In MYRRHA, availability is of critical importance. Numerically it is expressed in terms of the Mean Time Between Failures (MTBF), which in MYRRHA has to reach 750 hours, a failure being defined as a beam trip of more than 3 s. This is about an order of magnitude higher than what is currently obtained in comparable accelerators. Therefore the availability issue is the outstanding challenge for this accelerator, and the focus of its R&D. The “classical” means for optimizing the availability, essentially based on the reliability of individual components, will not suffice, and hence fault tolerance must be added.
Improvement of the availability
In order to improve the availability of the accelerator, some strategic principles are adopted:
Reliability of individual components
Ensure the highest reliability of individual components, e.g. by design, over-rating, prototyping, testing and quality control.
Include redundancy in the global design of the accelerator. It is fundamental to yield the possibility of reallocating the functionality of a failing component. This may be achieved in 2 ways:
By having a functionally compatible component readily installed (this is classical or parallel redundancy).
By distributing the missing function over a number of unaffected components (this is distributed or serial redundancy).
Include repairability in the global design of the accelerator. Whenever it is conceivable, the components themselves, their accessibility and the pertaining logistics are designed in such a way that it is allowed to recover a failing component during the operation of the redundant scheme.
Implement the fault tolerant scheme. The system will become fault tolerant if and only if the failure of a component is transparent for the availability of the proton beam. In practice this means that the total time of fault detection followed by setting up the redundant scheme (whether parallel or serial) must be smaller than 3 seconds (which is the definition of a fault in the MYRRHA context). This needs:
Fast, sensitive and failsafe detection systems.
Novel fast digital electronics, especially for the low level RF control system in charge of implementing the distributed redundancy scheme.
More information: The LINAC for MYRRHA