Page 1 of 1

How stringent is the ~100fb for 2016-2018?

Posted: Mon May 30, 2016 12:57 pm
by jmc2000
For 2016, they're hoping to accumulate ~25fb. But suppose instead they only manage ~15fb and ~80fb for 2016-2018. Will they extend the run for another year, say, or just accept what data has been accumulated?

Regards,

Jmc

Re: How stringent is the ~100fb for 2016-2018?

Posted: Mon May 30, 2016 9:27 pm
by DCWhitworth
jmc2000 wrote:For 2016, they're hoping to accumulate ~25fb. But suppose instead they only manage ~15fb and ~80fb for 2016-2018. Will they extend the run for another year, say, or just accept what data has been accumulated?

Regards,

Jmc
At a guess I'd say that's far too speculative a question to get an answer to right now. I would suspect the limit is not very important in itself, much would depend on whether the experiments feel they're close to getting some sort of result nearer the end of the run.

In the first run, they extended it for another year because the data was showing hopeful signs of there being a Higgs signal and more data would help them resolve this and it was felt the benefits of getting this result outweighed putting off the shutdown.

So I expect any decision will be based not on the abstract "Did we hit our data target ?" but on "will more data likely help us resolve any outstanding anomalies and is it worth postponing the shutdown for this ?"

Re: How stringent is the ~100fb for 2016-2018?

Posted: Wed Jun 01, 2016 11:14 pm
by jmc2000
Currently at 2048b, the LHC is accumulating ~1.5fb/week for 30% stable beams which is great news, considering the lost two weeks over bad luck. If they can get to 2800 bunches, that should mean ~2fb/week bringing 25fb/year on target!

Cheers,

jmc

Re: How stringent is the ~100fb for 2016-2018?

Posted: Wed Jun 01, 2016 11:54 pm
by terryburton
jmc2000 wrote:Currently at 2048b, the LHC is accumulating ~1.5fb/week for 30% stable beams which is great news, considering the lost two weeks over bad luck. If they can get to 2800 bunches, that should mean ~2fb/week bringing 25fb/year on target!
At the moment they appear to be limited to 72 bunches per injection due to a vacuum fault in the SPS dump. It doesn't look as though they can reach 2800 bunches whilst preserving the current inter-batch spacing.

Does anyone know whether they can reduce these gaps or whether they impose a limit?

Has anyone found the plans for resolving the issue with the SPS dump? During a technical stop, maybe - or perhaps this is something they will have to live with for longer?

Re: How stringent is the ~100fb for 2016-2018?

Posted: Thu Jun 02, 2016 11:49 am
by mfb
It is still unclear how long repair of the SPS vacuum leak will take. We'll have to live with the gaps in the meantime - they are the time necessary to ramp up/down the kicker magnets for injection, so they cannot be shortened. They are 38 bunches long, so ideally we get 72 bunches per 110*25 = 2750 ns. The LHC ring is 88924 ns long, 3000 ns space are necessary for the LHC beam dump, so we could get at most (88924-3000)/2750 = 31 injections with 72 bunches each, that would be 2232 bunches. For actual operation you want to start with a probe beam, then just a few bunches, and only then go for 72 bunch trains, so ~2100 is more realistic. We are at 2040 already, another train of 72 might fit in but probably not more. Source
Heat load in the magnets is another limitation, and we are close to the heat limit now. Power per beam intensity tends to go down slowly over time, so more bunches become possible in the future.

Edit: The necessary low-intensity injections don't allow more bunches - we are stuck at 2040 until the SPS allows the injection of larger bunch trains. With 288 bunch trains, 27xx bunches are possible.

Coming back to the original question: The 100/fb is a rough guideline what to expect, so the experiments can start studying which precision the analyses can reach. Those results don't change much if we get 80/fb or 120/fb, the plan won't be shifted for that. The long shutdown might be shifted if (a) something really surprising shows up and it is expected that a bit more data can improve the measurements notably (unlikely - 2019 won't be as in 2012 where we could take 4 times the integrated luminosity of the previous year) or (b) more than one experiment raises concerns about being in time with their experiment upgrades.

Re: How stringent is the ~100fb for 2016-2018?

Posted: Sat Jun 04, 2016 1:17 am
by jmc2000
terryburton wrote:
jmc2000 wrote:Currently at 2048b, the LHC is accumulating ~1.5fb/week for 30% stable beams which is great news, considering the lost two weeks over bad luck. If they can get to 2800 bunches, that should mean ~2fb/week bringing 25fb/year on target!
At the moment they appear to be limited to 72 bunches per injection due to a vacuum fault in the SPS dump. It doesn't look as though they can reach 2800 bunches whilst preserving the current inter-batch spacing.

Does anyone know whether they can reduce these gaps or whether they impose a limit?

Has anyone found the plans for resolving the issue with the SPS dump? During a technical stop, maybe - or perhaps this is something they will have to live with for longer?
The 126th LHCC meeting a week back goes as far as giving a luminosity plot on page 27 for running at 2100 rather than 2800 bunches for the whole of 2016. On page 23 it says a spare of the SPS dump is being prepared:

https://indico.cern.ch/event/527359/con ... tus_EB.pdf

Looking at the latest schedule dated June 1st 2016, the previous projection is probably based upon the current ~1.5fb/week at 2028 bunches:

https://espace.cern.ch/be-dep/BEDepartm ... e_2016.pdf

JMc

Re: How stringent is the ~100fb for 2016-2018?

Posted: Sat Jun 04, 2016 9:10 am
by sciing
Thanks for the details, this confirms why no further intensity increase was done.
We have another >300pb run with initial lumi above 8e33, so just 6 runs like this per week for 2fb/week.

Re: How stringent is the ~100fb for 2016-2018?

Posted: Sun Jun 05, 2016 1:01 am
by mfb
2/(fb*week) is challenging - possible, but needs more than 50% of the time in stable beams (55% assuming an average of 6000 Hz/µb). Usually 1/3 is more realistic, for ~1.4/(fb*week), about the plan linked above.