Probability Of Failure Research Articles

Optimising batch withdrawal strategies for non-safety-related defects under risk sensitivity

A significant challenge in warranty-protected products is determining the appropriate strategy for removing batches that are vulnerable to non-safety-related failures. Two main approaches exist for handling such failures: an active strategy that compensates customers for current faulty units but preemptively withdraws remaining batch units to prevent future warranty claims, and a passive strategy that compensates customers throughout the warranty window without removing the batch. Under two different risk quantification methods, we examine the connection between a product's warranty configuration and the optimal timing strategy for withdrawing a batch that is susceptible to non-safety-related defects. The first method reveals a deterministic optimal withdrawal strategy, and we identify conditions where a passive strategy is optimal, irrespective of warranty length or batch size. In contrast, the optimal strategy under the second method is a probabilistic decision rule that depends on a random event's outcome. We also derive the maximum likelihood estimator for the parameters governing the product's failure probability, and use numerical analysis to explore the performance cost of dynamically learning these parameters. The results suggest that maximum likelihood estimation is a robust method for identifying the parameters of the failure probability function when these parameters are initially unknown during the warranty period.

Quantitative Risk Assessment of Hydrotreated Vegetable Oil at an Oil and Gas Company

Introduction: An oil and gas refinery operates various equipment with specific functions for different processes. Each piece of equipment has potential hazards that can damage the equipment and injure or kill workers. This study focuses on the hydrotreated vegetable oil (HVO) export facility from the jetty loading area at an oil and gas company that processes flammable liquid using various equipment. Methods: The HAZOP method determined the hazardous spots, and the probability of each equipment failure corresponding to the system was also determined using fault tree analysis (FTA). Furthermore, every event tree analysis (ETA) output probability was also determined. The probability and radius of pool fire varied for different leak hole scenarios. The final steps are individual risk per annum and potential loss of life to measure the risk level of the system. Results: Based on HAZOP deviation scenarios, every operating equipment can potentially cause a pool fire. In FTA, scenarios were developed based on different leakage hole sizes, ranging from 1-3 mm, 3-10 mm, 10-50 mm, 150 mm, and >150 mm. The results indicated that leakage could occur across all operating equipment. Similarly, the ETA applied the same bore size scenarios. The consequence analysis yielded a worst-case outcome of pool fire and a best-case outcome of un-ignited fluid release. Subsequently, the pool fire output was modeled using ALOHA, which resulted in three heat flux zones: the red zone (10 kW/m²), the orange zone (5 kW/m²), and the yellow zone (2 kW/m²). Smaller leak holes had a higher probability but smaller pool fire radius. The initial risk of the export facility was unacceptable. Furthermore, insufficient safeguards contribute significantly to the resulting high-risk level. Two mitigations were implemented: adding safeguards and reducing worker hours. Conclusion: The final results showed that for every piece of equipment, the overall risk of the export facility became acceptable after mitigation..

Stochastic Risk Assessment Framework of Deep Shale Reservoirs by a Deep Learning Method and Random Field Theory

Risk assessment of deep shale reservoirs is very important for subsurface energy development. However, due to complex geological environments and physicochemical interactions, shale reservoir fabric parameters exhibit variability. Moreover, the actual investigation and testing information is very limited, which is a typical small-sample problem. In this paper, the heterogeneity and statistical characteristics of deep shale reservoirs are clarified by the measured mechanical parameters. A deep learning method for deep shale reservoirs with limited survey data information is proposed. The variability of deep shale reservoirs is characterized by random field theory. A variable stiffness method and stochastic analysis method are developed to evaluate the risk of deep shale reservoirs. The detailed workflow of the stochastic risk assessment framework is presented. The frequency distribution and failure risk of deep shale reservoirs are calculated and analyzed. The risk assessment of deep shale reservoirs under different model parameters is discussed. The results show that a stochastic risk assessment framework of deep shale reservoirs, using a deep learning method and random field theory, is scientifically reasonable. The scatter plots of the elasticity modulus (EM), cohesive force (CF), and Poisson ratio (PR) distribute along the 45-degree line. The different distributed variables of EM, CF, and PR have a positive correlation. The statistical properties of the measurement data and deep learning data are approximately the same. The principal stress of deep shale follows the normal distribution with significance level 0.1. Under positive copula conditions, the maximum failure probability is 5.99%. Under negative copula conditions, the maximum failure probability is 4.58%. Different copula functions under positive and negative copula conditions have different failure probabilities. For the exponential correlation structure, the minimum failure probability is 3.46%, while the maximum failure probability is 6.19%. The mean failure probability of the EM, CF, and PR of deep shale reservoirs is 4.85%. Different random field-related structures and parameters have different influences on the failure risk.

Soil Strength Parameters for the Sustainable Design of Unsupported Cuts Under Drained Conditions Using Reliability Analysis

Unsupported excavations are frequently performed in several geological and geotechnical projects, particularly for constructing roads and railways, and they are often carried out in different materials. The design of such cuts in soils needs the determination of representative values of its mechanical properties, particularly of the strength parameters, and the application of adequate safety factors. The procedure should ensure a sustainable design of those cuts, allowing for economical solutions that guarantee a low probability of geological–geotechnical failure. This paper assesses the reliability of unsupported cuts in soils, under drained conditions, assuming a Mohr–Coulomb strength criterion. Statistical meshes are generated considering the spatial variability of the friction angle and of the true effective cohesion, which are assumed to be uncorrelated. In this process, typical values of the coefficients of variation and of the horizontal and vertical scales of fluctuation are applied. Soil characterisation is simulated in each statistical mesh, and the characteristic values of the strength parameters are determined using statistical methods. Unsupported cuts of different heights and inclinations are designed using typical safety factors. Slope stability analyses are carried out using Random Finite Element Limit Analysis. The uncertainty in the actions is considered, and the probability of failure is determined by direct reliability analysis. The results show the relevance of the ratio between the scale of fluctuation and the excavation depth, the slope inclination, and the characteristic value of the soil strength parameters on the probability of failure. Values of adequate safety factors are proposed towards obtaining an appropriate probability of failure, compatible with the sustainable design of the cuts.

Regional bridge risk assessment due to the combined effect of flooding and overloading events in Manitoba

The challenges experienced by Canada’s aging infrastructure require probabilistic methods to address the threats that bridge systems are exposed to. This study provides a framework to evaluate risk associated with two major bridge hazards such as flooding and overloading in Manitoba. The probability of failure of highway bridges due to overloading is assessed as a function of traffic volume and overweight percentages. The probability of failure due to flooding is obtained from a spatial analysis of the water levels. Finally, consequences of failure are established from the analysis of insurance exposure, including climate change effects. Historic data are projected ten years into the future through an ARIMA model. The results of the analysis show that the risk for bridges in the south-west regions of Manitoba is expected to increase by 2035 up to 40% for water levels of 3.0 m and up to 0.6% for a threshold of 12.0 m.

Determining Targets for Antiretroviral Drug Concentrations: A Causal Framework Illustrated With Pediatric Efavirenz Data From the CHAPAS-3 Trial.

Determining a therapeutic window for maintaining antiretroviral drug concentrations within an appropriate range is required for identifying effective dosing regimens. The limits of this window are typically calculated using predictive models. We propose that target concentrations should instead be calculated based on counterfactual probabilities of relevant outcomes and describe a counterfactual framework for this. The proposed framework is applied in an analysis including longitudinal observational data from 125 HIV-positive children treated with efavirenz-based regimens within the CHAPAS-3 trial, which enrolled children < 13 years in Zambia/Uganda. A directed acyclic graph was developed to visualize the mechanisms affecting antiretroviral concentrations. Causal concentration-response curves, adjusted for measured time-varying confounding of weight and adherence, are calculated using g-computation. The estimated curves show that higher concentrations during follow-up, 12/24 h after dose, lead to lower probabilities of viral failure (> 100 c/mL) at 96 weeks of follow-up. Estimated counterfactual failure probabilities under the current target range of 1-4 mg/L range from 24% to about 2%. The curves are almost identical for slow, intermediate and extensive metabolizers and show that a mid-dose concentration level of ≥ 3.5 mg/L would be required to achieve a failure probability of < 5%. Our analyses demonstrate that a causal approach may lead to different minimum concentration limits than analyses that are based on purely predictive models. Moreover, the approach highlights that indirect causes of failure, such as patients' metabolizing status, may predict patients' failure risk, but do not alter the threshold at which antiviral activity of efavirenz is severely reduced.

Deficient postoperative anterior wall index adversely affects joint survival after transposition osteotomy of the acetabulum in patients with hip dysplasia.

This study aimed to determine clinical outcomes; relationships between postoperative anterior, lateral, and posterior acetabular coverage and joint survival; and prognostic factors for joint survival after transposition osteotomy of the acetabulum (TOA). Data from 616 patients (800 hips) with hip dysplasia who underwent TOA between November 1998 and December 2019 were reviewed. The median follow-up period was 8.9 years (IQR 5 to 14). A medical notes review was conducted to collect demographic data, complications, and modified Harris Hip Score (mHHS). Radiological indicators of acetabular coverage included lateral centre-edge angle (LCEA), anterior wall index (AWI), and posterior wall index (PWI). The cumulative probability of TOA failure (progression to Tönnis grade 3 or conversion to total hip arthroplasty) was estimated using the Kaplan-Meier product-limited method. A multivariate Cox proportional hazards model was used to identify predictors for failure. Median mHHS improved from 68 preoperatively to 96 at the latest follow-up (p < 0.001). The overall joint survival rate was 97% at ten years and 70% at 20 years. For the postoperative LCEA subgroups, survival in the deficient group was lower than that in the excessive (p = 0.006) and normal (p = 0.007) groups. For the postoperative AWI subgroups, survival in the deficient group was lower than that in the excessive (p = 0.015) and normal (p < 0.001) groups. Multivariate analysis identified age (p = 0.010), Tönnis grade 2 (p < 0.001), roundness index (p = 0.003), fair joint congruity (p = 0.004), and postoperative AWI (p = 0.002) as independent risk factors. Deficient postoperative AWI adversely affected joint survival after TOA, underscoring the importance of sufficient anterior acetabular coverage, along with precise surgical indications, to ensure successful hip joint preservation in the treatment of hip dysplasia.

Effectiveness of bictegravir/emtricitabine/tenofovir alafenamide (BIC/FTC/TAF) as switch strategy in virologically-suppressed patients: real world data from a monocentric cohort.

BIC/FTC/TAF showed efficacy and tolerability in randomized trials as a switch strategy in virologically-suppressed people living with HIV. We evaluated its effectiveness in a real-life setting. A retrospective monocentric cohort including 431 virologically-suppressed (HIV-RNA <50 copies/ml) people switching to BIC/FTC/TAF in the period 2018-2022 was evaluated. Probabilities of virological failure (VF, i.e.2 consecutive HIV-RNA ≥50 copies/ml or a single HIV-RNA ≥200 copies/ml) and of treatment discontinuation (TD) were estimated by Kaplan-Meier, and predictors of both outcomes were identified through multivariable Cox regression. Analysis-of-variance for repeated measures was used to examine changes in CD4 count and CD4-to-CD8 ratio. Overall, 16 VF occurred during 22 months of median follow-up time. Estimated probabilities of VF at 1, 2 and 3 years were 2.0% (95% CI 1.04.2%), 2.9% (95% CI 1.5%-5.6%) and 5.5% (95% CI 3.2%-9.2%), respectively. Caucasian ethnicity and a history of previous VF independently predicted VF. TD occurred in 42 cases, predominantly for simplification. One discontinuation due to VF was reported. No predictors of discontinuation were identified. An increase in CD4-to-CD8 ratio over 3 years was evidenced (p < 0.001). Total cholesterol decreased over 3 years (p < 0.001). Triglycerides did not significantly change (p = 0.465). BIC/FTC/TAF demonstrated high effectiveness, tolerability and safety.

Comparative analysis of deep learning models for various nonalcoholic fatty liver disease datasets

Fatty liver disease is caused by increased liver buildup or weight above 5-10%. This disorder is widespread in people with diabetes, overweight persons, and metabolic syndrome patients. Clinical decision support systems can improve liver failure diagnosis and prediction to reduce this situation. Many liver failure models have drawbacks, and liver failure prediction is still a problem. This work uses four large open-access critical care patient datasets to create and verify liver failure risk prediction models. This study aims to construct a clinically applicable diagnostic and predictive model that evaluates the probability or risk of liver failure in intensive care unit (ICU) patients using extreme gradient boosting (XGBoost), artificial neural networks (ANN), multi-layer perceptron (MLP), Modular Neural Network (MNN), and generalized feed forward (GFF). We evaluated performance metrics using these models: accuracy, sensitivity, specificity, and predictive accuracy.

Cost-effectiveness of viral load testing for transitioning antiretroviral therapy-experienced children to dolutegravir in South Africa: a modelling analysis

For children with HIV on antiretroviral therapy (ART), transitioning to dolutegravir-containing regimens is recommended. The aim of this study was to assess whether introducing viral load testing to inform new nucleoside or nucleotide reverse transcriptase inhibitors (NRTIs) for children with HIV and viraemia alongside dolutegravir-based ART is beneficial and of good economic value. We used the Cost-Effectiveness of Preventing AIDS Complications-Pediatric model to project clinical and cost implications of three strategies among a simulated cohort of South African children aged 8 years with HIV receiving abacavir-lamivudine-efavirenz: (1) continue current ART (no dolutegravir; abacavir-lamivudine-efavirenz); (2) transition all children with HIV to dolutegravir, keeping current NRTIs (dolutegravir; abacavir-lamivudine-dolutegravir); or (3) transition to dolutegravir based on viral load testing (viral load plus dolutegravir), keeping current NRTIs if virologically suppressed (abacavir-lamivudine-dolutegravir, 70% of cohort) or switching abacavir to zidovudine (zidovudine) if viraemic (zidovudine-lamivudine-dolutegravir, 30%). We assumed 50% of children who had viraemia after abacavir-lamivudine exposure had NRTI resistance; with resistance, we assumed zidovudine-lamivudine-dolutegravir was more effective than abacavir-lamivudine-dolutegravir. We designated a strategy as preferred if it was most effective and least costly or had an incremental cost-effectiveness ratio less than half the South African 2020 gross domestic product per capita. Under base-case assumptions, the viral load plus dolutegravir strategy would be the most effective (projected undiscounted life expectancy of 39·72 life-years) and least costly strategy (US$24 600 per person); the no dolutegravir strategy was the least effective (34·49 life-years) and most expensive ($26 480 per person). In sensitivity analyses, the 24-week virological suppression probability and subsequent monthly virological failure risks (ie, late failure) were most influential on cost-effectiveness. Only with a high late-failure risk for zidovudine-lamivudine-dolutegravir (ie, ≥0·3% per month in the base case or >0·5% per month if abacavir also confers low virological suppression probability in the presence of NRTI resistance [65%]) would the dolutegravir strategy become preferred above the viral load plus dolutegravir strategy. For programmes transitioning to dolutegravir-based regimens, our model predicted that doing so would be more effective and less costly than continuing current ART regimens, regardless of NRTI choice. Whether viral load testing for children with HIV is necessary to inform NRTI choice depends substantially on the comparative outcomes of abacavir and zidovudine after switching to dolutegravir-containing ART. The Eunice Kennedy Shriver Institute for Child Health and Human Development, the National Institute of Allergy and Infectious Diseases, the Massachusetts General Hospital Executive Committee on Research, the Massachusetts General Hospital, and the Medical Research Council.

An efficient algorithm for estimating Profust failure probability function under the assumption of probable input and fuzzy state

An efficient algorithm for estimating Profust failure probability function under the assumption of probable input and fuzzy state

Dynamic risk analysis of fire and explosion domino accidents at hydrogen refueling stations using Dynamic Bayesian Network

Hydrogen is a promising energy source and hydrogen refueling stations (HRS) are the main hydrogen supply infrastructures. Unwanted hydrogen leaks and releases at the hydrogen station may cause serious explosion accidents and even induce domino effects due to intensive hazardous equipment in the station. However, scant attention has been accorded to the dynamic risks of fire, explosion, and domino effects at HRS, posing a threat to the safe functioning of HRS and the advancement of the hydrogen energy. This study thus first develops Dynamic Bayesian networks to model the primary fire and explosion accidents as well as domino effects at HRS. By the developed models, the critical factors that lead to accidents and the critical equipment that contributes to the initiation or propagation of domino effects can be identified. Moreover, the failure probability of different equipment exposed to possible accidents can also be obtained, supporting safety management of HRS.

Probabilistic analysis framework for the deterioration of marine tubular steel structures based on random field simulation

Corrosion reduces the effective cross-section, leading to the bearing capacity loss of marine tubular steel structures; the structures’ time-variant performance are of great concern in practices. The paper employs random field simulation to account for corrosion's spatial variability in evaluating the bearing capacity of the axially-loaded steel tubes, and proposes a time-variant reliability analysis framework for marine tubular steel structures, from the member-level to the structure-level. Results are compared with those neglecting the corrosion's spatial variability. The proposed method prevents a significant underestimation of the failure probability comparing to that neglecting spatial variability, and provides a tool for designers to predict the time-variant reliability levels for marine steel structures.

Experimental research on fatigue behavior of concrete under acid rain corrosive environment

Concrete infrastructures such as high-speed railway bridges, road pavements, and airport runways in acid rain areas are severely affected by fatigue loading and environmental corrosion. To investigate the impact of acid rain corrosion on the fatigue resistance of concrete, 12 groups of concrete specimens were exposed to acid rain spray tests and then subjected to bending fatigue testing. These tests comparatively examined the impact of different stress levels and corrosion ages on the concrete's fatigue life, strain, and damage pattern. Additionally, the fatigue life reliability of the concrete was analyzed using the Weibull distribution theory and equations were developed to predict the fatigue life of the concrete under different probabilities of failure. The results indicate that both stress level and acid rain corrosion age significantly affect the ultimate strain and fatigue life of concrete. High stress levels combined with prolonged acid rain exposure severely degrade the fatigue performance of the concrete. The fatigue life of the concrete follows a Weibull distribution, and the P-S-N curves and fatigue life equations developed based on data from various failure probabilities show a high linear association with values larger than 0.9. These findings can serve as a reference for predicting the fatigue life of concrete in areas affected by acid rain.

An integrated system theoretic accident model and process (STAMP)-Bayesian network (BN) for safety analysis of water mist system on tanker ships

A water mist on tanker ships is a specialized fire suppression system intended to mitigate fire hazards on board. This innovative safety measure disperses microscopic water droplets, creating a mist that rapidly absorbs heat and suppresses fires by reducing oxygen levels and cooling the surrounding area autonomously. This paper attempts to perform a comprehensive safety analysis of the water mist system on a tanker ship under the system theoretic accident model and process (STAMP) and Bayesian network (BN) robust modeling, which is capable of presenting a dynamic structure comprising complex elements. In the paper, whilst the STAMP is utilized to determine failure scenarios by establishing a hierarchical control and feedback model structure, the BN quantifies the probabilities of potential failures based on the outcomes of STAMP. A water mist system in tanker ships is analyzed under STAMP-BN modeling since it is an essential part of the ship, including an advanced system comprising elements of human, software, and hardware. The research results indicate that the failure probability of the water mist system has a failure probability of 3.93E-02. Besides its robust theoretical background, the research findings will provide a helpful reference for maritime safety managers, safety inspectors, technical inspectors, HSEQ managers, and ship crew on improving fire-fighting safety in tanker ships.

Probabilistic stability analysis of inclined strip anchors in spatially variable clay

The stability of plate anchors has attracted significant attention, but most existing solutions are derived using deterministic soil properties. This paper demonstrates how natural spatial variability affects the undrained pullout resistance of inclined strip anchors embedded in clays. Both vented and attached bases are considered. Analysis uses an adaptive finite element limit analysis with random field model in Monte Carlo framework. The pullout resistance statistics and failure mechanisms of shallowly and deeply embedded anchors are presented for different anchor inclinations. Safety factors to achieve target levels of failure probability are proposed for the inclined anchors. It is revealed that ignoring the random nature of clay strength gives rise to the overprediction of pullout resistance of inclined anchors and it is more pronounced for deep anchors under attached conditions. The anchor pullout resistance is closely associated to the failure mechanism consisting of several failure planes following the weak clay paths that rely on the spatial pattern of the random clay.

Probabilistic analysis of progressive slope retreats due to weathering: An improved discretized-UBLA approach

Slope retreat phenomena have been a persistent area of interest for researchers; however, the majority of these studies have been conducted within a deterministic framework, thereby overlooking the numerous geotechnical uncertainties inherent in these scenarios. This study addresses this gap by considering the initial slope strength and its degradation rate as random variables, thereby enabling a probabilistic analysis of the progressive sliding behavior of a slope subjected to weathering. An upper-bound limit analysis (UBLA)-based discretized mechanism was established to serve as the deterministic model. This mechanism employs a ’point-to-point’ technique for generating the discretized sliding surface, effectively circumventing the limitations and complexity associated with the classic log-spiral mechanism. When integrated with time-varying slope strength, this model is capable of predicting the occurrence of each sliding event with remarkable accuracy and efficiency. To assess the likelihood of specific sliding events occurring at any given time, a reliability analysis was conducted using Monte Carlo simulation. The framework proposed in this study was validated through two case studies: one involving a homogeneous slope and the other a two-layered slope. These case studies produced a variety of reliability results of great value, including the time-varying failure probability (Pf) for each slide, the distance of slope retreat, and the probabilistic collapse within the slope. Additionally, this study also conducted a system reliability analysis. It is demonstrated that the variability in initial strength and degradation rate exerts contrasting effects on the time-varying slope strength. The variability in initial strength constricts the strength distribution, while the variability in degradation rate tends to expand it. Homogeneous and heterogeneous slopes exhibit distinct characteristics of slip surface degradation. These results provide novel insights into how the probabilistic characteristics of initial strength and degradation rate influence progressive failure. The proposed methodology provides a powerful tool for conducting probabilistic analysis of slope retreat resulting from weathering, and it has significant implications for regions that are prone to landslides.

Multi-objective maintenance strategy for complex systems considering the maintenance uncertain impact by adaptive multi-strategy particle swarm optimization

Effective maintenance optimization strategies are crucial for improving the complex equipment reliability and reducing the maintenance costs. However, the effectiveness of the maintenance procedures applied to industrial equipment is affected by uncertainty, e.g. due to the professional skills of maintenance personnel, the actual condition of the equipment being maintained. The quantification of the uncertainty on the effect of maintenance has practical significance and must be accounted in the development of maintenance strategies for reducing equipment probability of failure. This paper proposes a multi-objective maintenance strategy considering the uncertain impact of the maintenance actions. The impact of maintenance actions on the reliability of components is first studied and a reliability assessment model is developed, which considers the skill of maintenance personnel and the actual condition of the equipment. To optimize the multi-objective maintenance strategy, a multi-strategy particle swarm optimization (MS-PSO) algorithm is proposed. Two case studies are considered to verify the effectiveness of the proposed approach for multi-objective maintenance strategy optimization. In the case studies considered, it turns out that the maintenance cost rate (MCR) is reduced throughout the system life cycle and the cumulative availability is improved.

Computing upper probabilities of failure using optimization algorithms together with importance sampling

AbstractThis paper addresses efficient computation of the upper probability of failure of an engineering structure, when the uncertainty is modeled by a family of probability densities. We develop an algorithm significantly reducing the sample sizes required in the optimization algorithm by adopting a recursive approach. The algorithm is based on importance sampling, reweighting, comparing sample ratios and setting up an iterative scheme that allows one to re‐use previously computed sample points in the optimization. The efficiency of the method is analyzed by means of a moderate scale engineering structure.

Restoration of forestry-drained boreal peatland ecosystems can effectively stop and reverse ecosystem degradation

Ecosystem restoration will increase following the ambitious international targets, which calls for a rigorous evaluation of restoration effectiveness. Here, we present results from a long-term before-after control-impact experiment on the restoration of forestry-drained boreal peatland ecosystems. Our data comprise 151 sites, representing six ecosystem types. Species-level vegetation sampling has been conducted before, two, five, and ten years after restoration. With joint species distribution modelling, we show that, on average, not restoring leads to further degradation, but restoration stops and reverses this trend. The variation in restoration outcomes largely arises from ecosystem types: restoration of nutrient-poor ecosystems has a higher probability of failure. Yet, the ten-year study period is insufficient to capture the restoration effects in slow-recovering ecosystems. Altogether, restoration can effectively halt the biodiversity loss of degraded ecosystems, although ecosystem attributes affect the outcome. This variability in outcomes underlies the need for evidence-based prioritization of restoration efforts across ecosystems.