SIC 2024

The permanent-way (PW) maintenance process involves the use of several assets that support railway-track maintenance. One of the main challenges in this process is asset availability. The equipment is currently maintained every 250 hours and, combined with geographic dispersion, this requirement has generated significant process impacts. This study focuses on analyzing asset reliability and the optimal maintenance interval. As a result, it was possible to extend preventive-maintenance intervals for some assets, maximizing their use by the track-maintenance team. It was also possible to project the number of expected asset failures if preventive-maintenance frequency were increased. The study demonstrated that reliability methodology supports sound decisions that optimize cost and reliability while keeping operational risks under control.

The RAM study performed on the OBATI System at the São Paulo operating unit revealed a surprising reality for the company’s decision-makers. Faced with Petrobras’ requirement to increase transportation capacity through this pipeline, the project appeared capable of generating tens of millions of dollars in revenue over five years. Before a final decision was made, however, the operations team conducted short-duration tests to assess machine behavior under the new logistics demand. Only when the reliability management team became involved, almost at the final decision stage, did the true picture emerge. RAM analysis made it clear that proceeding without equipment modifications would be a mistake. The low reliability identified could lead to severe financial consequences. This case illustrates how RAM analysis not only reveals crucial insights but can also prevent significant organizational losses.

The objective is to perform a RAM (Reliability, Availability, Maintainability) analysis of NS04 — Vale’s first dry concentration plant — using disruptive dry magnetic-separation technology to produce mineral concentrate without water, eliminating the need for dams and mitigating environmental impact. The major challenge of this technology is achieving high feed rates, meeting iron-ore processing requirements and maintaining competitive cost. Reliability Engineering methodologies are essential for the technology to achieve expected results, ensuring that the developed equipment fulfills its design function in terms of performance, risk and cost. The NS04 RAM analysis, applied using the similarity technique, provides detailed data to forecast future budgets accurately, plan spare parts and determine optimal preventive-maintenance intervals. For magnetic separators, reliability applied to product development is fundamental to ensure operational efficiency, product quality, safety and customer satisfaction. Although the RAM simulation results were below expectations, implementing the NS04 project for industrial-scale testing is viable and will also serve as a laboratory for future installations.

This paper analyzes the change in the maintenance program for Brazilian Air Force A-29 aircraft. This approach is necessary because there is no guarantee that fleet performance improved in terms of reliability, failure incidence and operational availability. The objective is to evaluate the effectiveness of the change in the A-29 Aircraft Maintenance Program, verifying whether fleet performance improved with respect to system reliability, failure incidence and availability. This objective was achieved through a literature review and documentary analysis of the maintenance programs before and after the change. The comparative analysis was based on the collection and analysis of failure and availability data from 2017 to 2022. The study showed a positive impact on reliability and on failure incidence for the aircraft system. An increase in fleet availability was also observed after adoption of the new maintenance program. However, when the aircraft systems were evaluated individually, not all benefited from the change, since some showed worse reliability and failure-incidence indicators.

Human reliability can be defined as the probability that a person will successfully perform a task over a specified period under appropriate environmental conditions and with the resources available to execute it. Measuring human error is complex because each person has a unique cognitive profile, particular anthropometric characteristics, experiences and skills that are difficult to quantify. It is also important to recognize the many organizational factors that can induce, facilitate or even promote human failure. In addition, many organizations interpret human error simplistically as merely an indication of incompetence. Because human interaction with systems is unavoidable, the contribution of human reliability must be considered when quantifying overall system reliability. In this context, this paper presents practical cases involving analysis and prediction of human reliability in different scenarios.

In consumer-goods companies such as Ypê, the speed and accuracy of business strategies are determining factors for business sustainability. This operating context challenges industry to be flexible and highly effective in asset operation. This paper presents a step-by-step process for implementing reliability engineering using digitalization technologies to collect and process industrial-asset data in real time. It is essential to establish a work plan that keeps activities aligned with Ypê’s strategic objectives, building a framework of data and information that supports decision-making while balancing Cost, Risk and Performance. Reliability Engineering and Digitalization Technologies are relatively new and complex processes with many challenges, making the environment even more demanding because the people involved often have questions about implementation. To develop deliverables consistently, resources must be used accurately throughout every phase of the implementation process. Therefore, a roadmap aligned with the strategic plan of the asset-management policy is essential, considering the timing and movement of each stage while maintaining employee inclusion and training throughout the innovation process.

As Human Reliability has gained greater visibility in recent years, it has become increasingly important to align concepts and understand what is required to structure a Human Reliability program. The conceptualization phase becomes the most important, including literacy and awareness, always beginning with leadership, where cultural transformation starts. In large organizations such as Vale, this becomes even more challenging because isolated initiatives labeled as “human reliability” can compromise structured program development. Mapping, as one of the stages in this construction, aims to evaluate what will be considered part of the program with a preventive approach, what will remain reactive, and even what may have no direct connection to the program and can be discontinued or not; all cases are assessed during selection of the required data. Dividing reliability between preventive action focused on prediction and reactive action focused on correction requires qualitative and quantitative tools and methods that guide controls over critical tasks within the organization of work.

Given the need to achieve the 2023 ore-production volume together with safety and productivity targets, an analysis was developed covering all assets in each mine and plant production phase — drill, excavator, loader, trucks, primary crushing, storage and shipping — based on advanced reliability studies. The objective was to determine which equipment could affect production in 2023 and propose an action plan to strengthen Asset Management in mining. The result was an avoided loss of 670 thousand tons of ore in 2023.

DEGRADATION ANALYSIS OF OFF-HIGHWAY TRUCK BODY PLATES USING DIGITAL-SCANNING INSPECTION.

The oil and gas industry plays a fundamental role in the global energy market. In Brazil, the sector represents 10% of industrial GDP and is essential to the country’s economic and social growth. Quantitative reliability approaches support optimization of reliability cost based on estimates of equipment and system failure probability and frequency. Repairable systems can be represented by stochastic processes in which events represent failures of a system and its components. Monte Carlo simulation is an established tool for evaluating these processes, especially when assumptions in the mathematical model make analytical calculation impractical. In this case study, the operating-cost scenario of turbogenerators in an oil refinery is evaluated using a methodology that combines Monte Carlo simulation and fault-tree analysis.

Key reliability models applied to thermoelectric systems, with a focus on HVAC applications. The presentation addresses how these technologies operate, common failure mechanisms and life-prediction methods. Topics include thermoelectric applications, the importance of stress cycling to reliability, and how to predict performance over time.

The search for optimal maintenance policies and periodic testing of safety systems is a topic of major interest in industry, especially in the oil and gas sector. A balance between cost and availability of safety functions is constantly sought. When data are scarce and/or non-homogeneous, approaches based on Bayesian inference are also proposed to estimate model parameters. This paper uses classical and Bayesian models to represent the failure behavior of toxic-gas detectors and flame detectors, model maintainability, and evaluate detection-function availability by varying test and calibration frequency. The objective is to generate significant savings in costly offshore instrumentation labor while keeping the fire-and-gas detection system operating within acceptable safety limits. Established software such as Weibull++ and BlockSim was used to support the models. The paper presents two real case studies in which quantitative reliability supported decision-making to improve maintenance strategies and generated substantial benefits for the company.

An automotive clutch essentially has two functions in a vehicle. The first is to transmit and interrupt engine torque to the transmission. The second is to attenuate transmission vibrations, which is the focus of this work. The torsion springs in clutch discs play an important role in reducing vibrations generated by engine combustion before they reach the vehicle transmission. If these vibrations are not attenuated to acceptable levels, they can cause premature transmission wear. Therefore, it is essential to understand the reliability of these springs during product development, ensuring they perform their function throughout the life of the clutch disc. Durability tests are conducted on dedicated test benches to assess product failure risks. The tests consist of millions of clutch-disc actuations according to specification, recording when spring breakage occurs — usually due to fatigue — and the corresponding accumulated cycle count. Based on failures and suspensions, reliability curves are modeled to support spring validation decisions.

The objective of this work was to demonstrate an increase in the availability of a wind farm over recent years. Reliability Growth Analysis (RGA) was used not only for the wind turbines but also for the Balance of Plant (BOP). A detailed analysis of operational and maintenance data supported a strong defense of the company’s interests, demonstrating compliance with previously established contractual terms. In this sense, the paper illustrates the broad applicability of Reliability Engineering tools, which can be used even outside the classic O&M context.

The work presents the maturity journey and reimplementation of qualitative reliability processes in a pelletizing plant after five years of inactivity due to a dam failure and the loss of reliability professionals, progressing to the acquisition and treatment of data needed to model individual assets in the pelletizing process. The objective was to break the status quo and redefine the optimal hot-shutdown interval through simulation using RAM analysis.

With the growing complexity and diversity of assets in modern organizations, implementing an Asset Management System (APM) becomes essential to ensure operational efficiency and regulatory compliance. However, this process presents several challenges that can affect its success. This presentation shares Vale’s experience in overcoming these challenges during implementation of the system. Topics include integration of legacy systems, securing senior-management support, developing an organizational culture oriented toward asset management, ensuring data quality and establishing appropriate performance metrics. Understanding and mitigating these challenges is crucial to the long-term success of asset management in organizations of different sectors and sizes. Although implementation can represent a significant challenge, once consolidated it creates new opportunities to fully leverage its benefits.

This work will be valuable to SIC professionals because it covers several Reliability Engineering concepts. It includes developing the life curve to first failure for each component using Life Data Analysis (LDA), analyzing the life curve of the cooling system after a period combining partially refurbished components with new components using Reliability Growth Analysis (RGA), and applying system reliability in BlockSim across several “What If” scenarios. The application of these concepts together with Life Cycle Cost (LCC) analysis completes the study, highlighting how quantitative reliability analysis supported more accurate life-cycle decision-making for the plate cooler.