Reliability and availability considerations for CLIC modulators Daniel Siemaszko OUTLINE : Give a specification on the availability of the powering system of the drive beam LINAC klystrons. Evaluate the reliability of a given topology/solution. Evaluate the reliability of modular/redundant systems. TE EPC

Hypothesis (Decelerator example) N+1 redundancy allows one failure of the modules in a power converter. The whole converter fails when two module failures occur. A factor  describes the converter failures that are saved by redundancy. The trimmers allow a tolerance up to twenty failures [ref: Adli]. Estimated repair time includes machine cool down and walking time in the tunnel (4h). 11/05/2010Daniel Siemaszko2 TE EPC

Composite MTBF model Failure rates  =MTBF -1 combined with the same association rules as impedances. Reliability calculated as a function of failure rate and mean time between preventive maintenance (or technical stops or horizon h). Serial reliabilities are multiplied. 11/05/2010Daniel Siemaszko3 TE EPC

Markov chains Each converter defined as a set of states with probability transitions after each time step R. Failure probabilities (F=1-R) defined as a function of failure rate and time step. Matrix P m contains all transition probabilities. The failure probability of the whole system is a combination of all components probabilities. 11/05/2010Daniel Siemaszko4 TE EPC

Poisson modelling Failure rates are assumed to be Independent Identically distributed exponential variables. The expected number of failures is given as a function of horizon time (namely days between preventive maintenance) with an envelope corresponding to 95% probability. Down time is a function of number of failure and MTTR (Mean time to repair) and considered the down time of the maintenance days. 11/05/2010Daniel Siemaszko5 TE EPC

Modularity / Redundancy (1) Modularity and redundancy is a way for increasing the reliability of a power converter. However, modularity decreases the overall MTBF of a system by increasing the number of components. Redundancy increases the reliability if and only if failures can be saved and that the added redundant system does not add additional failures. Short circuit is ensured with a dedicated crowbar that must not be fired under normal operation. Open circuit is ensured with a dedicated breaker that must not open under operation. When a converter fails, depending on the redundant structure, the short-circuit or open circuit must be ensured. Factor κ stands for the probability of saving a failure by redundancy. Its value is crucial when predicting the global reliability of a system. 11/05/2010Daniel Siemaszko6 TE EPC

Modularity / Redundancy (2) If parameter κ= 60% (left-hand side graphic), then modularity adds more failures to the system than an individual converter. Only one case (one converter and one redundant module) can increase the system reliability. If parameter κ =80% (right-hand side graphic), then modularity can help increasing reliability but the price to pay stays high for small increase. For higher values of κ, then modularity and redundancy increase the reliability of system with a decisive value. 11/05/2010Daniel Siemaszko7 TE EPC

Drive Beam Linac If the klystron modulators are designed with a MTBF of 50,000 hours, the powering of the 1,638 units will reach some 93,5% availability, counting on individual powering. With a solution, including one hot spare for 20 modulators (and some few minutes for remotely swapping a failed converter), then 97% is reached for a horizon time of 100 days. The design of the modulator is still under research. It will include reliability optimisation with either redundancy or hot swap. Modulator MTBF target is: 100,000 hours. 11/05/2010Daniel Siemaszko8 TE EPC

CLIC machine availability (1) The minimum expected machine availability with individual powering of all magnets does not reach 80% when considering failure tolerance in the drive beam decelerator (which comes for free.) The availability due to maintenance is defined as one maintenance day out of h days which is the horizon in reliability calculations. Availability is the proportion between MTBF and MTTR 11/05/2010Daniel Siemaszko9 TE EPC

CLIC machine availability (2) When considering failure tolerance in the main beam quadrupoles, hot swap in the klystron modulators, and hot swap redundancy for each magnet, then a peak in availability of 93% is reached for a horizon of some 60 days. 11/05/2010Daniel Siemaszko10 TE EPC

Towards specifications The modulators have same requirements on voltage cycles than in traction but for a shorter length of time. Given the specification of 100,000 hours on one modulator’s MTBF, what are the possibilities? Single Individual modulator: high reliability required on every single component. Modular-Redundant approach on single modulators: Need for high κ factor. Hot spare approach: Need for ready hot spares for quick replacement of bulky modulators. 11/05/2010Daniel Siemaszko11 TE EPC CLIC life time10 years Availability97% Thermal cycles1500? Modulator stops every two days? Voltage cycles1.3·10 10 (50Hz operation)

Modulator reliability The reliability of one single modulator or modulator’s module is a function of the reliability of: Charger – Pulse transformer – Solid state switch – Bouncer – Capacitors – Control – Measurement. IGBT reliability depends on a thermal factor, environmental factor, quality factor plus voltage stress factor. All together define a number of cycles to failure. When evaluating the reliability of a topology, reliability or failure rates in FIT of each component should be known. So a global reliability can be drawn on the modulator. However, they will be based on assumptions, namely when talking about thermal effects and environment. When going for a modular – redundant approach, the reliability of the bypass system must be high enough to ensure high availability. 11/05/2010Daniel Siemaszko12 TE EPC