Minutes and Summary

Main purpose of the meeting: Data taking progress; BPT plot, lumi imbalance, MD2 plans, intensity increase plans, 2026 TS, SMOG issue

LPC minutes 28 July 2025

Introduction (Chris Young)

Paula Collins: Was the beam-beam test not 12h rather than the anticipated 9h, even though it was efficient?
Chris Young: We stopped doing the beam-beam test after 9h and then did Michi's end-of-fill MD. We actually stopped before they had taken all the data they wanted.
Paula Collins: I will check the times again including until the next fill.
Joanna Wanczyk: As we had a following activitiy the ramp-down wasn't considered as part of the beam-beam MD.
Chris Young: This won't be happening again next year so it isn't relevant for how much time we allocate for the lumi program in 2026.

Jorg Wenninger: At the moment we have set aside a week for the beam commissioning. We have started to look at a plan and if it is started 1 day earlier so there is 8 days then it appears this can work but it relies on keeping everything the same as it is now. We should try to be consistent; if we have people from the machine asking to change things then the assumptions used when considering shortening the VdM might not be valid and it would also take more commissioning time. Optics changes would significantly extend commissioning for example.
Filip Moortgat: It is important that there are no changes on the machine side if we are considering reducing the VdM program.

David Stickland: CMS doesn't have a 2-3% uncertainty on their luminosity. We are not prepared to say better than 5%, and we have a difference between the VdM and Z-counting results.
Eric Torrence: We do agree with each other better than 5% from each experiment. For a paper ATLAS would also currently not go below 5% but believe this is conservative which is why I gave you the 2-3% number, although I didn't realize it would go on a slide.
Filip Moortgat: CMS will have a slide on this later but we are not willing to go below 5% uncertainty.
Jorg Wenninger: A scale difference can explain the difference before the x-angle leveling but unless there is a x-angle dependence in one of the experiments the shift after the x-angle change cannot be explained.

Georges Trad: Is there any evidence that this is caused by effects local to the IP rather than those which would be seen by eg. BSRT.
Chris Young: The BSRT does show the beams growing over time, but it doesn't seem to jump when we do the x-angle leveling.
Georges Trad: But do you see the pattern within the trains that is shown here?
Chris Young: I haven't seen a bunch-by-bunch measurement.
Michi Hostettler: For this to be done precisely we would need to fix the camera position in a certain fill such that it doesn't move as this introduces differences at the level of the effects.
Georges Trad: For the test as the x-angle was changed quickly the camera shouldn't have moved. There is a pattern through the train seen in BSRT due to emittance growth. 
Michi Hostettler: There are two effects overlaid here. There is the emittance though the train due to time at injection in LHC/SPS and then wiggles that look like noise on top. This is not statistical but looks like LRBB effects.

Jorg Wenninger: One thing that could be done at the end-of-fill, so it is relatively cheap, is to check the local coupling knob to check we aren't doing anything silly.
Eric Torrence: Is there any correction for the solonoid? Could this not have been updated after the crossing plane rotation?
Jorg Wenninger: No, there are no solonoid corrections at the LHC. They have always been seen to be small, although CMS's should have a larger effect.

Jorg Wenninger: What do you mean by prefer? Is it not defined by intensity and emittance?
Chris Young: I believe that there will be separation leveling to the target.
Eric Torrence: For us it is testing the HL-LHC prototypes so 132 is preferred and if a variety of different luminosities are given then that is even better.

Georges Trad: Part of the merit of the MD is that the experiments would be keen to explore these range of pile-up?
Chris Young: The experiments have some prototypes that can be tested.
Filip Moortgat: Most the the detector is replaced so "keen" is a strong statement. We are not requesting this data.
Eric Torrence: ATLAS are generally keen on this data.
Paula Collins: If it is planned in advance what is needed from the VELO then we can consider this.
Jorg Wenninger: The heating for 600b should be low so it can be treated as a normal fill.
Jorg Wenninger: We can insert the Roman pots as in a standard fill.
Filip Moortgat: Agreed.

Jorg Wenninger: It isn't the orbit feedback but the crossing/separation plane optimization. In fact we could use this to optimize the position of ALICE/LHCb but there isn't real-time feedback which makes this difficult.
Silvia Pisano: It has been discussed with the physics board and management and 1cm is really the limit, and a bit of margin is appreciated.

Filip Moortgat: What would be the limit in intensity we could reach before we would have to change the scheme?
Chris Young: The rest of the slide explains when we get tight.
Jorg Wenninger: We would need to go slowly with cryo or more probably change the scheme.
Andrej Gorisek: Why do we not see much conditioning?
Chris Young: There is very slow conditioning, and it is not much slower than last year as the plot is on a shorter timescale.
Andrej Gorisek: Is there a chance it will kick-in and we can get there.
Chris Young: For the value needed for 1.8e11 this seems unlikely.
Georges Trad: Would there be a possibility of mixing trains of 4x36b and 3x36b to lighten the heatload?
Chris Young: The 4x36b scheme actually has some 3x36b trains and it is some of these becoming 2x36b that is the change to the suggested scheme. This seems optimal.

Paula Collins: This is potentially disastrous for LHCb but we will inquire. [on the length of the TS in 2026].
Eric Torrence: We will ask but ZDC doesn't take 3 days.
Jorg Wenninger: Some cool-down could be desired and using all of the MD time might be difficult.

Eric Torrence: Is the triplet heating limits and the S78 e-cloud limits at all connected?
Chris Young: I don't believe so -- we can check with cryo but these are limitations on different parts of the system.

ATLAS (Eric Torrence)

Filip Moortgat: The scaling means it was mu=65.5 with the adjusted lumi scale.
Jorg Wenninger: It is always difficult to know how long a downtime will be.
Chris Young: This isn't for accesses?
Eric Torrence: No this is other data-taking/calibrations/tests. For example the Cs scan of the tile which takes 8h. 
Michi Hostettler: We often don't know for sure so are usually conservative.
Eric Torrence: Even if it is just likely that it will be 6h for many of these tests this is useful as we can always come back 

Michi Hostettler: For the triplet limits we need to discuss with the cryo colleagues but many are away. The limitation is on the regulation to prevent flooding as this would result in a long stop so it is a question of how close to this we can go. ie. how much risk can be taken.
Eric Torrence: Why should it be different between point 1 and 5?
Michi Hostettler: In point 5 there is also one which is closer to the limit. But there are many mechanical things so it isn't exactly the same for all triplets.
Chris Young: I had an email discussion with Benjamin during the Oxygen run and it is clear from the plots he sent me that all 4 are different with the L and R differing at point 1 and point 5, and it just happens that one of those at ATLAS is most limiting. I will follow up with cryo if there is anything that can be done to raise the limits.

CMS (Andrea Massironi)

Chris Young: Are you going to update the MASSI files from before the TS with the new scale?
David Stickland: We have not updated the previous runs and don't intend to. They will continue to give the online calibration.
Filip Moortgat: Note that this 4% was done very quickly and other checks have large discrepancies with this, actually showing 2% in the opposite direction in one case. We didn't realize that this would be such a big deal as in previous years we weren't being limited by what luminosity scale we publish and now we are.

Chris Young: Would the limitation be due to the ECal degrading?
Andrea Massironi: This is one of the potential things that could cause this limitation? 
Andrej Gorisek: Is this proportional to the total taken luminosity?
Andrea Massironi: Not really, and it is also not an edge effect when it suddenly appears. We believe the thresholds set at the beginning of the year are high enough that we can continue, but have back-up scenarios of raising thresholds or reducing the pile-up.

Chris Young: The problem is that the online luminosity values are the best current knowledge at the time of data-taking. Obviously improving the cryo situation would be the best solution.
Filip Moortgat: In 2023, when I was LPC, I didn't believe any of the luminosity measurements from the experiments but relied on the beamspot size in z. This was one of the years where we got the most similar luminosities between the experiments. It doesn't seem desirable to limit based on numbers that change by large amounts at the end of the year.
Chris Young: There is a 1-1.5% effect on the total luminosity between leveling at 63 or 64.
Filip Moortgat: We only need a couple of % from cryo to mitigate this so hopefully they are able to find this after they are back from holidays.

ALICE (Silvia Pisano)

Chris Young: The plot you show here of the z position is flipped in sign with respect to the one on the LPC page?
Silvia Pisano: Yes, exactly this is our x,y,z definition rather than the LPC one.

Jorg Wenninger: There are no changes on the machine side that should do this. Also I am confused why some are quite extended in time?
Silvia Pisano: Some of this is that the recovery takes some time, and some systems might report they are in error a little later than others.
Chris Young: Were there the little ones before the TS and only the big ones are new?
Silvia Pisano: For the very serious events no and the beams are dumped, the smaller ones can probably be dealt with. We are trying to be more quantitative, but it is certainly not ideal as time is lost recovering etc.
Jorg Wenninger: Could we try to correlate with losses in the machine.
Eric Torrence: I think that one on Friday 18th is ATLAS's fault. There were bunches that blew-up in the ramp and then after some time in stable beams ATLAS inserted AFP and this resulted in the halo hitting AFP quickly and showering ALICE. This is what caused the dump.
Chris Young: There has been a fix in the ADT that prevents the blow-up in the ramp. So if spikes continue after the ADT fix was implemented then it must be something different.
Silvia Pisano: We can make plots for the fills after the ADT fix for the next meeting.
Chris Young: When you have these can you tell which side they come from.
Silvia Pisano: For the sub-detector that makes this plot no, but for the ITS we can tell which side they are coming from.

LHCb (Paula Collins)

Chris Young: Why can't the flip be done just after the MD block?
Jorg Wenninger: I think after should be ok, and we don't have an intensity ramp-up.
Michi Hostettler: What I would avoid is to do in the couple of fill after the MD, but you could do it in the 1st fill after the MD but this puts the tuning over the weekend.
Chris Young: Therefore we stick with what is on the slides.

Jorg Wenninger: It would be good to know if this can be done in 2 interventions, for example split over the Wednesday accesses in 2 subsequent weeks.
Paula Collins: We will pass this on and let you know the outcome of the meeting this afternoon.

Chris Young: There will probably be another FASER or SND emulsion exchange which will take similar to your temperature cycle so maybe this would be the easiest opportunity to do this again.