CRICOS No J a university for the world real R Aging Equipment Replacement in Power Systems, Using Reliability and Cost Considerations By Dr. Ghavam Nourbakhsh

CRICOS No J a university for the world real R Reliability in Power Systems load (magnitudes, number of customers) network topology reliability data economic appraisal Reliability improvement versus investment cost

CRICOS No J a university for the world real R Introduction: Reliability – without Aging Continuity of supply Random Failure

CRICOS No J a university for the world real R Introduction: Reliability – with Aging Continuity of supply End of life aging failure

CRICOS No J a university for the world real R Introduction: Objectives Model include both useful life & aging failure, Provides failure frequency & duration indices, Components failure contribution on indices, Measure the effect of aging failures on bulk loads and customer loads, Risk & economic replacement decision.

CRICOS No J a university for the world real R Markov State Space Model, Frequency and Duration (no aging) Markov model for a component enumeration of selected system states; single contingencies: switching repair double contingencies: switching & switching repair & repair switching & repair maintenance & switching maintenance & repair

CRICOS No J a university for the world real R Markov State Space Model, Frequency and Duration General Procedure; 1.Based on the events in each system state, corresponding protection or isolation switching actions are performed, 2.Power Flow algorithm is performed to record any load on outage, 3.Load point and system indices are calculated; Load point outage frequency, average duration, probability, unavailability, energy on outage, average load on outage etc. Similar system indices for the whole network performance The main indices: Expected Frequency of Load Curtailment (EFLC) Probability of Load Curtailment (PLC) Average Duration of Load Curtailment (ADLC) Energy Not Supplied (ENS)

CRICOS No J a university for the world real R Components Failure Contribution to Indices Percentage contribution of components to reliability indices; percentage contribution to PLC: percentage contribution to EFLC: percentage contribution to PLC & EFLC weighted by load curtailment:

CRICOS No J a university for the world real R Application to Small Sub-transmission System Overall system indices Load point indices Percentage contribution to probability Small sub-transmission system

CRICOS No J a university for the world real R Application to Medium Substation System Medium substation system, with loads 1, 2 and 3 are 10, 50 and 100 MW Overall system indices Percentage contribution to probability Load point 1 indices

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Typical electric network equipment with end of life aging failure distribution and corresponding failure rate characteristic.

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Markov model for a component, including aging state The model is used for system states probabilities and frequency calculations, 3 rd order contingencies are used for accuracy, The model includes most important outage causing states, Components’ aging contribution to indices can provide valuable information on weak links.

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration State probabilities of the Markov model, including the aging state

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Brief steps to calculate system reliability using this method; 1. Using normal or Weiball, find aging failure rates, 2. Find all component state probabilities and frequencies, 3. Using 3 rd order enumeration, find system state probabilities and frequencies, 4. Reconfigure the network for every system state, check for any outage loads, 5. Increment the planning year and follow 3 & 4 again.

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Effect of Components Aging on Overall System Load Unavailability Planning YearsTX1TX1,TX2CBTX1,CBTXs,CB Small sub-transmission system

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Contribution of system components to load(1) PLC index

CRICOS No J a university for the world real R Application of Aging Markov Model, using Frequency and Duration Medium substation system, overall system results, as all transformers and breakers are aging.

CRICOS No J a university for the world real R Effect of Aging in Sub-transmission Network on Distribution Load Customers Distribution system reliability, customer related indices: Load point indices; failure rate annual unavailability average outage duration energy not supplied System Indices; SAIFI (system ave. interrp freq) SAIDI (system ave. interrp dur) CAIDI (cust. Ave. interrp dur) Aging failure frequency and average outage time at load buses are applied to distribution feeders and the effects are measured at load points. The indices at customer load points can be used as part of risk criteria for upgrading the facilities in sub-transmission system.

CRICOS No J a university for the world real R Effect of Aging in Sub-transmission Network on Distribution Load Customers SAIFI (int/cust-yr) SAIDI (h/cust-yr) CAIDI (h/cust-int) ASAI ENS (MWh/yr) Feeder Feeder Feeder Feeder System RBTS Bus 2, stand alone distribution system indices.

CRICOS No J a university for the world real R Effect of Aging in Sub-transmission Network on Distribution Load Customers YearSAIFI intrp/cust yrSAIDI hrs/cust yrCAIDI hrs/cust intrpASAIENS MWh/yr RBTS Bus 2 distribution system – supplied from load Bus 1 of the aging sub-transmission network Base Case SAIFI (int/cust-yr) SAIDI (h/cust-yr) CAIDI (h/cust-int) ASAI ENS (MWh/yr) System Indices

CRICOS No J a university for the world real R Aging Replacement Decision Aging replacement decision are made based on two criteria; economic, and risk The economic procedures used are as follows; (1)Consider the aging failure only for one equipment at a time. (2) Evaluate the damage costs for two cases with and without the aging failure of the selected equipment, the difference in the damage cost between the two cases is the damage cost caused by the equipment's aging failure. (3) Repeat Step (2) for every planning year starting from the current year until the year in which the damage cost caused by this equipment is equal to or larger than its investment cost plus maintenance cost. The equipment should be replaced in this year. (4) Repeat the above process for other equipment one by one. Obtain a list of replacement times for all equipment in the system or substation.

CRICOS No J a university for the world real R Aging Replacement Decision

CRICOS No J a university for the world real R Aging Replacement Decision Total expected interruption cost (EIC) is: Cash flow for annual capital investment is calculated using the capital return factor (CRF), as follows; A = S x CRF, where A = annual equivalent investment S = actual capital investment in some year i = discount rate n = economic life of investment (years)

CRICOS No J a university for the world real R Aging Replacement Decision i=0.08 discount rate n=45 economic life of investment S=3.80E+06 actual capital investment A= annual equipment investment TX1 & TX2 aging, with 10% annual load growth.Small sub-transmission, with 10MW for each load.

CRICOS No J a university for the world real R Aging Replacement Decision

CRICOS No J a university for the world real R Aging Replacement Decision i=0.08 discount rate n=45 economic life of investment S=1.80E+06 actual capital investment A= annual equipment investment Breaker and a half substation. C.B.s replacement target year, with 10% load growth.

CRICOS No J a university for the world real R Questions ?