LHC Update

During the last two weeks while I have been away the LHC has made a rapid start to its 2012 run. Peak luminosities are already around 4.5/nb/s in CMS although it is hard to be sure how accurate this is since they have not yet had time to do the scans required for proper calibration. This exceeds last years records already. About 0.6/fb has already been delivered and is now increasing rapidly. Stability and efficiency is looking very reasonable at this early stage.

Update 20-Apr-2012: According to CMS the LHC has now delivered over 1/fb this year. The ATLAS figure is a fair bit lower and we wont know who is right until the recent Van de Meer scans have been analysed. The milestone was reached thanks to the Machine Development break being reduced by one day. Paul Collier from beam operations has siad that there is still a little scope to increase the bunch intensity from 130 billion to about 150 billion. This could give as much as 33% more luminosity. With these figures they should have no trouble collecting about 1/fb per week and there are 20 weeks of proton physics left after the technical stop next week.

There is now good hope for some preliminary results at the new higher energy of 8 TeV during conferences scheduled for the next few months, with high expectations for the big ICHEP conference in July.

Update 18-Apr-2012: The LHC operations has now reached the maximum bunch numbers of 1380 per beam given the current parameters. This means that it is now fully up to speed and should run like this for the rest of the 2012 run. Progress towards this point has been very smooth and has taken just three weeks since first collisions.

If operating efficiency stays as good as this for the rest of the year they will have no trouble meeting their stated targets. Luminosity peaked at 5.6/nb/s in CMS. This may increase a little if they try a little more bunch intensity but there is not much room for improvement, so this years run will be very different from last year when the luminosity was continually increasing throughout the year. (I am assuming they wont go for something crazy like improving the squeeze or switching to 25ns spacing)

After one more day of running the first Machine Development and Technical Stop will interrupt data collection for about 8 days. If tomorrow goes well they could reach the first inverse femtobarn before the stop.

19 Responses to LHC Update

  1. Shroon says:

    That’s a startling amount of typos for such little content.

  2. Lubos Motl says:

    5.6/nb/s is pretty cool. It seems to continue now…

    • Philip Gibbs says:

      Yes and it is about what would be expected given the squeeze to 0.6m. Last year they had 3.5/nb/s at 1.0m and 3.5*1.0/0.6 = 5.8/nb/s

  3. JollyJoker says:

    “Luminosity peaked at 5.6/nb/s in CMS. This may increase a little if they try a little more bunch intensity but there is not much room for improvement, so this years run will be very different from last year when the luminosity was continually increasing throughout the year.”

    This means we now know fairly accurately how much data we’ll have for summer conferences and by the end of the year, right?

    • Philip Gibbs says:

      The biggest remaining unknown is the Hubner factor which depends on how efficiently they can run. They have 20 weeks of proton physics and about 6 weeks before the cutuff for ICHEP. If we assume a peak lumi of 7/nb/s (they were about to try a bunch intensity of 1.45E11 before the MD) and a HF of 0.2 they will collect 120/pb per day, so 18/fb in the year and 6/fb for ICHEP which is in line with targets. HF could be better than that, hopefully not worse.

      • carla says:

        Going by last year, it takes 3-5 days to recover from a technical stop which gives more like 5 weeks before the end of the first week in June. And last year they were managing 500/pb/week so let’s say they manage 750/pb/week. Then that gives 3.750/fb for the 5 weeks plus 0.75 they already have giving a grand total of 4.5fb which I think they’d be happy with.

  4. ondra says:

    CMS has reduced their lumi numbers, so no 1/fb yet and peak under 6 /nb/s.

    • Philip Gibbs says:

      Not unexpected. It just means they can celervrate the 1/fb mark twice.

      • ondra says:

        Now there is also change on ATLAS side, both experiments now have almost the same numbers on delivered and recorded lumi. I wonder if we will see same numbers also in stable beams when they get through the recovery after first TS.
        Hopefully they finally fix the BLM and injection issues.

  5. Tony Smith says:

    Phil – and ondra, carla, JollyJoker et al,
    is it a serious matter of concern that according to Andrew Elwell
    “… Comments 08-05-2012 23:19:47
    lost cryo compressor for sector 78 …”
    that according to CMS and ATLAS Luminosity public pages
    the maximum Luminosity/Day in May is less than 100/pb
    (average of days in May looks like less than 50/fb)


    are the difficulties listed in Andrew Elwell’s Comments just more or less start-up glitches that should be expected ?


    • Philip Gibbs says:

      From last year’s experience we know that sometimes the LHC runs like a gazelle for a couple of weeks and then it runs like a lame dog the next. During the last week they have had four cryo failures and various other problems. They are not yet back up to 1380 bunches and are using lower intensities. There are no big issues as far as I know so it should run better soon. There are still about six weeks until the cutoff for ICHEP so no need to panic.

    • ondra says:

      Tony, Philip,
      true is we have seen LHC in more trouble and it always recovered and i hope they fix the issues, but to be honest i am starting to be concerned about luminosity target.
      I think they should stop changing things and put priority to stability of the machine, even if you would run at lower intensity but you would make 100+ pb-1 a day, it would be much better than 2 hours a day with 30 pb-1 :).

    • carla says:

      Yes, these difficulties are entirely expected based upon last years problems, but hopefully should be reduced in occurence from last years lessons. So expect loss of the entire electrical network for a few hours and consequent loss of cryogenic conditions for a few days. Also expect the LHC to run like a champ for a week over the next few weeks and accumulate 1/fb/week.

      It’s no big deal delivering 5/fb for the first week in June, whereas their priority is accumulating around 15/fb this year to rule-out/discover a Higgs. They also want to run at 25ns and carry out sufficient machine development to help with the upgrade over the long shutdown. They have the option of running for two extra months if needed.

      • Tony Smith says:

        Carla said:
        “… priority is accumulating around 15/fb this year to rule-out/discover a Higgs …”.

        Will 15/fb(2012) + 5/fb(2011) be enough to give a really clear picture of the Higgs ?

        Please consider, criticize and correct the following scenario:

        Might it be that a really clear picture of the Higgs will require that the high-resolution low-background Higgs to ZZ to 4l Golden Channel has a statistically useful amount of data in the critical region about 50 GeV wide between 110 GeV and 160 GeV
        which, for bins at least as narrow as 2 GeV, means at least 25 bins.

        If you roughly define “statistically useful” as an average histogram height of 10 events for the 25 bins,
        then a clear picture of the Higgs would require at least 250 events in the Golden Channel from 110 to 160 GeV.

        In the 5/fb of the 2011 run from 110 to 160 GeV in the Golden Channel
        ATLAS saw about 6 events and CMS saw about 15 events
        so if you guess that you might get (6+15)/2 = about 10 events per 5/fb
        to get the required 250 Golden Channel events
        you would need about (250/10)x5 = the order of 100/fb

        Would it not be reasonable to reconcile ourselves to waiting until after the long shutdown for repair/upgrade
        and then hope to get 100/fb fairly quickly
        and then see the Higgs clearly
        instead of through a glass darkly ?


      • carla says:

        @Tony, I’m neither a theoretical nor experimental physicist and merely regurgitating what I’ve learnt from the LHC news page.

      • Philip Gibbs says:

        Nobody ever said that they would get a complete picture of the Higgs this year. What they expect is to get enough evidence for discovery of the Higgs if it is there or clear evidence against it if it is not. If it is there at 125 GeV they should be able to see if there are any big deviations away from standard model cross-sections for some of the channels. That will be followed after the shutdown by more detailed checks that would tell us about more of the decay channels and alao the production modes. Precision measurments could start to tell us if there are any BSM efects in case we dont see them directly.

        I don’t get this idea that the LHC is struggling. It has been performing beautifully and has justified the decision to delay the long shutdown until next year. It looked like a risky decision when they took it because it was not clear then that the LHC would do enough to discover the Higgs this year, but now I think if it had been shutting down people would be questioning why. Full marks to the LHC directorate for getting it right.

      • carla says:

        @phil some people, including myself last year, don’t understand that a complex accelerator like the LHC cannot realistically run 100% of the time. They expect it to run like the family car – turn it on, engine starts, everything works 100% generally.

        And if I recall correctly, Phil, weren’t you the same last year predicting 11/fb for the year at 1nb/s intensity;)

  6. Tony Smith says:

    Ondra said on 9 May:
    “… i am starting to be concerned about luminosity target.
    I think they should stop changing things and put priority to stability of the machine …”.

    Now the latest comment on lhcstatus.elwell.org.uk is
    “… Comments 12-05-2012 02:20:11:
    access for power converter change in S45 has started
    then we precycle and fill for physics …”.

    This makes me think that Mike Harney may have been both prescient and correct in his 7 September 2009 comment on Tommaso Dorigo’s blog entry of 4 September 2009:

    “… It just seems like the LHC is in what I call “oscillation mode” –
    a highly technical term that describes the state of a project when it has lost convergence on becoming a reliable piece of hardware because it is in the process of not fixing root problems but rather doing a patch job which results in releasing a new set of gremlins that are waiting in the wings. …
    I can say without a doubt that LHC is definitely in “oscillation mode” and as somebody previously put it, they really need to get to the root of the problems …”.

    Would it be better to just go ahead now with the long shutdown for repair and upgrade so that a full power/luminosity LHC could start taking data almost a year earlier than it would on the present schedule of long shutdown beginning around early 2013 ?

    If they are determined to keep to the present schedule for 2012
    then would the machine be more nearly somewhat stable
    if they were to restrict peak luminosity to about 4/nb/s
    which is about the level of around 14 April which was the best day for luminosity (about 180 /pb) ?


  7. Tony Smith says:

    There are about 5 weeks before the June LHC Machine shutdown.

    For the 4 days 12,13,14,15 May
    luminosity has been 125, 130, 158, 127/pb/day
    with peak luminosities for those days 4.2, 4.3, 4.9, 5.6 /nb/s
    and an average luminosity of 135/pb/day

    If the LHC were to stay in this stable mode for the remaining 5 weeks until the June shutdown prior to the July conferences
    the LHC should have 1 + 5x7x0.135 = 5.7/fb for the 2012 run prior to July conferences which is comparable to the 5/fb for 2011
    by the July conferences we should be able to compare diphoton observations and see whether structures seeming to emerge in 2011 survive in 2012.

    Note that the 2011 ATLAS and 2011 CMS diphoton plots did not fit together very well.
    The red line at tony5m17h.net/CMSLHCdigamma2peaks.png
    connects a high value for ATLAS with a low value for CMS,
    a discrepancy described by Matt Strassler in his blog on 9 Feb 2012:
    “… ATLAS’s largest excess (126-127) is at a point where CMS has a deficit,
    and … ATLAS has a deficit across some of CMS’s excess …
    this discrepancy … had better go away …” with addition of 2012 data if ATLAS and CMS diphoton results are to be reconciled.

    If the diphoton combination of 2011 and 2012 data is as inconclusive as the attempt to combine ATLAS and CMS data for 2011
    the resolution of the situation will fall on the high-resolution low-background Golden Channel of Higgs to ZZ to 4l.
    Although the additional 2012 5.7/fb may be enough for the Golden Channel to give a clear rough view of some structures above 180 GeV such as ATLAS around 240 GeV,
    the number of events in the 110 to 150 GeV range may still be too small to shed much light on structures there.
    Even another 10/fb in the second half of 2012 may not yield enough events for the Golden Channel to clarify structures in the 110 to 150 GeV range.


%d bloggers like this: