1. Test production run list from Devika:  first priority golden list, day 82-128, 477 runs, ~75.8 pb-1

2. MC luminosity: 1000/pb? ~10 times large than Data will be normalized

3. Zero bias events based on X-section from Pythia

   •  signal: 
       W⁺→ e⁺ν : 110k
       Wˉ→ e⁻ν : 35k

   •  background:  
      W⁺→ τ⁺→e⁺ ν : 110k   
      Wˉ→ τ⁻→ e⁻ν : 35k
      Z⁰→ e⁺e⁻ : 25k 

4. request ZB events for each run is besed on the luminosity 
ZBi=Li/Ltot * Embeding Sample (total) 

5. Here is the run list with ZB event require in the last colum. 

We will use the zerobias daq files contains enough events ( >= ZBi ). 

Ten runs have no zerobias daq files.
(14101053 14101054 14101055 14101061 14101062 14101064 14101066 14101067 14110060 14127047)

About 50 runs need more than one daq file. Remove runs with very low luminosity, ZB require < 10 events. 
 
6. Estimation of the disk space for the zerobias daq files.

I tested to request daq files of 3 runs from HPSS.   

3.2G    st_zerobias_adc_14082029_raw_2440001.daq  185 events
1.7G    st_zerobias_adc_14082031_raw_2440001.daq    95 events
4.3G    st_zerobias_adc_14082033_raw_2430001.daq  201 events

We have 474 runs. The average ZB events are ~230, so 4GB will be the average footprint of the runs. About 2-3 T disk space is needed. 
If we have enough disk space, we can copy all the dap files. If not, we can do this in 2 or 3 times.

7. Procedure of embeding
  a) Generating MC events by PYTHIA. 
    key words: vertex distribution; kumac file; fzd file ...
  b) Embedding MC events into zero bias events and producing the muDst files. 
    key words: fzd file; daq file; bfc macro; geometry; BFC options..