LHC Luminosity Estimates for 2011 (and poll)

Last night the Large Hadron Collider clocked up another record luminosity using 624 bunches per beam. In ATLAS the peak luminosity was measured at 737/μb/s while CMS saw about 10% less at 660/μb/s. In theory they should be seeing about the same amount. Paul Collier (head of the beams department at CERN) has said that they suspect the ATLAS figure is too high. A more accurate measurement will be made after the next technical stop in two weeks time when a Van der Meer scan will be run through. Meanwhile I will quote the lower CMS numbers and will be updating the figure in the viXra Log banner above.

An important date coming up in the LHC calendar is the 6th June when the “Physics at LHC 2011” conference starts. The experiment teams will be keen to show results using as much data as possible collected before that date. That will be challenging given the rate at which data will then be accumulating and the need to get about 3000 physicists in each collaboration to sign off on any results before they are presented.

A few weeks ago I said the LHC should be able to deliver 200/pb before the conference. In fact they have already surpassed that figure for 2011 and can add last years 40/pb as well to make nearly 240/pb. I will also be tracking this number above. Two weeks ago I upped the estimate to 500/pb for the conference but even that is looking a little unambitious now. According to the plan there are about 30 days left for physics before the 6th June and they can now easily collect about 30/pb each day. My new estimate is at least 1000/pb which is 1/fb in time for the conference assuming they don’t lose too many days to machine development or faults. Recall that 1/fb is the official target for the whole of 2011! It may sound like their original expectations were a bit low but really this amazing result is due to the exceptional early performance of the collider when compared with previous models.

The 6th June also marks the point when commissioning of the LHC for this year should officially stop. If all goes well they will be aiming to reach this years maximum luminosity by this date. Which will mean filling with 1404 bunches per beam to provide something like 70/pb per day. From 6th June there are 124 days marked on the plan for running proton physics. Do the sums and you will get my new estimate for this years total of 10/fb, easily enough to find the Higgs boson whatever mass it has, or rule it out.

Update: If you dont like my estimate of 10/fb here is your chance to say how much luminosity you think the LHC will deliver this year

30 Responses to LHC Luminosity Estimates for 2011 (and poll)

  1. katzeee says:

    I wonder if they will reconsider their decision not to stop in 2012, if they really get 10 fb^-1?

    • Philip Gibbs says:

      It’s a good question. Unless they change their mind early it may be hard to reverse that decision because a lot of preparations have to be made for the long shutdown and the plans would have to be different.

      It also depends on what they find. There is always going to be a good case for doubling the data if it will turn observations into discoveries.

      The best scenario will be that they find they can increase the energy for 2012. That will certainly make it still worth doing.

  2. karla says:

    Can they easily collect 30/pb each day now? I’m not sure.

    Take a look at the current luminosity graphs:


    For the past 6 days, they had the potential to collect around 25/pb each day whereas it’s more like 15/pb on average.

    What will be the next number of bunches and when?

    What are the important mile stones in data collected – 100/pb? 500/pb, 1/fb, 10/fb?

    • Philip Gibbs says:

      A lot of the delays happen after they change the filling scheme and encounter something unexpected. Once they settle into a fixed scheme it should go more smoothly, of course there will still be the occasional glitch that dumps the beams, but I’m allowing for some of that. There are still a lot of caveats for that number, in particular it might prove difficult to maintain stability at 1404 bunches in which case they will settle for less.

      They are currently going up in 144 bunch steps so presumably it will be 768 next. They need three good fills and 20 hours of stable beams at each step. The coordination vista says 624 bunches until Friday so expect the next step after that.

      • carla says:

        Any idea what a good fill is? If they need 20 hours, then maybe it’s at least 7 hours for each 3 fills, and they’re on the 3rd right now.

      • Philip Gibbs says:

        They are already qualified by the usual standards but seem to be in less of a hurry at this point. They need about one step every four days of operation to reach max luminosity by 6th June.

        The morning meeting slides today end with “One or two more 624 bunch fills. Probably step up to 768 Sunday morning”

      • carla says:

        Any idea what’s involved in the machine development and technical stop phase over the next 9 days?

      • Luboš Motl says:

        In the case that the U.S. hurts Osama bin Laden, Al Qaeda has promised to detonate a nuclear device in Europe. Some Americans said Help yourself. At any rate, the officials in Al Qaeda determined that the most powerful nuclear device was the LHC, so they detonated its UPS – power source – this morning. Surprisingly, the planet didn’t explode. Extra experts were brought at 9 am and they moved the problems from the UPS elsewhere. I am sure you watched it, too. 😉 Otherwise, no idea!

        After he corrected 1/pb to 1/fb, I voted for the minimal-luminosity option in Phil’s poll so I am not surprised. 😉

      • Philip Gibbs says:

        I have to admit to indulging in an evil chuckle every time the “Uninterruptible” Power Supply gets interrupted. I wonder how often this device actually keeps the power on during an external power failure.

        We are now into time for machine development and a technical stop so don’t expect any more data for a couple of weeks anyway.

      • Philip Gibbs says:

        Carla, I only know they were going to do some unsqueezing tests for Totem and Alfa. There will be plenty of other things they have to do but that’s too much detail for me.

  3. JollyJoker says:

    I’m a little surprised at how they can get the luminosity so much higher than expected just a few months ago. I seem to recall that they assumed something like 3-5/fb before the end of 2012.

    ps. You have pb instead of fb in quite a few places towards the end of the post.

    • Philip Gibbs says:

      The only thing that really improved was using 50ns instead of the 75ns spacing they planned. They must have not expected to be able to get the luminosity up so fast, or perhaps they are not expecting a high running efficiency. There is still a lot of uncertainty.

      Wasn’t the 3-5/fb estimate for end of 2011 though?

      • JollyJoker says:

        “How much luminosity will they collect next year? 1/fb is still the official target and 1/fb-3/fb is the official estimate, but some people are thinking optimistically about up to 5/fb .”

        I’m not sure where I got the 3-5 for 2012, but the official estimate for 2011 was only 1-3 as late as January. Not that I’m complaining, by any means 🙂

    • Bill K says:

      The 50 ns, plus their estimates assumed a Hubner factor of 0.2 – 0.3, but we’ve been doing much better than that!

    • Philip Gibbs says:

      I think the half life for luminosity has improved a lot too.

      • carla says:

        Even if the half life has improved, I don’t see the point running for long periods at this stage if their priority is to see if they can commission a stable system that can consistenly run for periods much larger than the turn-around time.

        In the paper The LHC Machine, L. Evans, JInst 2008, the theoretical minimum turn-around is supposed to be 1.15 hours. The HERA machine had a theoretical min of 1 hour, put in practice averaged out as 6 hours and so the expected average for the LHC becomes 7 hours. With a luminosity life time of 15 hours, they give an optimum run of 12, 5.5 hours for 1.15, 7 turn-around times. If you’re interested, the paper is available for download, if it’s still there:


      • Philip Gibbs says:

        Thanks for the link. A good turnround takes about 3 hours but the average will be higher due to faults. Power glitches are always a problem, often requiring access and precycles that take time.

        One of the biggest problems in recent runs has been the vacuum spikes. i dont know much about them but perhaps these are related to the effect of the e-cloud or similar problems. These may get worse as intensity increases, but they may also get better as the natural scrubbing effect of the beams cleans the pipes out.

        This is why there is so much uncertainty about how much data they can collect. My estimate assumes some solutions to these issues but I would not bet on it.

  4. Luboš Motl says:

    It happens that I am the only one who gave the correct answer in your poll: in 2011, the LHC will delived more than 20 inverse PICObarns, indeed. 😉

    I can easily imagine how you feel to screw every line of the poll. 🙂

  5. Bill K says:

    Don’t count your 1/femtobarns before they hatch. The last few weeks they’ve been doing great, but it’s difficult to bet on a string of 124 good days in a row… Wait! Are you actually going to bet on this?

  6. […] a adivinar la luminosidad integrada total durante 2011, puedes jugar con Philip Gibbs en ”LHC Luminosity Estimates for 2011 (and poll),” viXra log, April 28, […]

  7. carla says:

    Phil, are you sure you have the right figure for the total integrated lunimosity at the top?

    If you go to http://atlas.ch/ and hover the mouse pointer over “proton run” you get a figure for the recorded integrated luminosity for 2010 & 2011 as currently 237/pb which includes the 43/pb from last year. It’s usually correct within 6 hours of the last run max and yours is showing 262/pb.

  8. […] couple of days ago we had some fun estimating how much integrated luminosity the Large Hadron Collider could collect this year. My estimate was […]

%d bloggers like this: