, imprecise, qualita...

Comparative Analysis of Life-Cycle Costing for Rehabilitating Infrastructure Systems

Metro systems usually offer an attractive alternative for mass transit transportation system in most large cities. Such infrastructures require proper maintenance and rehabilitation (MR (2) continuous rating approach; and (3) dynamic or time-dependent TPM. Results revealed that the continuous rating approach provides lower values compared to the traditional approach. Dynamic TPM reflects better the infrastructure behavior but necessitate additional data gathering. This research mainly benefits metro management agencies and enhances the MDP practice by overcoming some downsides of the traditional methodology.

ASSESSMENT MODEL OF WATER MAIN CONDITIONS

One of the greatest challenges facing municipal engineers is the condition assessment of buried infrastructure assets. It is a mandatory process to establish and employ management strategies for these assets. Condition assessment of water mains is challenging compared to other infrastructure assets because they are typically underground, operated under pressure, and mostly they are inaccessible. To assess the condition of water mains, current research considers physical, environmental, and operational factors and their effect on different types of water mains. A condition assessment model, using the analytic hierarchy process (AHP), is developed in order to set up rehabilitation priority for water mains. Various factors are incorporated in the developed model, such as physical (pipe type, size, age, breakage rate), environmental (Cathodic protection, ground water level, soil type, surface type, and road type), and operational (Hazen-Williams factor, operational pressure). Data, which are collected from municipal engineers who are experts in water system, include pair-wise comparison matrices among factors and their sub-factors. The AHP procedure is applied to these pair-wise comparison matrices in order to generate the relative weights of each factor on a scale out of 1.0. A model is developed to determine the condition of water main based on the AHP results. Each factor weight represents the relative importance of this factor among other factors that affect water main condition. Results show that pipe age has the highest relative contribution factor among others (20.95%); then pipe material (17.49%); however, the third factor is the breakage rate (13.13%). On the other hand, the least factor is type of service (2.85%). The developed model will assist municipal expertise to prioritise pipe inspection and rehabilitation planning for their existing water mains.

Contractor Selection Model for Highway Projects Using Integrated Simulation and Analytic Network Process

AbstractSelecting the appropriate contractor is a significant step to highway project success. The process of selection is subjected to uncertainty and influence of criteria other than bid price. This paper presents an approach to prioritize competitive contractors at the prebidding stage for highway projects by utilizing the analytic network process (ANP) and Monte Carlo simulation. Both techniques are integrated on a single platform to build the proposed model. Initially, the main quantitative and qualitative criteria affecting the contractors’ selection process are identified and studied. The effective criteria in the selection process were obtained from experts and literature. A structured questionnaire was designed and sent to experts in highway projects. The ANP was used to prioritize these criteria subsequently, and Monte Carlo simulation was utilized to develop the selection model. The applicability of the proposed model was tested using four real cases of highway projects. It was concluded that u...

Structural Condition Assessment of Sewer Pipelines

The need of immediate supportive measures for sustainability of municipal infrastructures calls for better understanding of the behavior of various infrastructure network systems and their components. This paper presents a study which uses artificial neural networks to investigate the importance and influence of certain characteristics of sewer pipes upon their structural performance, expressed in terms of condition rating. In this study, back propagation and probabilistic neural network NN models were developed and validated. The data used in the development of these models were provided by the municipality of Pierrefonds, Quebec. It comprised of parameters related to sewer pipelines, pipe diameter, buried depth/cover, bedding material, pipe material, pipeline length, age, and closed circuit television CCTV based structural condition rating. The first six parameters are the independent variables of the models whereas CCTV based condition rating for these pipes is the dependent variable i.e., the output of the models. The developed NN models were used to rank the parameters, in order of their importance/influence on pipe condition. It was found that, among the studied parameters, material attributes have highest influence on pipe structural condition, respectively, followed by the geometric and physical attribute group. Sensitivity analysis was then performed to simulate the structural condition of a pipe at a range of values of each input parameters. Results of sensitivity analysis describe the nature and degree of the influence of each parameter on pipe structural condition. The developed models are expected to benefit academics and practitioners municipal engineers, consultants, and contractors to prioritize inspection and rehabilitation plans for existing sewer mains. DOI: 10.1061/ASCECF.1943-5509.0000081 CE Database subject headings: Neural networks; Assessment; Parameters; Sensitivity analysis; Sewer pipers; Rehabilitation. Author keywords: Neural networks; Condition assessment; Parameters; Sensitivity analysis; Sewer mains.

Deterministic models for assessing productivity and cost of bored piles

The assessment process of productivity and cost of bored pile construction is dictated by unseen subsurface obstacles, lack of contractor experience and site planning. These problems complicate the estimators role in evaluating pile equipment productivity and cost. Current research discusses the assessment of piling process productivity and cost using the deterministic technique. Data are collected through questionnaires, site interviews and telephone calls to experts in various construction companies. Many variables have been considered in the piling construction process, such as pile size, depth, pouring method, soil type and construction method. Five deterministic models have been designated to assess productivity, cycle time and cost. The developed models are validated whereas 79% of the outputs have been predicted with more than 75% accuracy. Consequently, three sets of charts have been developed to provide the decision‐maker with a solid planning, scheduling and control tool for piling projects. If a pile has 60′ depth with φ‐18 (18″ diameter pile) in clay soil using a 5′ auger height, the cycle time is estimated as 56 and 65.5 minutes; however, productivity is 6 and 5 holes/day for dry and wet methods, respectively.

Method for Analyzing Time-Series GPR Data of Concrete Bridge Decks

AbstractGround-penetrating radar (GPR) has been extensively studied in North America as a nondestructive evaluation (NDE) technology for inspection of concrete bridge decks. With current practices, however, GPR has only proven to be an indicator of potential damage. Basically, to obtain the condition map for a concrete bridge deck, one would try to analyze one-time GPR data based mostly on the relative difference between reflection amplitudes at the top rebar layer. With a hypothesis that time-series GPR data can provide better information on bridge deck deterioration progression, this study investigates and proposes a new method to interpret those time-series data sets. Based on a correlation coefficient between A-scans, the proposed methodology was implemented and validated for a bare concrete bridge deck in New Jersey. The map provided by the proposed method clearly shows deterioration progression between the two consecutive scans, whereas the traditional analysis technique using the top rebar amplitud...

Fuzzy-Based Model for Predicting Failure of Oil Pipelines

AbstractOil and gas pipelines transport millions of dollars of goods worldwide every day. Even though they are the safest way to transport petroleum products, pipelines do still sometimes fail, generating hazardous and irreparable environmental damages. Many models have been developed in the last decade to predict pipeline failures and conditions. However, most of these models were limited to one failure type, such as corrosion failure, or relied mainly on expert opinions. The objective of this paper is to develop a fuzzy-based model to predict the failure type of oil pipelines using historical data of pipeline accidents. The model was able to satisfactorily predict pipeline failures due to mechanical, operational, corrosion, third-party, and natural hazards with an average percent validity of 83%. This research contributes to the body of knowledge by developing a robust failure type prediction model for oil pipelines using a fuzzy approach. The model will assist pipeline operators to predict the expected...

Condition Prediction Models for Oil and Gas Pipelines Using Regression Analysis

Although they are the safest means of transporting oil and gas products, pipelines can sometimes fail with hazardous consequences and large business losses. The decision to replace, repair, or rehabilitate depends mainly on the condition of the pipeline. Assessing and predicting its condition is therefore a key step in the maintenance plan of a pipeline. Several models have recently been developed to predict pipeline failures and conditions. However, most of these models were limited to the use of corrosion as the sole factor to assess the condition of pipelines. The objective of this paper is to develop models that assess and predict the condition of oil and gas pipelines based on several factors including corrosion. The regression analysis technique was used to develop the condition prediction models based on historical inspection data of three existing pipelines in Qatar. In addition, a condition assessment scale for pipelines was built based on expert opinion. The models were able to satisfactorily predict pipeline condition with an average percent validity above 96% when applied to the validation data set. The models are expected to help decision makers assess and predict the condition of existing oil and gas pipelines and hence prioritize their inspection and rehabilitation planning.