A COMMON OPERATION METRICS FOR STORAGE RING LIGHT SOURCES
Storage ring light sources aim for a high operation reliability. Very often the beam availability is used as the operation metrics to measure the reliability of the accelerator. A survey of several light sources revealed that the calculation of this statistics varies significantly between the facilities. This prevents a useful comparison of their reliabilities. The authors propose a specific metrics for the reliability of storage ring light sources; a metrics that will allow a detailed and meaningful comparison of these particle accelerators.
The calculation of the operation metrics of an accelerator serves several purposes: it quantifies the improvement of a specific facility over time, or it is used to compare the performance of similar facilities. In the first case one should select an operation metrics that is close to the requirements of the users of the facility (Lüdeke 2014). For the latter case the operation metrics of the compared facilities should follow a common standard. In particular the beam availability is often used to compare light sources. While these statistics are published for most light sources, very few facilities do add exact definitions on how these numbers are calculated. A survey of several light sources revealed (Lüdeke 2009) that the calculation of these metrics do vary considerably. The conditions under which beam is considered available are often defined in common sense terms and even if there are formal definitions, these do differ between the facilities. But if beam availability is not identically defined for the facilities, then these metrics are not useful for a meaningful comparison. Our aim is to propose a simple, well-defined, formal operation metrics for storage ring light sources, to make the reliability of these facilities comparable. The metrics itself clarifies for each facility the beam parameter promised to the users, and the statistical data do reveal, how well these promises are kept.
The definition of beam availability is important in order to judge the validity of a comparison of the numbers from different facilities. A survey on failure analysis in 2008 of nine light sources1 revealed significant differences for the calculation of beam availability (Lüdeke 2009). In the following we’ll summarise the main findings of the survey. In many cases the beam availability rules were determined by common sense: any event that prevented the majority of the users to measure was considered to be downtime. Some facilities considered “long” injector outages - causing decaying beam operation - to be downtime, others accounted for these events individually. Most facilities were only counting beam delivery between two outages if it exceeded a minimum duration. But the minimal required duration varied between 15 and 60 minutes between facilities. In the case of long beam outages most facilities organised compensation time for the users, to allow them to finish their experiments. The compensation time was accounted for in different ways, depending on the facility: some fully subtracted the extra beam time from the downtime, others ignored this extra time for the availability calculation. All light sources did record other events than beam outages, like increased beam sizes or orbit problems, but no one did publish statistics of these other failure modes in regular intervals. During a discussion round at the ARW 2013 in Melbourne (Web page of the Accel...) we polled the calculation of beam availability from participants of ten different light sources2 with the same result as the survey of 2008. The authors then concluded that a common operation metrics is needed, that allows a standardized calculation and consequently a meaningful comparison of the reliability of different storage ring light sources.
Survey participants for the failure analysis questionnaire came from: APS, ESRF, SPring-8, Diamond, SOLEIL, BESSY II, Elettra, ANKA and SLS.↩
Participants of the ARW’13 presenting their calculation of beam availability came from: ALBA, Australian Synchrotron, BESSY II, Diamond, SPEAR, NSRRC, SOLEIL, Elettra, SLS and PETRA III.↩
We compared the current situation at our facilities and discussed the advantages and disadvantages of the different definitions of downtime and beam availability. Based on that we developed a proposal for a common operation metrics. The comparison soon revealed, that currently many failure modes are summarised under the terms “downtime” and “beam availability”. We propose to disentangle those failure modes and record statistical data for each mode individually. It turned out that the definition for each failure mode has to be adaptable for the different light sources, because nominal beam parameters do differ between facilities and variations of a specific parameter may have severe conse