Cooling Working Group (CWG) Minutes #4

Date of the meeting:
06/03/97
Place of the meeting:
40 2-A01
Present:
G. Benincasa, P. Bonneau, M. Bosteels, M. Bozzo, A. Carraro, M. Edwards, G. Feofilov, J. Godlewski, R. Gregory, J. Gulley, M. Hatch, G. Hallewel, N. Lupu, C. Nuttal, T. Ohska, E. Perrin, P. Petagna, W. van Sprolant, V. Vacek.
Agenda:

Short Summary:

G. Feofilov (ALICE/St Petersburg) et G. Hallewel (ATLAS/Marseilles) presented their respective work on the evaporative systems intended for Tracker cooling. The two teams had different objectives and their state of progress and prospects also differed.
M. Hatch had drawn up an almost complete list of the heat sources and cooling requirements of the electronics in the ATLAS detector, which totalled about 2 MW,
C. Nuttal (TIS) was very doubtful about the massive use of a water/ice/methanol mixture in the caverns and strongly encouraged his colleagues to find other fluids for the detectors below 0°C which posed fewer safety hazards.

Detailed Minutes:
To be more precise, M. Bozzo would be co-ordinating cooling efforts in the CMS tracker, not on the whole CMS.
There had been a misunderstanding about the date of the Advanced Ceramics Corporation's presentation; it was to be given on 19/3 and not 19/2, and was therefore on the agenda for meeting # 5.
St. Petersburg was in charge of the cooling of the Inner Tracking System of ALICE and G. Feofilov gave an account of the state of progress with their R&D programme.
The ITS consisted of 3 sub-detectors, in the order starting from the beam pipe::
-Silicon Pixel Detector (SPD) - 450W
-Silicon Drift Detector (SDD) - 350-1900W
-Silicon double sided Strip Detector (SSD) - 2700W
Unlike ATLAS and CMS, the ALICE tracker did not need to be cooled to temperatures below 0°C and was to be run at 20°C. The SDD, however, required a thermal stability of 0.1°C and the whole of the ITS structure had to be kept within a very tight materials budget. St. Petersburg was therefore examining very thin carbon structures for the SDD and SSD with integrated cooling channels and their calculations showed that an evaporative system with fluorocarbons (C5F12) provided the best effectiveness/materials ratio. There nevertheless remained the problems inherent in that type of system:
- To achieve thermal stability (0.1°C), the evaporation pressure (Å 0.7 bar) had to be adjusted within a range of a few mbar for each exchanger in parallel.
- Owing to the appearance of the vapour phase with bubbles, the thermal transfer values varied over the whole length of the exchanger tube and a temperature gradient appeared.
To deal with the latter problem, St. Petersburg was tending towards a phase separation system, i.e. with a double-walled tube for the gas return.
Another problem examined was the influence of less thermally stable detectors on the SDD. It would probable be necessary to blow gas axially to limit the radial heat transfers.
The preliminary examinations and tests on a prototype were thus far from being completed to the point where a final design could be considered.
G. Hallewel (CPPM) took advantage of his presence at the meeting to report on the tests conducted at Marseilles for several years on an evaporative system intended for the ATLAS Pixel detector. The stakes were different, as there purpose was to cool the detector to around -7°C and the status was different, too, since the current choice for all the ATLAS Silicon detectors had fallen upon an underpressure binary ice system.
Nevertheless, the Marseilles collaboration strongly believed in its technique and G. Hallewel presented the complete file (technical installation and results) on the tests made with 4 ladders in parallel.
The table of the various possible fluorocarbons was (Ref. 3M-1995): Table 4-1

The selected fluid was C4F10, which made it possible to work in the -10/-20°C range with the circuit still at underpressure. There was no safety problem and the only unknown factor was the radio-chemical yield G which should be very low.
The installation was developing regularly and its layout was relatively complex. To put matters simply, the C4F10 at ambient temperature was pumped to the inlet of the 4 ladders where it was expanded through injectors. The inlet pipe simulated actual conditions with 16m of i.d. 4 mm tube on the main pipe and 3m of i.d. 2 mm tube at the inlet to each ladder; each ladder had 2 pneumatic valves (inlet and outlet) and all 4 ladders were installed in an insulated box. The vapour was collected at the outlet via an accumulator tank and recompressed towards the condenser, itself connected to a fridge unit. The pressure was monitored by a valve controlled by a pressure sensor at the ladder outlet. The recondensed C4F10 was then returned to the liquid circuit. A large number of measuring instruments was fitted in the circuit to measure the pressures, flow rates and temperatures and the whole system was monitored on LABView. The installation also had provision for the evacuation and bleeding of the circuits.
G. Hallewel presented various results corresponding to different configurations: 1 ladder, 2 ladders in parallel and 4 ladders in parallel, and load at 0.33 and 0.6 [W/cm2].
His conclusions were:

Questions:

Finally, G. Hallewel showed that in terms of the materials budget, an evaporative system with a modularity of 1 was equivalent to a binary ice system with a modularity of 4. That was thus a good argument in favour of the evaporative systems and Marseilles was encouraging the rest of the ATLAS SCT community to take part in their tests, as a great deal of work remained to be done to demonstrate the feasibility of such a system. In sum, the aim was to cool a prototype barrel and a prototype disc (several ladders in parallel at 0.6 [W/cm2] to -7°C with an underpressure system and a load loss of under 200 mbar in the exchanger tube.

The documents presented by G. Feofilov and G. Hallewel were available.

M. Hatch listed all the heat sources (electrical and electronic) dissipating into the main cavern and the annexes and arrived at a total of about 10 MW for the whole of ATLAS, 2 MW of that in the detector. A technical note ATL-TC-CERN-EDN-0004-00 was available, and the tables below were taken from it: Table 4-2

Certain sub-systems were already well defined and their integration was well advanced (e.g. the calorimeter electronics), while others had yet to be confirmed (binary ice in the SCT or electronics in the muon chambers) or were completely unknown to M. Hatch (vacuum beam bake-out or SCT/TRT active insulation). Those in charge were invited to contact him.

It was currently intended to install 3 binary ice machines in the cavern annex to supply 8 Si and 3 Pi circuits.
One of the great unknown factors was the method to be used to cool the power supply cables in the detector.
Leakless cooling was scheduled for the racks (control rooms and main chamber) and would be dealt with at subsequent meetings.

Various tests and the literature clearly showed that methanol was the best anti-freeze for the water or binary ice systems, in that it was the additive which resulted in the least increase in the viscosity of the fluid and hence in the materials budget.
The ATLAS SCT was to use a binary ice/methanol mixture of 25 to 30%, or about 600 kg of methanol in the cavern.
M. Edwards of Rutherford Appleton Laboratory Health & Safety Group gave an account of the measurements made there (Drager tubes) on the tanks containing the mixture at ambient temperature and at -10°C. The values measured represented no health hazard and the tanks would simply have to be ventilated or closed, as small quantities of vapour (1000 ppm) were detectable when the system was stopped (fluid not circulating).
For information:
Occupational Exposure Limits:
Long Term Exposure Level: 200ppm / 260mg/m3
Short Term Exposure Level: 250ppm / 310mg/m3

For C. Nuttal the problem was rather the flammability of the methanol and, perhaps, that of the mixture. In the more general context of safety in the underground experiments, TIS had been working for many years with the physicists to eliminate flammable gases from the detectors and, in the case of ATLAS, there was only one gas which currently raised problems, which was already an enormous success. C. Nuttal thus stressed the fact that authorising the storage and use of several hundred kilos of methanol in a cavern would be a very retrograde step where safety was concerned. Moreover, if, as desired, the final flammable gas were replaced in ATLAS, the use of methanol alone would require a very expensive fire detection system.

That was a very delicate point as it was controversial, and a decision would have to be taken quickly, as there were not many alternatives (TIS' comments on methanol also applied to acetone).
To mention only the ATLAS SCT (area at ~ -10°C, fluid at ~ -15°C), the pipework, which had already exceeded the allocated materials budget, were designed for a binary ice/methanol mixture. Any other mixture (glycol) or heat-conveying fluid (silicone oil, hydrofluoroether, Dynalene, etc.) would increase the material for the piping by at least 30%, either because they were more viscous or because they had a lower specific heat than water.
What remained, therefore, was either to work at high pressure to make up load losses and thus hope to reduce diameters, or to use evaporative systems (see item 2 above).
It had been pointed out that the last alternative was to review the operating temperatures of the silicon detectors upwards; the collaborations would in any event have to be aware of the technological and financial stakes entailed by wishing to operate at -10°C.
The next meeting would be held on Wednesday 19/03/97, Building 112, Room R018 from 10 a.m. to 12 noon.
Agenda:

CERN/P.BONNEAU/30/05/97