Engineered Systems Applications

Reliability & Throughput

System Reliability and Process Throughput Analysis

Reliability & Throughput

For many engineered systems, it is necessary to predict measures such as the system's reliability (the probability that a component or system will perform its required function over a specified time period) and availability (the probability that a component or system is performing its required function at any given time). For some engineered systems (e.g., processing plants, transportation systems), these measures directly impact the system's throughput: the rate at which material (e.g., rocks, chemicals, products) move through the system.

Reliability models are frequently used to compare design alternatives on the basis of metrics such as warranty and maintenance costs. Throughput models are typically used to compare design alternatives in order to optimize throughput and/or minimize processing costs.

By combining the flexibility of a general-purpose and highly-graphical probabilistic simulation framework that can directly model the movement of material through a system, with specialized features to support reliability analysis, GoldSim allows you to create quantitative and transparent reliability and throughput models to allow you to ask "what if" questions regarding various designs and make defensible risk management decisions. 


These analyses are facilitated by the GoldSim Reliability Module. The fundamental outputs produced by the Reliability Module consist of predicted reliability metrics (e.g., reliability and availability) for the overall system, and for individual components within that system. The Reliability Module tracks and analyzes failure scenarios, which allows for key sources of unreliability and risk to be identified.



White Papers

Technical Papers

  • Proving a New Refinery Design Using Reliability Proving a New Refinery Design Using Reliability Throughput Modelling

    2011 International Applied Reliability Symposium, North America: San Diego, California – June 7 - 9, 2011

    Damien Willans

    The design, construction and operation of large-scale resource projects are now subject to extreme levels of competition and cost, with many projects running into billions of dollars. For this reason, resource companies, bankers and every participant in the design train are now compelled to understand every aspect of risk to be undertaken, usually well before the project becomes a reality. A key element of the design process is knowing if the plant can actually produce the required output – along with a clear understanding where the areas of potential risk might be. Rather than ignoring the reality of failure effects or, more commonly, de-rating the entire design by some “accepted” operating availability number and trying to justify it, companies are now turning to plant reliability modeling up-front to prove more efficient designs, even with the uncertainties of failures well before plant startup. The concept of calculating “throughput” provides an absolute plant (design) model output for given inputs, by directly linking plant flows and consequences to the equipment availability. This presentation describes a project recently undertaken, including the project background, modeling stages and processes, and the successful outcomes provided for one client who is currently undertaking a very large-scale resource investment.

    This presentation describes how GoldSim and the Reliability Module were used to simulate the throughput of an alumina refinery, accounting for failures and preventive maintenance schemes, to support facility design.


  • A Reliability and Availability Sensitivity Study of a Large Photovoltaic System

    25th European Photovoltaic Solar Energy Conference, Valencia, Spain – September 2010

    E. Collins, S. Miller, M. Mundt, J. Stein, R. Sorensen, J. Granata, and M. Quintana, Sandia National Laboratories

    A reliability and availability model has been developed for a portion of the 4.6 megawatt (MWdc) photovoltaic system operated by Tucson Electric Power (TEP) at Springerville, Arizona using a commercially available software tool, GoldSim™. This reliability model has been populated with life distributions and repair distributions derived from data accumulated during five years of operation of this system. This reliability and availability model was incorporated into another model that simulated daily and seasonal solar irradiance and photovoltaic module performance. The resulting combined model allows prediction of kilowatt hour (kWh) energy output of the system based on availability of components of the system, solar irradiance, and module and inverter performance. This model was then used to study the sensitivity of energy output as a function of photovoltaic (PV) module degradation at different rates and the effect of location (solar irradiance). Plots of cumulative energy output versus time for a 30 year period are provided for each of these cases.


  • The Comparison of Three Photovoltaic System Designs Using the Photovoltaic Reliability and Performance Model (PV-RPM)

    Sandia Report SAND2012-10342, Sandia National Laboratories, Albuquerque, New Mexico – December 2012

    Steven P. Miller, Jennifer E. Granata and Joshua S. Stein Sandia National Laboratories

    Most photovoltaic (PV) performance models currently available are designed to use irradiance and weather data and predict PV system output using a module or array performance model and an inverter model. While these models can give accurate results, they do so for an idealized system. That is, a system that does not experience component failures or outages. We have developed the Photovoltaic Reliability and Performance Model (PV-RPM) to more accurately model these PV systems by including a reliability component that simulates failures and repairs of the components of the system, as well as allow for the disruption of the system by external events such as lightning or grid disturbances. In addition, a financial component has also been included to help assess the profitability of a PV system. This report provides some example analyses of three different PV system designs using the PV-RPM.


  • Accelerated Testing of Metal Foil Tape Joints and Their Effect on Photovoltaic Module Reliability

    25th European Photovoltaic Solar Energy Conference, Valencia, Spain – September 2010

    N. Robert Sorensen, Michael A. Quintana, Michael J. Mundt, Edward V. Thomas, Steven P. Miller, and Samuel J. Lucero, Sandia National Laboratories

    A program is underway at Sandia National Laboratories to predict long-term reliability of photovoltaic (PV) systems. The vehicle for the reliability predictions is a system performance model, currently being run under a simulation software called GoldSim™. The model includes inputs for module performance, irradiance, and degradation. In order to be truly predictive, physics-informed degradation processes and failure mechanisms need to be included in the model. This paper describes accelerated life testing of metal foil tapes used in thin-film PV modules, and how tape joint degradation, a possible failure mode, can be incorporated into the model.


Learn More