Development of a portable CO 2 cooling system

Development of a portable CO 2 cooling system for CO 2 R&D in HEP Bart Verlaat 2 nd October 2013

Traci project • Development of a portable CO 2 laboratory cooling unit called Traci – • TRACI=Transportable Refrigeration Apparatus for Co 2 Investigation. Development in AIDA framework together with interested partners. – – – Nikhef & CERN lead development with engineering support of the Cracow University of Technology. Co-funding from clients Collaboration with Sheffield, Oxford, Liverpool and Milano? • • Would be great to boost the project to the next level! Designed for applications like: – – – Test beam telescope (AIDA) Micro channel development (LHCb) Pixel development (CMS) Detector thermal/mechanical support structure development (Atlas IBL, ILC-TPC) Detector commissioning (Atlas IBL) • Focus on simplicity for easy production in series • Focus on simplicity for easy non expert use Bart Verlaat 2 nd October 2013

Traci design overview • The need for laboratory CO 2 cooling is in the order of several 100 watt from room temperature down to -40ºC. – Capacity is a function of temperature. – A kw version under design in collaboration with GSI (Traci-XL) • Stable CO 2 temperature, heat load independent • System must be easy to operate by non expert users – On/off button + temperature set point – Stand-alone operation (PC only for monitoring or sending set point commands) • Simplified concept called I-2 PACL. – Modified 2 PACL concept developed for AMS and LHCb. • Several functions are integrated for a simpler use and operation • 2 PACL = 2 -Phase Accumulator Controlled Loop • I-2 PACL = Integrated 2 PACL (several 2 PACL functions are integrated into 1 component) • I-2 PACL concept is patent pending (CERN&Nikhef) Bart Verlaat 2 nd October 2013

I-2 PACL concept Integrated 2 -Phase Accumulator Controlled Loop Patented ! (patent owners: CERN & NIKHEF) 2 -Phase Accumulator Controlled Loop Used in LHCb-Velo, Atlas IBL & CMS-Pixel Bart Verlaat 2 nd October 2013

Transportable Refrigeration Apparatus for CO 2 Investigation TRACI-2 (Uni GVA, CMS, Ke. K) TRACI-1 a (LHCb) TRACI-1 b (Atlas) • • • Brrrrrrr! Portable laboratory cooling unit Cooling power <100 W – 250 W> Temperature range <-400 C; +200 C> Turn key Very simple to operate ”fridge like” Bart Verlaat 2 nd October 2013

Typical Traci performance, stable cooling over a large temperature range Experiment Evaporative CO 2 Liquid CO 2 Sub cool protection Bart Verlaat 2 nd October 2013

Current Traci Capacity Traci capacity is comparable with a Julabo FP 40 -ME thermostatic bath cooler Julabo FP 40 -ME Traci-2 Prototype Bart Verlaat 2 nd October 2013

Door panel with control interfaces Traci 1&2 design CO 2 transfer hose Conditioner and relay based control logic Other side Foam box with CO 2 piping CO 2 accumulator CO 2 liquid pump Off-the shelve R 404 a condensing unit Bart Verlaat 2 nd October 2013

Traci experiences Traci set-up in Atlas SR 1 • Traci set-up for IBL stave tests Traci-1&2 LHCb, Atlas and CMS has now been running for a decent amount of time. Several lessons learned: – Operation of the systems turns out to be easy, no experts were needed for operation. – Protection against empty start up incorporated. Users are not looking at the numbers apparently (also when you tell them to do so…). – Serious pump wear problem, chosen gear pump is not suitable for long term CO 2 pumping. Traci set-up for LHCb micro channel tests Bart Verlaat 2 nd October 2013

Pump issues Too small? HNP pump • CO 2 is a problematic fluid to pump – Very low viscosity (15% of water) • • – • Currently using Gather gear pumps Long duration use of current Traci’s show short and variable life times of Gather pumps. Exploring different new pump options – – • High system pressure No reliable small pump on the market found so far. – – • Limited or no lubrication Low pumping efficiency Large and expensive HNP pump Problem giving gather pump HNP pump 3 m pump Lewa pump (LHCb style) Design special pump in house 3 Mpump under study (Too big for a small Traci) LHCb pump with excellent track record too big for small systems. – Being installed in KEK system, but severe performance degradation at colder temperature Lewa pump: reliable but heavy (and pulsating) Bart Verlaat 2 nd October 2013

Gather pump behavior Atlas-SR 1 1 pump, slowly wearing out 1 st dead 2 nddead CMS 3 pumps in half of the SR 1 time Still alive Why CMS dies and SR 1 not? No idea. . Bart Verlaat 2 nd October 2013

Traci SR 1 Elapsed time Zoomed in Run time without stops Bart Verlaat 2 nd October 2013

Traci CMS Bart Verlaat 2 nd October 2013

Pump requirements • For a 300 Watt system based on 50% VQ – – – – Liquid density range: 780 - 1130 kg/m 3 Liquid viscosity range: 0. 07 - 0. 2 c. P (0. 07 at start-up, nominal 0. 15 to 0. 2 c. P) Trange: -50'C to 20'C Flowrate: 5 g/s => 270 ml/min Minimum sub cooling of inlet liquid: 3'C Pressure head: 10 bar (at 0. 15 c. P viscosity) Static design pressure: 110 bar Efficiency: >15% (d. T over pump must remain <3'C, this is work and external heatleak. Typical Work is d. Px. VF=>10 bar*270 ml/min=5 watt, Qpump total (Qext+Work) is ca. 30 Watt at -40'C) Backwards leak is not a problem as long as the over effeciency remains above the requested 15%, infect the efficiency is backward flow related. Low pulsation >25 Hz Oil free Leak tight < 0. 1 g/*y CO 2 Limited liquid pollution due to wear. (Avoid rubbing concepts) Maintenance: Pump should be able to operate full time without intervention for at least 1 year. § What are your specific needs? Bart Verlaat 2 nd October 2013

HNP 7255 pump Requirement Current Traci Bart Verlaat 2 nd October 2013

Future Traci developments • Now working on Traci-3, with improvements on: – Control: PLC control included – Different pump • Traci-3 is build with AIDA funds Traci-3 design Traci-3 assembly has started Bart Verlaat 2 nd October 2013

Future goals of the Traci project GOAL: General purpose CO 2 laboratory cooler competitive with water / glycol / silicon oil chillers Size reduction • • • CO 2 -I-2 PACL is clean CO 2 -I-2 PACL has a very accurate distant temperature control Small hardware in experiment Bart Verlaat • High risk of contamination of expensive equipment by leaking Fluids 2 nd October 2013

Traci-2 for KEK/Desy • For reliability reasons Traci. Ke. K was supplied with a Lewa pump (LHCb - Track record) – Performance degradation expected due to too large pump. – Ke. K specs allow a performance degradation. – Question for us was: How much? • New configuration, new surprises? • Status of today: – Exploration has started and 1 st signs look positive – Further tuning needed to deliver a solid product and give us the experience for further development. Bart Verlaat 2 nd October 2013

Traci Ke. K (With Lewa) Bart Verlaat 2 nd October 2013

Traci-2 capacity estimate Preliminary Ke. K Requirement Traci-Lewa (0. 16 lpm) r he t Ga iac Tr 250 W accu heater Traci-Lewa (0. 28 lpm) Loss of cooling power at low temperatures Calculated capacity based on 50% vapor quality Bart Verlaat 2 nd October 2013

Traci collaboration • If Liverpool, Oxford, Sheffield and Milano can join that would be perfect! • Work packages: – Redesign Traci concept for easy production (Outsourcing, so we can order for demand) – Prototype production – Pump development • With industry • Test stand at CERN – Control development • Under control at CERN • How to continue? • I look forward in a good working collaboration! Bart Verlaat 2 nd October 2013