Down Time

The down time statistics are based on individual fault reports. In Table 1 I give the general down time statistics for period 32.

A total of 83 fault reports were submitted, with an average time lost of 9 min per fault, for a total down time of 0.7% (0.6% on scheduled observing nights). Of these, 60 reported no time lost, 21 reported $<$ 2 hrs lost, and 2 reported 2 or more hrs lost.

Table 1: Technical down time statistic period 32: 2005-10-01 to 2006-04-01

Night included	Time lost	Nights	Percentage$^a$	Last period
All nights	755 min	182	0.7%	2.8%
Scheduled observing nights$^b$	435 min	116.5	0.6%	2.3%
Technical nights	185 min	26.5	1.2%	0.7%
Service nights$^c$	30 min	17.5	0.3%	1.4%
Visitor instruments	135 min	39	0.6%	6.8%
$^a$ Assuming an average of 10 hr per night
$^b$ Excluding technical nights and visitor instruments
$^c$ Excluding service nights with SOFIN

This compares to a down time of 2.8% over all nights (2.3% on scheduled observing nights) in period 31. For that period 101 faults reports were submitted with an average time lost of 24 min per fault. Of these, 58 reported no time lost, 38 reported $<$ 2 hr lost, and 5 reported 2 or more hrs lost.

The reported down time is at an all time low. One might argue that this is because the reporting period covers the winter period, but the past winter was not particularly bad. For that matter, the typical number of hours that the telescope operates in Summer and Winter is actually not very different. However, given that there are typically only a few big problems causing most of the down time in each observing period there is likely some effect from small number statistics which can appear to make things as for down time look relatively positive or negative. If you look at the total down time over all nights in a year combining consecutive observing periods (i.e., looking from the 1st of April to the 1st of April) one gets: [P25/P26] 2.2%; [P27/P28] 2.5%; [P29/P30] 1.3%, and; [P31/P32] 1.6%.