Electrolytic Capacitor Tests Konstantinos Papastergiou TEEPC 25 May

Electrolytic Capacitor Tests Konstantinos Papastergiou TE-EPC 25 May 2018

Summary § § § Background work Motivation of the test Test objectives Test protocol Test Results Conclusions Capacitor banks for energy recovery

Electrolytic Capacitors for Energy Storage Purposes • Applications: Energy recovery of power converters supplying magnets - SIRIUS power converter family (e. g. TT 2 Consolidation Programme) - Future medium power converters with energy recovery capacitor banks • Benefits: - Increased energy efficiency of power converters supplying magnets - Improved power quality from power network Single brick SIRIUS power converter Master Thesis work by Luca Sburlino

Single Capacitor Ratings: Worst Case Requirements Worst case cycle acting on single capacitor Capacitor bank Min – max voltage Cycle duration (freq. ) Configuration 600 V – 900 V 1. 2 s (0. 833 Hz) 3 series x 5 parallel Single screw-terminal Al-ELCAP ratings RMS voltage 400 V RMS current 11 A Peak current 50 A Voltage swing 300 V – 206 V Capacitance Ambient temperature Minimum Lifetime ≥ 29 m. F 40 ºC 100’ 000 h operation Master Thesis work by Luca Sburlino

Construction Technology • Enhanced heat dissipation by using: - Extended cathode (axial direction) - Corrugation (radial direction) - Thermal conductive glue (may clog safety vent) Screw-terminal aluminium electrolytic capacitor

Manufacturing Process • Increase capacitance density: Anode and cathode foils etching (1) surface increase of 60 x for 400 V capacitors • Obtain capacitor rated voltage: Anode foil forming (2) Vformation ≈ 540 V for 400 V capacitors • Increase capacitor quality, reduce early failures: Ageing (7) Vr < Vageing < Vformation Manufacturing process

Lifetime Influences: Charge-Discharge Applications Tab-heating in charge-discharge applications Charge-discharge applications: Elevated heating on tabs. Consequences: Local electrolyte vaporisation and damage on tabs more likely Transient Thermal, Electrical and Lifetime Analysis of Large-Can Al-ELCAPs, Cornell Dubilier – Sam G. Parler, Jr. , P. E.

Lifetime Influences: Charge-Discharge Applications Ageing and failure modes diagram

Motivation for the Tests by the manufacturer § § Previous experience with capacitive energy storage shows the need to understand the failure mechanism of capacitors. To have enough input in order to decide compensatory safety measures ú ú ú ú Visualisation of the cabinet (video) during the test Visualisation of the vapour release from the cabinet (video) during the test Visualisation of the test area (video), to assess the potential projection of debris Measurement of the voltage/current during the test Measurement of the sound pressure at 1 m from the cabinet Photos of the cabinet (exterior+interior) after the test Detailed photos of the capacitor bank after the test Qualitative assesmnet of impact to the floor.

Objectives § § § Qualitative assessment of the impact of a capacitor failure Verify Sirius metallic enclosure withstand to a destructive failure Qualitative assessment of the impact on surrounding equipment ú ú ú § From heat and fire From debris and flying parts From the electrolyte evaporation Identify the audible magnitude of a failure

Test Protocol § Test 1: Continuous rated peak voltage across a capacitor ú ú ú § Test 2: Continuous over-voltage (x 2) across a capacitor ú ú ú § Single point of failure will result in this condition Withstand over long time to be determined Repeat on three different samples A double failure will result in this condition Worst case if energy source not disconnected Repeat on three different samples Test 3: Reverse polarisation ú ú ú Reverse polarisation of a capacitor May occur as a result of manufacturing error Repeat on three different samples EDMS 1919992

Test setup description § Test 1 & 2 ú ú § Overvoltage on DUT With and without external source Test 3 ú ú Reverse polarisation of DUT Gradual increment of source voltage

Test 1 Results: continuous peak voltage Shot 1: 450 V on DUT -> no failure during 1 h Shot 2: 550 V on DUT -> instant destructive failure Shot 3: gradual increase 450 V to 520 V on DUT: • Failure at 520 V after 1320 sec

Test 2 Results: over-voltage (x 2) Shot 1: 900 V on DUT with external source -> failure after 73 s Shot 2: 900 V on DUT, external source disconnected -> no failure, DUT operational after test Shot 3: 900 V on DUT, external source disconnected -> no failure, DUT operational after test Shot 4: 900 V on DUT with external source -> destructive failure Test 2

Test 3 Results: reverse polarisation Shot 1: Reverse polarisation of DUT with external source -> failure, open vent

Summary of test results • Test 1 & 2: single point of failure • Over-voltage <37%** : No failure after 1 hour of exposure • Over-voltage >37% with Sirius capacitor banks energy: Failure mode is uncertain • Over-voltage >37% with Sirius capacitor bank energy and external energy source: Destructive failure mode after some time • Test 3: polarity reversal • Failure with open-vent after 900 sec if source remains energised. ** Over-voltage with respect to the application operating voltage

Sirius Capacitor Failure Tests - Conclusions • Electrolytic capacitors are prone to failure if exposed to over-voltage. • There is no evidence that energy stored in a Sirius brick can cause by itself destructive failure or fire. • A failure may become destructive in case an energy source remains energised after the event. • Key findings: • Destructive failure of capacitors (fumes, debris) is contained by the enclosure • Destructive failure of capacitors has no structural impact on the Sirius metallic enclosure • Destructive failure has limited impact on the equipment directly above • The peak observed sound level was 135. 5 db at 0. 5 meter from the cabinet A report is circulated internally for approval and will be placed under CERN-0000189820 *Destructive is a failure that results on visible damage to the capacitor can or its surroundings *Reports and media in EDMS folder CERN-0000189820