| Description: | EDITOR'S NOTE: This DQR refers to data collected by the EBBR at E13 prior
to the begin date of regular ARM data. At that time, the data streams which
contain the EBBR.E13 data were named
Dsgp15ebbr1.a0
Dsgp30ebbr1.a1
Dsgp5ebbr1.a0
These data are not readily available from the ARM archive, but can be
made available by special request. The actual time range of the problem
described here is 921230.0000-930423.1200.
EBBR1, Central Facility reported incorrect home value signals from 30
December 1992 through 23 April 1993. The home signals are used in the
calculation of fluxes by the CR10 data logger. The fluxes for this period
are thus invalid as reported and must be recalculated from the archived 15
minute data. The proper home signals are less than 35 for 30 minute home
signal and greater than 35 for 15 minute home signal. Home signals were
the same value for 15 minute and 30 minute during the period of interest,
although they were not a constant value. The chronology of the situation
is reported below. On 7 April 1993 I noticed the problem in the A0 data.
On 30 December 1992, the automatic exchange mechanism (AEM) of the Central
Facility EBBR froze up with ice and stopped working (surmised from looking
at the EBBR1 data). This caused a fuse to blow and stop the mechanism from
operating. However, one channel of the CR10 data logger was apparently damaged
in the process. The condition of the AEM only, was noticed on 4 January 1993
by site operations personnel. The blown fuse was replaced, which started
the AEM working; but AEM home signal values between 30 Dec and 04 Jan were
around zero. Home values remained around zero until at least January 8, 1993.
I have a gap in data archived at ANL between 8 and 15 January 1993, during
which I do not know quite what was happening. On 11 January 1993 site
operations personnel found that the EBBR1 battery was run down and the EBBR
thus was not working. A new battery was installed. On 15 January the home
values were about 34. I am assuming that they became 34 upon installation
of the new battery on 11 January. Until at least 15 January the home signals
stayed around 34. After that period the home signals became greater than
35 and generally were around 36. Only occasionally (a few hours here and
there)were the home signals just above or below 35 to produce the proper
flux outputs. On 23 April 1993, at my direction (and after a week of tests
on other parts of the system), the CR10 data logger was replaced. This solved
the problem. In fact, previous to this I had suggested reloading the software
into the previous CR10, but the CR10 would not accept it; as soon as the
machine was deactivated before that procedure it stopped working altogether.
I find it amazing that it worked at all, although all data other than the home
signals, and the calculated quantities dependent on them looked perfectly
acceptable during the period of interest; there is no evidence that the
CR10 was malfunctioning in other areas.
The road to restoring the 30 Dec to 23 April data would be a rocky one, as
each 15 minute data period would have to be examined to determine the
proper position of the sensors on the AEM. However, it is possible to
reconstruct most of the 30 minute calculated values from the 15 minute data. |