Here are some notes from today's meeting:
1) the power required for the whole ILC - (a partial rollup presented at Vancouver in July ~ 358 MW) - is now 274 MW. We (mostly) limited our discussion to linac power systems today. The linac RF system power use is 93 MW (unchanged from Vancouver). An additional 69 MW is used for support systems (29 for cryo, 22 for support electrical - e.g. klystron magnets/filaments, electronics, 17 for water and 1 for magnets). Non linac systems use 102 MW and will be discussed Friday.
2) some of the total ILC use is 'climate dependent' and will be perhaps 10% less on cold dry days.
3) roughly 10% of the (total ILC) power consumed is for tunnel air 'conditioning' - the term refers both to thermal and humidity control. This can be reduced most effectively by reducing heat rejected to air and overall tunnel air flow. There are some savings by using recirculated air. This is done at DESY.
4) The overall figure of merit ratio 'watts required to remove one watt power dissipated' is about 0.18 for the ILC. This is not far from existing machines. 26% of the dissipated power is removed via chilled water / 74% is removed through 'process' (LCW) water. We think this should be 15 / 85 or better. Both the power consumed and the capital cost of the installed CFS mechanical equipment will be reduced.
5) The klystron collector delta T should be increased to 40 degrees F. This allows 28% reduction of the process water pump-related power. Initial discussion with CPI indicates this should be ok. Heavy insulation of the collector and surrounding piping will be needed.
6) The baseline modulator / charging supply system needs modification to provide >0.95 power factor. There are no fundamental cost drivers or technical challenges associated with achieving an appropriate power factor. Nevertheless, the impact of 0.95 power factor (added distributed cable plant infrastructure and transformer heat dissipation) has not been included in the baseline design. There is a recent DESY note on this topic.
7) We recommend changing the baseline design to one process water skid / two RF units, rather 1/1. This reduces equipment cost and consumed power. We will investigate reducing the number of skids further; and try to identify an optimum.
8) We recommend a re-evaluation of the linac electronic (relay-rack) loads.