Refining Reliability Analytics Engineer

Marathon Petroleum Corporation•Findlay, OH

4d•Onsite

About The Position

The Refining Reliability Analytics Engineer develops, deploys, and sustains analytical capabilities that improve the reliability, availability, and performance of refinery equipment and process systems. The position operates at the intersection of reliability engineering, refinery process analysis, statistics, machine learning, and industrial data analytics. The engineer combines process historian, condition-monitoring, equipment-performance, maintenance, inspection, and operating-context data to identify abnormal behavior, emerging degradation, and changes in asset or process performance. The role requires hands-on development of analytical models and sufficient engineering understanding to ensure that model inputs, relationships, and outputs are technically credible and operationally meaningful. Core applications include rotating-equipment health and performance analytics for pumps, compressors, turbines, motors, and associated auxiliary systems; process-unit and reactor analytics, including weighted average bed temperature (WABT) monitoring for hydrotreaters; anomaly detection and early identification of equipment degradation, process deviation, instrumentation concerns, and changing operating relationships; and reusable analytical methods that scale across comparable assets, process units, and refinery sites. The Reliability Analytics Engineer works with process engineers, rotating-equipment specialists, reliability engineers, operations personnel, maintenance professionals, data scientists, data engineers, and IT teams to move analytical concepts from exploratory analysis into validated and sustainable Asset Health Monitoring (AHM) capabilities.

Requirements

Bachelor's degree in engineering, computer science, data science, statistics, applied mathematics, operations research, or a related technical discipline.
Significant experience developing advanced analytical solutions in refining, petrochemical, energy, manufacturing, or another asset-intensive industrial environment.
Demonstrated experience applying statistical modeling, machine learning, anomaly detection, predictive analytics, or time-series analysis to industrial operating data.
Demonstrated ability to connect analytical methods with equipment behavior, process relationships, degradation mechanisms, or reliability outcomes.
Advanced proficiency in Python or R and working knowledge of SQL.
Experience combining time-series, operating, equipment, maintenance, or other industrial datasets and validating findings with engineering or operational SMEs.
Ability to communicate analytical findings, uncertainty, limitations, and operational relevance to technical and nontechnical audiences.
Willingness to travel up to 25% to support refinery sites.

Nice To Haves

Experience developing health and performance analytics for pumps, compressors, turbines, motors, or associated auxiliary systems.
Experience working with rotating-equipment vibration, temperature, pressure, flow, speed, load, efficiency, lubrication, seal, and maintenance data.
Experience identifying rotating-equipment anomalies using change-point detection, statistical process control, multivariate analysis, residual modeling, or dynamic baselines.
Experience performing hydrotreater or catalyst-bed reactor analysis using WABT, individual-bed temperatures, thermocouple profiles, temperature rise, process rates, pressure, hydrogen relationships, quench behavior, or catalyst age.
Understanding of reliability principles, failure modes, predictive maintenance, condition-based monitoring, RAM, RCFA, and reliability-centered maintenance.
Experience developing, deploying, monitoring, and sustaining analytical models in production environments.
Experience with industrial historian and analytics platforms such as PI and Seeq and with cloud, API, visualization, or model-operationalization technologies.
Master’s Degree or PhD in Computer Science, Statistics, Mathematics, or Operations Research preferred.
Six (6) years or more of relevant experience required.

Responsibilities

Translate equipment failure modes, process-degradation mechanisms, and refinery operating concerns into observable variables, analytical features, and detection methods.
Develop, validate, deploy, and sustain statistical, machine-learning, and engineering-based models for equipment and process health.
Apply change-point detection, statistical process control, dynamic baselining, multivariate analysis, residual monitoring, classification, regression, and time-series methods.
Combine operating conditions, equipment configuration, process loading, maintenance history, and instrumentation quality to interpret model results.
Design explainable outputs that identify the abnormal condition, contributing variables, operating context, confidence, and reason engineering review is warranted.
Validate outputs with engineering subject-matter experts using known events, operating history, retrospective analysis, and false-positive and false-negative reviews.
Translate AHM data models and predictive algorithms into practical reliability strategies that support enterprise initiatives.
Support development of KPIs and dashboards that measure equipment health, RAM performance, and intervention effectiveness.
Partner with engineers, analysts, and data scientists to validate models and ensure outputs align with engineering and operational realities.
Work collaboratively with refinery teams and site-level engineering/operations organizations.
Develop and deliver training to refine site staff on reliability concepts, AHM tools, and best practices.
Foster a culture of continuous improvement and knowledge-sharing.
Monitor industry trends, regulatory requirements, and emerging technologies in reliability engineering and asset monitoring.
Benchmark against peers and incorporate lessons learned into AHM program enhancements.