The downtime has been monitored from October 2013 to March 2022. We do not monitor downtime by category. In any case, technical downtime statistics wouldn't be very telling. A non-working instrument doesn't lead to much downtime, since we switch to service mode with the remaining ones.
Lockdowns in 2020-2021 have been excluded from the statistics but all other causes are considered: weather, technical problems, human errors, instrument changes, etc. Our experience is that time losses are dominated, by far, by weather issues (clouds and humidity). Quite expectedly they do not feature clear long-term trends, but there is a marked seasonal variation.
Time loss: 23 ± 3% in Summer, 36 ± 5% in Winter. Yearly mean 29 ± 2%, stable during the period 2014-2022.
GROND programmes feature losses of over one-third during MPIA time, which is due to the large fraction of time-critical observations (transits) with this instrument. WFI and FEROS programmes lose typically 20% but many programmes also receive more than their share, typically when there is additional time available towards the end of a period
There seems to be an improvement in the completion of programmes, probably because of a more cautious scheduling.
The number of open tickets evidence three epochs: ESO service mode operations (1997-2008), ESO visitor mode operations (2009-2013) and MPIA operations (2013-2021). The number of tickets that are unresolved within the same calendar year has increased significantly in 2020 and 2021, but it will be difficult to determine if it arises from the lockdowns or an increased instability of the system.
The number of new tickets hase been stable (44 ± 6 per year) since 2016.
