Modernizing automation in industrial control/cyber physical systems through the system engineering lifecycle

2021 Fall.Includes bibliographical references.The systems engineering process seeks to develop systems beginning from a need and ending with an operational system. The systems engineering framework is acknowledged as an effective tool for building complex systems, but this research seeks an expansion in scope and emphasis to include more detailed methods for managing, operating, and upgrading existing subsystems when they are challenged by obsolescence, functional degradation, and upgrades/commissioning. System development from a blank slate is often the default for the systems engineering field, but often an individual subsystem (in this case studied here, the automation system) must undergo upgrades much sooner than the rest of the system because it can no longer meet its functional requirements due to obsolescence. Partial system upgrades can be difficult to conceive and execute for a complex industrial system, but the fundamentals of the system engineering process can be adapted to meet the requirements for maintenance of an industrial control/cyber physical system in practice. Cyber physical systems are defined as systems that are enabled by interactions between computers and physical systems. Computers and other automation components that control the physical processes are considered part of this system. This dissertation seeks to engineer industrial automation systems to enable identification of obsolescence in cyber physical systems, simulation testing of the automation subsystems before/during upgrade, and integrity testing of alarms and automation after completion. By integrating some key aspects of the systems engineering approach into operations and maintenance activities for large-scale industrial cyber physical systems, this research develops and applies 1) novel risk-based approaches for managing obsolescence, 2) novel techniques for simulation of automation controls for fast commissioning in the field, and 3) an automatic alarm configuration engineering and management tool. These systems engineering developments are applied over the course of 5 years of continuous operation and 14 large upgrades to automation systems in the process industry (gas processing, chemical, power generation). The results of this application illustrate consistent improvement in the management, upgrading, and engineering of industrial automation systems. Metrics of system performance used to quantify the value of the proposed methodological innovations include commonly used metrics such as number of alarms, cost, and schedule improvement. For the research contribution which develops novel obsolescence identification and replacement strategies, the results show that using a modified risk management approach for automation and cyber physical systems that can quickly identify components that required upgrade. The results indicate a reduction of roughly 70% of reactive replacements due to obsolescence after the major upgrade and a 24% reduction in unplanned downtime due to part failure during normal operations. For the research contribution illustrating that automation system simulation can confirm that the upgraded subsystems meet functional requirements during upgrade on continuously running sites, results are similarly positive. A new metric is developed to normalize the cost of simulations per system which measures the amount of simulation inputs (I/O) divided by cost. Results show that using the proposed simulation tools can reduce the cost of simulation by 40% on a normalized basis and reduce alarms for a system by 55% during system startup and early operations. Lastly, an audit system was developed for the automation systems to ensure that the subsystem continued to meet functional requirements after the upgrade. Deploying the audit system for alarm configuration was successful in that it resulted in no unauthorized alarm changes after the subsystem upgrade. It also resulted in improved alarm performance at sites since causes of alarm deterioration were eliminated. Results show that these added controls resulted in 52% fewer alarms (post implementation) and the elimination of alarm flooding (periods where more than 10 alarms occur in under 10 minutes). The goal of this dissertation is to document innovative means to develop systems engineering towards operational and maintenance upgrades for industrial automation systems and to provide examples of ways this process can be applied. The values of the proposed engineering methods were validated through its application to over a dozen industrial sites of varying processes and complexity. While this research focused on heavy process industries, the process for identifying obsolete components and making major subsystems upgrades can also be applied to a broad set of industries and systems and provide research contributions to both the fields of industrial automation and system engineering