RAM STUDY METHODOLOGY PART 1 of 3
The use of RAM (Reliability, Availability, and Maintainability) studies in the oil and gas industry has a long history and has become a standard tool to quantitatively predict a production and/or processing facility's performance. Such studies are generally performed in the early design phase through to the detailed design phase and into the operational phase to verify that a facility configuration and design meets the required availability, providing operators and their partners confidence that contracts or production outputs can be achieved.
WHAT IS RAM?
RAM generally uses Monte Carlo simulation to predict availability, by modelling the production critical equipment in its series or parallel configuration design and applying equipment failure rates and repair times. Each Monte Carlo lifecycle simulation randomly assigned failures for all modelled equipment according to their specified failure rates. Such simulation represents one possible scenario of the system's availability. Only the simulation of a sufficiently high number of lifecycles ensures that the average availability is based on a statistically valuable pool of possibilities.
Levels of Detail in RAM Modelling
Depending on a project phase and the available design information, the level of details of a RAM model varies and so does its predicted availability. The following types of availability are usually predicted by a RAM study:
Inherent Availability predicts the availability of a production facility for which only the equipment failure rates, repair times and physical configuration are modelled. Therefore, it represents a constant availability that only considers the corrective maintenance downtime of a system throughout the system's lifetime.
Operational Availability is a measure of the average availability over a time period, and it includes all experienced sources of unplanned downtime, such as administrative downtime, logistic downtime, mobilisation downtime etc.
Production Availability as defined by ISO 20815 standard (ISO 2008), is “the ratio of (actual) production to planned production, or any other reference level over a specified period of time”. Production availability is, therefore, a measure used to quantify a system’s ability to deliver a quantity of product and is presented as a percentage of this quantity.
Another key aspect of the RAM study is its ability to predict the percentage of loss of its modelled components. The tool is used to identify which components are the highest loss contributors or which subsystems may cause a bottleneck in the production and how much availability can be gained by applying a higher redundancy. The predicted system losses and the system availability shall sum to 100%.
The reporting of a RAM study includes the system lifecycle average availability, the list of the top loss contributors and a breakdown of subsystem loss contribution.
Advantages of Monte Carlo Simulation for Probabilistic Analysis
The most important advantage of Monte Carlo simulation lies with the fact that it creates large data pools (i.e. number of simulations) that can be used for probabilistic analysis, able to provide a level of nearly 100% confidence. Hence, simulated lifecycles are generally analysed for a lifecycle distribution that is used to calculate the probability for a certain availability target or to predict the minimum availability for a specific probability (i.e. confidence).
Technical Manager
Bureau Veritas Australia
"The strength of using Monte Carlo simulation for availability studies lies with its probabilistic analysis potential."
