Forward Pixel Cooling System Information

1
Forward Pixel Cooling System Information

• Joe Howell

2
Sources of Information

• Presentation document by Paola Tropea
• Engineer CERN TS/CV Department
• CMS Tracker Cooling system overview presentation
  Dec 03, copy captured in docdb 590
• Very good overview of the entire CMS Tracker
  cooling system
• Invitation to Tender IT-3230/TS/CMS for cooling
  system
• Technical Specification for design, construction
  and installation of a double loop fluorocarbon
  150 kW cooling plant for the CMS Tracker and
  Preshower detectors
• From CERN EDMS document system
• One of many related documents, yet to review
  others
• Meeting with Scott Reeves and reference
  information Scott provided
• Fermilab AD engineer responsible for Fluorinert
  cooling systems
• Notes and much reference information captured in
  docdb 571
• Meeting with CMS Tracker engineers during Dec 05
  CMS week
• Paolo Petagna,  Hans Postema
• My notes and others to be captured in docdb 623
• Presentation material and notes from meeting with
  Kurt Gabathuler, PSI
• The contact person for barrel mechanics
• Very familiar with forward pixel
• Lots of information mining still remains to be
  done

3
Overall system diagram
These units must operate continuously, even in
case of power or water cut, and they will be
cooled by an emergency refrigerator, using the
available Emergency Power Supply (EPS, provided
by CERN).
4
Location of pixel cooling rack
5
Specifications provided by Kurt Gabathuler, PSI
Forward Pressure map similar to this Barrel
version
6
Pixel secondary system diagram
7
Hydraulic symbols
8
Details of secondary system branches
Pixel Detector
9
Pixel secondary system notes

• 4 branches for Z side and 4 branches for Z side
• Each branch has
• a manual throttle valve to set flow
• Pneumatic shutoff valve at inlet and outlet
• A calibrated orifice plate
• Schrader (self sealing quick disconnect) valve on
  each side of orifice on supply side which could
  be instrumented to measure flow
• Schrader valve on return side
• Reservoir tank is instrumented with a strain gage
  balance to measure weight change for leak
  detection
• The CERN engineers described a system to perform
  a leak check on each branch by isolating each
  branch with the pneumatic valves and then
  monitoring for a pressure drop/rise. (I have not
  seen a PID diagram for these elements yet)
• My understanding of the CERN engineers
  philosophy for leak detection/action is If you
  have a small leak that is not detectable with the
  reservoir balance you would not stop running even
  if you knew it existed. If the leak is large
  enough to be detected with the balance you would
  then proceed to leak check each branch and then
  leave that branch turned off if necessary.

10
Notes from discussions with CERN engineers

• They are well aware of Fluorinert/elastomer
  compatibility issues
• Fluorinert decomposition
• Charcoal filters were tested and shown to remove
  Fluorine ions
• Paola Tropeas presentation lists Alumina filters
  for acid removal
• Cleaning and material selection to remove
  hydrogen sources was emphasized
• Discussed continuous chemical monitoring vs
  periodic sampling
• They requested a concise description of
  conditions that caused problems in Fluorinert
  system at Fermilab
• CERN engineers expressed concern with long term
  integrity of fluid connections
• Seemed less confident of radiation resistance of
  epoxy used at blade connection points
• Also suggested metal connection instead of
  silicon tubing
• They were interested in 3M document that gave
  guidance on how to seal fluid connections. I got
  the impression they have struggled with leaks in
  an existing system.
• Thermal screen is intended to maintain un-powered
  silicon cold in the event of a power outage or
  primary system failure
• Sniffer lines for smoke detection in tracker
  volume
• Supply and return lines are copper and there is
  some experience with copper/aluminum
  compatibility during tests of the thermal screen

11
Miscellaneous notes

• Experience with a radiation induced Fluorinert
  decomposition in Fermilab AD led to extensive
  investigation
• Information exchange with 3M and others
• Special filter developed by 3M is used in all
  applications
• My impression is that every novice user of
  Fluorinert has problems with leaks and leak
  detection

• Notes from discussion with Jeff Spalding about
  integration area infrastructure
• C6F14 Plant capable of delivering 10 C (uses 6
  C chilled water CERN standard)
• Could be upgraded to run colder
• Interlock system (one to share with Pixel and
  Barrel) 16 temp, 8 humidity, 8 outputs
• Could be loaned when available