RAM (Reliability Availability Maintainability)

This video is one of a 10 part series. Visit our channel's playlist to watch the full set. If you would like more information about Opgrade or Nate Barber visit www.opgrade.com. Narrator: Nate Barber Video slides: RAM How to Size Equipment to Maximize ROI -What is a Monte Carlo simulation? *Usually uses random numbers as opposed to deterministic algorithms *Uses a repetition of “trials” or “missions” to predict overall production levels (On-Stream Factors) *Mission length often the life of a facility *Each mission: -Calculate production level, move forward in time -Equipment fail “randomly” according to their PDF -Repeat *Many missions needed for statistical significance -Why perform a RAM study? *NOT to improve reliability! -Reliability is an input to the study, not an output -Real output is On-Stream Factor (OSF) *Show the value of improved reliability or capacity -Which equipment configuration yields the highest ROI? -Where are the equipment availability bottlenecks? -Where are the maintenance activity bottlenecks? *A bank financing a project requires it -Ability to pay loan depends on ability to make products -Why perform a RAM study? *NOT to improve reliability! -Reliability is an input to the study, not an output -Real output is On-Stream Factor (OSF) * Show the value of improved reliability or capacity -Which equipment configuration yields the highest ROI? -Where are the equipment availability bottlenecks? -Where are the maintenance activity bottlenecks? *A bank financing a project requires it -Ability to pay loan depends on ability to make products -How to perform a RAM study *Collect and document assumptions in a Study Basis -Collaborative document to gain team agreement *Define failure logic in a RBD (ABD) -The effect to the system if this thing goes offline *Start developing stochastic model -Keep it as simple as possible until simple isn’t enough -System complexity may require Monte Carlo simulation *Start running model according to Study Basis -Does the model need to change to better reflect reality? -Do assumptions need to change to better reflect reality? *Collect and document results in a Final Report -Available Software *ReliaSoft -BlockSim *WS Atkins plc -RAMP *DNV GL -Maros -Taro *Opgrade tools -Reliable Process Manager (RPM) -Microsoft Excel custom functions -How to perform a RAM study *Items in parallel: Generalized Binomial Function p = probability equipment is online = n = number of equipment in parallel k = number of equipment offline *Items in series: simple product *OSF for one item = p *On-Stream Factor of 3 x 50% pumps = *If MTBF is 3yrs & MTTR is 24hrs -2 x 100% pumps, OSF = 99.99992% -3 x 50% pumps, OSF = 99.99988% *Difference is 22 min lost productivity every 100 years PROCESS EVALUATION TOOLS [x] Linear Programming Optimize profitability based on unit selection & configuration [x] RAM Studies (Reliability, Availability, Maintainability) Evaluate equipment bottlenecks based on reliability [x] Dynamic Process Simulations Evaluate pressure, temperature and flow profiles in systems [x] Computational Fluid Dynamics Evaluate pressure, temperature and flow profiles in equipment