TEVSC group seminar 10 th June 2016 Chemistry

TE-VSC group seminar (10 th June 2016) Chemistry lab contributions for the diagnostic and cleaning of the CMS cold box contaminated with Breox B 35 oil CHARVET Colette DENIS Maxime (IUT chimie de Besançon) TEISSANDIER Benoit Courtesy Nebojsa Smiljkovic PH/UMC 1/29

Summary Introduction – Motivation Problem understanding and cleaning strategy q Analysis for solvent validation q Development of a cleaning quality control method q Development of a drying quality control method Chemistry lab contributions q Responsible for quality control during cleaning operation Conclusions B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 2/29

Introduction - Motivation CMS experiment Superconducting magnet (cold mass : 225 t) Nominal operation at 4. 5 K (-268. 65 ° C) Cooled with liquid helium B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 3/29

Introduction - Motivation CMS surface cryogenic systems Compression of helium gas (Adding "Breox B 35" lubricant during compression) Helium compressor station (on surface) Separation Breox B 35 / helium Gravity separator, coalescers, activated carbons Acceptable Breox concentration in helium lower than 10 ppb B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN Helium injection to the underground cold box 4/29

Introduction - Motivation CMS underground cryogenic systems The CMS Cold-Box Composed of turbines, heat exchangers, filters, etc. Feeding the superconducting magnet with helium 4. 5 K B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 5/29

Introduction - Motivation CMS cold box with nominal behaviour Stable temperature gradient over first heat-exchanger No pressure drop evolution over turbine 1 & 2 inlet filters B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 6/29

Introduction - Motivation 2015 CMS cold box run: clogging filters and loss of cold box efficiency Temperature gradient developing over first heatexchanger (°K) Pressure drop developing over B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN turbines 1 & 2 inlet filter 7/29

Introduction - Motivation CMS detector performance is degraded in absence of the magnetic field provided by its solenoid magnet. Thermal cycles can affect the life time of CMS magnet Need to quickly and clearly identify the Cold-Box problem Find a solution to solve it B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 8/29

Problem understanding - cleaning strategy B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 9/29

Problem understanding - cleaning strategy Gas chromatography analysis on helium CMS Cold-Box (CERN chemistry lab) No significant gas contamination condensing on the filters (e. g. expected : CO 2 / N 2 ) Liquide or solid contamination ? FT-IR analysis need to be perform on filter H 2 O Very small traces B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 10/29

Problem understanding - cleaning strategy Basic FT-IR spectroscopy analysis B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 11/29

Problem understanding - cleaning strategy FT-IR analysis on turbine filter 1: identification of the responsible for the pressure drop Turbine 1 filter analysis Breox B 35 main component 100 mg of Breox B 35 on filter generate the cold-box perturbation B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 12/29

Problem understanding - cleaning strategy New specific analysis methods were developed Around 50 complete analyses (EDMS reports) Reference Pure Breox B 35 CMS cold-box turbine contamination Contamination was clearly identified everywhere Breox B 35 oil B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 13/29

Problem understanding - cleaning strategy First meeting : 19 th august 2015 Clean the cold-box on site underground by solvent circulation during YETS (in January 2016) Task force : CMS - TE Leonel Ferreira : - Design, select and buy the cleaning machine - Select and validate an adapted solvent (efficient, safe, compatible) - Responsible for cleaning machine operation Chemistry lab : - Analysis for solvent validation - Develop a cleaning and drying quality control methods - Responsible for quality control during cleaning operation Cleaning machine TE-VSC-SCC mandate : B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN Polluted solvent (1 m 3) Distiller Clean solvent (1 m 3) 14/29

1 PA 7 M lv T oso Prom Vertrel ® Sion TM Analysis for solvent validation 1, 1, 1, 2, 2, 3, 4, 5, 5, 5 -Decafluoropentane Vertrel ™ XF B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 15/29

Analysis for solvent validation Cleaning efficiency tests of Vertrel XF against Breox contamination - Stainless steel samples were contaminated with Breox oil Test procedure (EDMS 997350) Before cleaning - Samples were cleaned with solvent - FT-IR analysis was performed to evaluate the solvent efficiency Breox B 35 contamination Results and report (EDMS 997350) Reference Vertrel ™ XF is efficient After cleaning No traces of Breox B 35 oil B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 16/29

Analysis for solvent validation Examples of material compatibility tests Drying speed (in air) of Viton seals after contact with Vertrel XF for 24 h 60 V Green Viton seal e. . . V Black Viton seal 40 e. . . 30 Pumpdown of valve placed in contact with Vertrel XF for 24 h 1. 00 E-02 20 10 0 Residual solvent “trapped” in seal 0 10 20 30 40 Drying time (hours) All Viton seals need to be replaced after cleaning Pressure [mbar] % of mass 50 0 1. 00 E-03 1. 00 E-04 1. 00 E-05 100 200 300 400 500 600 Valve open AFTER solvent Vertrel XF outgassing from seal Valve open BEFORE solvent contact 1. 00 E-06 Temps [s] Antonio Mongelluzzo (TE-VSC) B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 17/29 700 800

Develop a cleaning quality control method (online) [Breox] < 10 mg. l-1 B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 18/29

Develop a cleaning quality control method FT-IR In-Situ measurement of Breox B 35 in Vertrel XF Substrate Mirror 20 µl Breox B 35 + Solvent IR Beam Detector Solvent Evaporation Breox B 35 B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 19/29

Develop a cleaning quality control method FT-IR Calibration curve of Breox B 35 in Vertrel XF 100 Detection limit : 10 mg. l-1 Reflectance (%) 99 98 0. 02 µg 97 96 The analysis time : 5 min 95 VALIDATED method R 2 = 1. 00 E+00 94 93 0 10 20 30 40 50 60 70 80 90 100 110 120 Breox B 35 in solvent (mg/l) B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 20/29

Develop drying quality control method (After cleaning) Vertrel XF < 10 ppm (v/v) 2 B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 21/29

Develop a drying quality control method Lab setup to measure traces of different solvents in nitrogen: 3 methods UV-Visible Vertrel XF + N 2 M Spec ass trom (sniff eter er) FT-IR B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 22/29

Develop a drying quality control method FT-IR accessory modification to measure solvent vapour traces in nitrogen Grazing angle module (82˚) with gold mirror Gas inlet Gas outlet Handling FT-IR N 2 IR Beam Gold mirror B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 23/29

Develop a drying quality control method FT-IR calibration curve: Vertrel XF in nitrogen FT-IR - Abs at 1222 cm-1 1 Abs. a 1222 cm-1 1 10 100000 Vertrel XF in N 2 0. 1 VALIDATED method Calibration peak at 1222 cm-1 (CF 3 stretching) 0. 01 0. 001 Detection limit : 2 ppm Vertrel XF in N 2 (ppm v/v) B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 24/29

Develop a drying quality control method Sniffer (Mass spectrometer) calibration curve: Vertrel XF in nitrogen MS Signal (ppm de R 134 a) Mass 69 Linear(Mass 69 ) R 2 = 0. 9993 100 Ecotec E 3000 multi-Gas detector (Internal calibration with R 134 a ) 10 VALIDATED method Detection limit : 2 -3 ppm 1 1 10 100 Vertrel XF in N 2 (ppm v/v) B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 25/29

Cleaning and drying operations in CMS cavern B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 26/29

Cleaning and drying operations in CMS cavern Austin Ball (CMS Technical Coordinator) Team work during 1 month 4 circuits were cleaned by solvent circulation ~ 400 g of Breox oil were removed Equivalent to 4000 clogged turbines filters More than 100 in-situ analyses were performed by chemistry lab team Less than 10 ppm (v/v) of residual vapour solvent in the cold-box left after drying B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 27/29

Conclusions - perspectives CMS cold-box works very well !! Successful project for CMS, CERN, TE department Very nice team work !! June 2016 On going at the chemistry lab : - Breox B 35 content analysis in activated carbon (filters) is developed - Breox B 35 analysis by Gas Chromatography (GC/MSD) will be developed If needed, Dream team of the Chemistry lab is ready to repeat it somewhere else at CERN B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 28/29

e h t o t s Thank r's e n a e l C C. Charvet - TE/VSC C. Fabre - TE/CRG D. Majournal - TE/CRG J. Bremer - TE/CRG J. Gremion - TE/VSC L. Bardo - EP/ADO L. Ferreira - TE/VSC M. Taborelli - TE/VSC P. Maurin - TE/VSC R. Consentino - TE/CRG Thanks for your attention ! B. Teissandier, TE-VSC Seminar, June 10 th 2016, CERN 29/29