We ran over the pp500 data from 2011 looking for upsilons. Our goal was to get an idea of the yield seen in 2011 so we can make plans for the upcoming p+p 500GeV run in 2013 where we expect to have upwards of ~150/pb of integrated luminosity. This is what we found.

The raw signal (no pt cut) contains 577±41 upsilons:

This is more than enough data to make a strong cross section measurement and to seperate the 1S from the excited states. However, let's see how far in pT we reach:

The blue curve is unlike-sign candidates. The black curve is like-sign. All candidates pairs have a parent mass in the range 8-11GeV. We can see that there is a noticable seperation out to around 10GeV or so.

If we look at the 8-10GeV bin, we see ~35±10 Upsilons:

This is right on the limit of our detection capabilities, but we most certainly could not do a 1S seperation with this sort of statistics. However, it does suggest that such a measurement is possible with more data.

To be conservative, let's assume that we need 150 events to seperate the 1S upsilons from the excited states. This means we need to increase our statistics by a factor of ~4.5. In terms of numbers, that means ~2600 raw upsilon candidates, ~750M BHT1-triggered events, or ~100/pb of integrated luminosity.

Now let's move to trigger rates. We assume a base rate of ~230 Hz for the L0 trigger. Also, we assume a 25:1 rejection rate if we use the HTL to match a track to the L0 tower. Scenarios are summarized in the following table:

intLumi
Max. Prescale
L0 Rate
HLT Rate

160/pb	1.6	~140	~6
200/pb	2.0	~115	~4.5
300/pb	3.0	~75	~3
500/pb	5.0	~45	~2.5
750/pb	7.5	~30	~1

Note that we've negelected any increase in the base L0 rate that would come with increased luminosity. Furthermore, using the HLT would reduce the dataset size from ~750M events to ~30M events.