Tuesday January 15th,
14:00
CCC conference room
874/1-011
There were no further comments on the minutes of the 37th meeting, which had been written by Stefano. Roger briefly announced the topics of this 38th meeting.
Following up an issue that had arisen at the 36th LHCCWG meeting, Massimo had contacted the persons responsible for the power converters of the spectrometer magnets in ALICE and LHCb, Gilles Le Godec, Hugues Thiesen, and David Nisbet. The results of the tests of these power converters demonstrated that the ALICE and LHCb magnets can be powered at the correctly scaled injection currents (1/16th of their nominal current) without any problem. Therefore, from the power converter point of view there is no constraint on operating these magnets.
Roger
and Massimiliano commented that at the 5th
LPC meeting this very morning, LHCb had made a
strong statement that it will not want to ramp its magnets every day during LHC
routine operation, but that it would prefer to leave them powered at full
strength. During the commissioning in 2008, however, LHCb
will accept magnet ramps from 1/16th of nominal strength, at
injection, to full nominal strength, at 7 TeV.
Massimo remarked that in later years too high a current in the spectrometer
could limit the aperture at injection. LHC
Project Report 1009 on “alternative running scenarios for the LHCb experiment”, by W. Herr and Y. Papaphilippou, had already considered some issues
concerning crossing angles for LHCb. Massimo and Massimiliano suggested that Werner publishes another note
summarizing the constraints and considerations pertinent to even larger
crossing angles. Helmut remarked that a careful study and compromise will be
needed. Massimiliano asked whether at 1/16th of the
nominal strength the current ripple would be much bigger. Inferring from the
specifications and the recent powering tests, probably the ripple will be
similar in absolute terms. Roger added that aperture is likely to be an issue
mainly when we run with a crossing angle. Massimo cautioned that there is
always a crossing angle and orbit excursions when the spectrometer is on, which
can be zeroed at the location of the IP with the help of superimposed external
bumps. Massimiliano asked for the constraints which
determine the acceptable value of the crossing angle. Massimo replied that we
could perhaps accept two times the nominal crossing angle at injection, but not
much more than this due to aperture limitations. He reiterated that it would be
good if Werner could write a note describing the various constraints. Gianluigi
inquired the reason for not wanting to ramp the LHCb
magnets on a daily basis. Massimiliano mentioned the
effect of friction, and a possibly resulting damage of
the insulation as reasons, though he wondered whether this would not be an
issue as well for
Magali presented a short overview of the global access system tests which were conducted under the leadership of Silvia Grau and Pierre Ninin, by a large TS/CSE team and in collaboration with AB/OP, during the previous week, from 7 to 13 January, 2008. This exercise represented a first test of the access system integrity.
On Monday 7 January the system was switched from “installation” (“myfare” system) to “operation” (i.e. access with biometrics control). From Tuesday to Thursday, the LHC was closed and searched. Then, from Thursday to Sunday various tests were performed. Searching a single point took between 4 and 5 hours; only the search of point 1 was completed within 1 h. Magali explained that no real patrol was done, but a simplified “quick search”, since we knew that no real beam would be injected. All emergency-exit handles were sealed (“plombage”). The sealing proved to have a very dissuasive effect which is welcome, and which was therefore left at most handles after the exercise, so that nobody would get accustomed to using the emergency handles. Magali pointed out that some people were still working and that access to the experiments was allowed throughout the entire week (a guardian was placed next to each missing grid). Overall there was no interference, except for one accidental door opening in ATLAS. The LHC search was done much faster than expected, thanks partly to the simplified “quick search” procedure and partly to the dedication of the teams participating. However, the tests could not start earlier then foreseen, since it had been agreed to search the experiments not before Thursday afternoon.
Tests were done both in access mode and in beam mode. In access mode the response to one or more EIS-beam devices (EIS=elements importants de sûreté) becoming unsafe was studied, in beam mode the response to one real AUG (arrêt d'urgence général), to a forced door, as well as to the discordance between two signals. Magali explained that “EIS-beam” devices are for example certain power converters, some TEDs, RF electron stoppers, etc. In the access mode tests, one power converter was switched on by hand (where the system should go to access off), then two (where the system access should go off and the BIW (“beam imminent warning”) sirens should start). The systematic tests were highly successful. In the beam mode tests from inside the LHC a sample of doors were forced open and one AUG was triggered, in point 7.
Reyes asked how the response of the dump was checked. Magali explained that the last signal sent to the dump was detected. A similar approach was adopted for the BIC (the signal was sent to the SPS BIC), just as a check whether the last signals were correct. When a door is opened or an AUG occurs, the last signal sent to the dump becomes “false” and the beam should be dumped. Jorg added that in case of the BIC the signal had been sent to the user interface box.
The general outcome of the tests is that the LASS (LHC Access Safety System) functions as it should. The human interface to the LACS (LHC Access Control System) also works, but it should be further improved in view of its slow response. Some minor non-conformities that were encountered will need to be fixed. A few relays had to be changed; a few problems were solved on line. Also one serious problem was found, namely on one occasion there was a non-detection of a missing key in a pad key distributor.
Ralph S. asked whether the response to a missing key will be checked for all keys in all LHC distributors. Magali responded in the affirmative. Stefano recalled that at the time of the November-2007 cold check-out meeting, Magali had not been happy with the interface. Magali replied that, yes, indeed, this was the reason why she said the system should be further improved.
Responding to a question by Massimiliano, Magali elaborated that the LASS would have given the green light to send the beam despite of the missing key, since it did not detect the latter. The signal is implemented by hardware and it is transmitted via cables.
Some unexpected false alarms, like evacuation sirens, fire alarm, ODH alarm in point 6, and AUG in point 8 were observed and are not yet understood. These false alarms resulted in fire-brigade interventions. In the future we should forewarn the firemen of the tests and instruct them how to enter the LHC without forcing doors. Ralph S. cautioned that we could not exclude the occurrence of a real fire during the time of a test. Magali mentioned that the correct BIW, and the wrong siren, false fire alarm and false ODH alarm had all been implemented by the same contractor, indicating a possible cabling problem. However, the effect could not be reproduced later on.
Some other problems found relate to “EIS-beam” equipments. For example, in one occasion, TEDs could not be moved. The reason is still unclear. At another occasion, LASS saw the TEDTI8 as “out” though it actually was “in”, which might have been caused by a faulty end-position contact. Electrical problems with two power converters were also observed.
Magali remarked that it had not been possible to test the full nominal system, since a few elements were not yet installed or implemented, e.g. some doors were not included, the SSA (safety system atlas) was not ready and the new LASS console not yet installed. Therefore, some future tests must still be foreseen.
Magali listed a few other issues that came up during these tests and which will need to be followed up, such as the responsibilities and procedures for searching the rf zones, the proper response when two EIS-beam devices go unsafe (evacuation siren instead of BIW?), the present treatment of a power converter and its corresponding cell as two independent EIS-beam devices, and the need for OP to have a portable ODH (oxygen deficiency hazard) detector. The latter appears necessary, since the firemen’s ODH detector was not working and they also stated not to have the qualification necessary to enter the tunnel in case of an ODH alarm.
Gianluigi asked what would happen if people were allowed to
enter during an ODH alarm. Thorsten Wengler commented that for ATLAS the ODH alarm triggers the
fire brigade. Magali remarked that the handling was
apparently different for the machine. Jan and Django
both remarked that the ODH should be a level 3 alarm, and therefore it should
be the task of the fire brigade. Magali replied that
according to the information she had been provided with the ODH was an alarm 1,
and not 2 or 3. She will follow up all open safety issues with
Roger
mentioned that in the next 3 months, Paul would like to render the access
system fully operational. For this purpose, systematic tests at every door will
be required. Roger then asked Magali whether she
plans to conduct further similar global tests. Magali
answered that the access people would like to do more tests, but whether these
would be local or global tests is not yet known. When the access system will
have been put into its nominal configuration and further qualified by TS/CSE,
DSO and OP acceptance tests will be executed with the full machine closed.
Preparation of these tests is presently ongoing. When these tests will happen
and for how long will they last is not yet defined and this has to be further
discussed with
Massimiliano wanted to hear more details about the quick search, and how much longer the search would take in real operation He asked whether all the locks had been checked during the test. This was the case, but Magali explained that there were many holes left unchecked and other places would need to be patrolled in the right order.
Thijs asked whether the beam would be dumped in case of a fire alarm. Magali did not know whether a dump was automatically enforced, but clearly if the fire brigade enters, the beam must be dumped. Gianluigi stated that no machine at CERN has the fire alarm connected to the beam abort, the sole exception being CNGS as pointed out by Jorg. In the days of LEP the control room had decided whether a fire alarm was real or not. Magali emphasized that the firemen can force any door. Laurette wanted to know whether they may force their entry even into radiation areas. Jan commented that they have a key to do so.
Roger reviewed the status of the procedures. 3 EDMS documents have already been released, for phases A.1, A.4, and A.11. The documentation for phase A.2 will be submitted for engineering check on Wednesday January 16th by lunch time, and the release is expected a month later (Magali), A.3 is ready for the engineering check (Verena), A5 will be ready for engineering check by the end of the week (Laurette), the check of A.6 can be closed (Walter), A.7 should be ready for the check within a couple of weeks (Magali), A.8 might take two weeks to be finalized (Reyes), A.9 is also closed (Walter), A.10 is ready to start with the engineering check (Reyes), and A.12 is in work (Delphine). Massimo commented that some modification may be needed for the latter, in view of recent news on LHCb magnet ramps.
Roger stressed that we now need to use these documents. Stefano remarked that several open issues had been left in each document, adding the question how we would address these. Roger commented that this was an excellent question. There are different types of open issues; some need to be sorted out, others not.
Roger now addressed our getting ready for beam. The power tests of the magnet circuits are ongoing. Power testing of the first sectors will be finished in May; for three sectors the testing will extend into July, based on latest information from Katy. A 4-step procedure exists for bringing the other accelerator systems into operation. The entire machine will be ready for high-energy operation some time in the summer. There is a hardware commissioning (HWC) hole of 4 weeks in February, after which HWC will restart around 8 March.
We will take over finished sectors for operational testing. To prepare the beam commissioning, we will organize a series of readiness reviews, dry runs, and the machine checkout. Roger now flashed the detailed commissioning phases, and the groups in charge of each phase. He then indicated the structure of the readiness reviews for each phase, primarily referring to the instruments and tools needed in each phase.
He announced that the readiness will be reviewed at an “extended LTC meeting” from 3 to 7 March 2008, for which a first preparation meeting had been held in the previous week by a small group. Five principal topics will be covered by this extended LTC meeting: (1) readiness and beam plans for accelerator systems [organized by Roger], (2) issues for injection into part of the machine [Mike], (3) from HWC to full machine commissioning [Gianluigi], (4) AP and BI systems [Oliver], and (5) control systems [Jorg].
Lastly, Roger listed a few LHCCWG topics to be treated early in 2008. Werner will discuss zero-crossing angle filling schemes on February 12. Massimo will give an update on relaxed tolerances on January 29. The 75 (and 25) ns operation is another outstanding item. He suggested that Werner could publish a note about the details of the filling schemes for these and other modes.
Tuesday January 29th, 14:00
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
Update on tolerances (Massimo)
Exceptions for safe beam flags (Jorg)
AOB
Reported by Frank