SF6 Management Challenges Version 2016 & Beyond Lukas Rothlisberger CEO DILO Company, Inc.

Topics of Discussion General Info / Current GHG Reporting Legislation Identifying Major SF6 Management Challenges Emission Types Tracking Problems SF6 Alternatives

General Information – SF6 Properties Manmade Colorless / Odorless Inert Non Flammable / Non Toxic / Non Poisonous Unmatched Dielectric Strength / Arc Quenching Capabilities Great Thermal Stability / Heat Trapping Re-Association / Self-Healing After Arcing Event

General Info – GWP – Global Warming Potential

SF6 – The Most Potent & Highly Persistent GHG Relatively few uses Electric power equipment – high voltage transmission and distribution pressurized equipment Electronics manufacture – etching silicon wafers Primary magnesium production – inert cover gas Miscellaneous low-volume applications (tracer gas, medical applications, insulator in accelerators, AWACs) Relatively small emissions Less than 1% total US annual GHG emissions Worst of the worst greenhouse gas 22,800 greater heat trapping capacity than CO2 over 100yrs “GWP100” 1 lb. SF6 = 22,800 lbs. CO2 Highly inert - 3,200 atm. lifetime

GHG Pollution By Sector and Gas Type

SF6 Equipment Lifecycle Emissions Emission Sources – mostly inadvertent Equipment Production Installation Servicing Operation – intermittent leaks Decommissioning – venting discouraged but not banned Reduction strategies – best handling and management Leak Detection and Repair Recovery/Recycling Employee Education/Training Track emissions

US SF6 Emission Reduction Drivers EPA Voluntary SF6 Emission Reduction Partnership for Electric Power Systems Emissions vary from <1% to over 12%; avg. 2.4% rate 2011 - Greenhouse Gas Reporting Program Large emission sources only No mandatory emission reduction requirements planned State Regulation – 1% emission rate by 2020 2012 – California emission standards 2013 – Massachusetts emission standards

SF6 Environmental Regulation – US EPA Mandatory Greenhouse Gas Reporting Rule (40 CFR Part 98) Subpart DD (Users of Electrical Equipment) and Subpart SS (OEM’s) US EPA Requires SF6 users with at least 17,820 lbs. SF6 Nameplate capacity to report emissions annually

GHG Pollution Shaming http://GHGdata.epa.gov/ghgp/main.do EPA data publication tool allows stakeholders and the public to access the key data elements and to sort data by location, sector, and by gas. 2014 data now available

Environmental Reg.’s – CA Air Resources Board (CARB) Established annual emission rate of 10% compared to nameplate capacity Requires GIE owners to reduce annual emission rate by 1% per year over a ten year period Beginning 2020 sets max emission rate not to exceed 1% Massachusetts passed similar Reg. starting 2013

Reported Emissions Actual / Hard Emissions Leakage from GIE Handling Emissions Tracking Emissions Recordkeeping Inaccuracies Nameplate Issues Nameplate Discrepancies Nameplate Inaccuracies

Major Challenges for SF6 Users Additional reporting requirements Operational / Procedural Inexperience / Infrequent SF6 handling Handling / Measuring equipment Missing info on work orders Tracking / SF6 Management Nameplate issues Data transfer not in real time Patchwork of different customized programs and Excel spreadsheets to monitor inventories and emissions Disconnect between E&H and field personnel Cylinder tracking / Multiple entry points for cylinders

Real World SF6 Emission Tracking Problems 4/14/2016 – Substation personnel removes SF6 from GIE containing 340 lbs per nameplate. Upon completion, 245 lbs. is stored in cylinders. 5/10/2016 – Environmental questions discrepancy – Substation personnel blames inaccurate nameplate Problems & Possible Errors: Long delay between work being performed and records being updated Density was not checked/recorded prior to gas recovery Recovery blank-off pressure unknown – incomplete recovery Residual SF6 in recovery system not accounted for Weighing inaccuracies (weight scale and/or cylinder TW) SF6 emission on the pressure side of equipment (i.e. recovery system/hose leak) Nameplate Inaccuracy All of the above

Eliminating SF6 Recovery Emissions

Impact of Blank-off Pressure on SF6 Recovery GIE containing 340 lbs @ 85 PSIG Recovery to 0 PSIG / 760 Torr 290.0 lbs removed / 60 lbs lost Recovery to 200 Torr 326.0 lbs removed / 14 lbs lost Recovery to 50 Torr 337.0 lbs removed / 3 lbs lost Recovery to 5 Torr 339.7 lbs removed / 0.3 lbs lost Personnel recovering should be instructed to always reach a blank-off pressure of 5 Torr / for GIE containing < 50 lbs 35 Torr

Step by Step SF6 Recovery – Verifying Nameplate Verify Temperature/Pressure for proper density Any deviation in pressure will result in nameplate discrepancy Recover to < 5 Torr (< 35 Torr for GIE containing < 50 lbs) Difference in recovered SF6 not measurable in GIE containing < 50 lbs Use calibrated mass flow scale or weight scale Mass flow scale preferred as it eliminates cylinder TW inaccuracies If using weight scale verify that residual SF6 has been removed from recovery system Document blank-off pressure

Real World SF6 Emission Tracking Problems Sample Utility - Equipment Owned CB 1 Nameplate 100 lbs. TOTAL NAMEPLATE 100 lbs. Cylinder A Empty Cylinder B 115 lbs. TOTAL SF6 IN SYSTEM 215 lbs.

Real World SF6 Emission Tracking Problems Calendar Year SF6 related work to be performed 2016 CB1 to be maintained 2017 CB1 to be decommissioned 2016 CB1 maintenance Recovered 100 lbs. from CB1 and stored in Cylinder A Filled CB1 from Cylinder B – it required 115 lbs. to get to proper density

Small Errors = Potentially Huge Impact What went wrong during SF6 Recovery? Temperature/Pressure was not checked prior to SF6 Recovery Personnel unaware of remaining SF6 in Recovery System Incomplete recovery Nothing How could this have been avoided? Real Time Tracking System immediately alerts to discrepancies Utilizing mass flow scale Better understanding of SF6 Recovery System (Training) What went wrong filling CB 1? Nothing (assuming CB1 was filled to proper density)  Why does any of this matter?

Status 12/31/16 Sample Utility - Equipment Owned CB 1 Nameplate 100 lbs. / Real 115 TOTAL NAMEPLATE 100 lbs. Cylinder A 100 Cylinder B Empty TOTAL SF6 IN SYSTEM 215 lbs.

How is this event reported? User Emission = (Decrease in storage inventory) + (Acquisitions) – (Disbursements) – (Net increase in total nameplate capacity of equipment operated) Calculating without nameplate adjustment: (15) + (0) – (0) – (0) = 15 lbs. Emission Calculating with nameplate adjustment: (15) + (0) – (0) – (15) = 0 lbs. Emission USEPA Currently does NOT allow nameplate adjustments at time of maintenance for the purpose of calculating emissions, but requires this at decommissioning and to calculate emission rate CARB allows nameplate adjustments if confirmed by OEM

(-115) + (0) – (0) – (-100) = 15 lbs. (negative emission) 2017 CB 1 decommissioned – How is this reported? Without nameplate adjustment in 2016 Status 1/1/2017 Status 12/31/2017 CB1 100 lbs. Decommissioned Cylinder A 100 lbs. 100 lbs. Cylinder B Empty 115 lbs. (-115) + (0) – (0) – (-100) = 15 lbs. (negative emission)

(-115) + (0) – (0) – (-115) = 0 lbs. emission 2017 CB 1 decommissioned – How is this reported? With nameplate adjustment in 2016 Status 1/1/2017 Status 12/31/2017 CB1 115 lbs. Decommissioned Cylinder A 100 lbs. 100 lbs. Cylinder B Empty 115 lbs. (-115) + (0) – (0) – (-115) = 0 lbs. emission

How much SF6 did Sample Utility release in this time period? Comparing 1/1/2016 with 12/31/2017 Status 1/1/2016 Status 12/31/2017 CB1 100 lbs. CB1 decommissioned Cylinder A Empty Cylinder A 100 lbs. Cylinder B 115 lbs. Cylinder B 115 lbs. TOTAL 215 lbs. 215 lbs. How much SF6 did Sample Utility release in this time period? Following EPA’s reporting rule, Sample Utility shows a 15 lbs. emission for 2016, and 0 emission for 2017 Following CARB’s reporting rule, Sample Utility shows either a 15 lbs. emission for 2016 and a negative 15 lbs. emission for 2017, or a 0 emission for both years.

Necessary Changes to Minimize SF6 Emission Tracking Problems Tracking should be in real-time Will immediately alert personnel of potential problem (Incomplete recovery, nameplate issues, weighing issues) Improve work instructions Field personnel needs detailed specs as opposed to “Remove SF6 from GIE” A) Check and record temperature/density Will immediately identify possible discrepancy with nameplate info B) Stop recovery only after achieving blank-off pressure < 10 Torr C) Record blank-off pressure Will eliminate recovery emission and can be used to document nameplate inaccuracy Utilize highest accuracy weighing tools Use mass flow scales whenever possible Will eliminate discrepancies due to incorrect cylinder TW and residual SF6 remaining in recovery equipment and hoses

EPA SF6 Partnership Emission Rates

Subpart DD – Use of Electric Power Equipment 2011 2012 2013 Total # Reporters 110 114 108 Total Emissions* (MMTCO2) 3.5 4.0 3.3 Electric power systems with a total nameplate capacity of SF6 and PFCs containing equipment exceeding 17,820 lbs. Several systems with emissions over 200,000 MTCO2e Note: EPA data now reported using IPCC 4th Assessment GWP100 values - SF6 = 22,900

EPA SF6 Partnership Emission Rates EPA Partnership Utilities showing much lower emission rates Domestic usage has increased Atmospheric concentration of SF6 raising faster than expected Possible Emission Sources SF6 used for other applications Actual emissions by partners higher than reported Anecdotal Evidence suggests a lot of GIE being slightly over-filled Entities not reporting have higher emissions as estimated (miles of transmission lines) Too much virgin SF6 being purchased vs using recycled SF6

Virgin vs Recycled SF6 Why use recycled SF6? Recycled SF6 works just as well – even in new GIE – provided it meets purity standards established by CIGRE / IEEE / IEC What if OEM requires virgin SF6? Ask to talk to somebody else Provide comparative test result (GIE filled with virgin vs recycled SF6) Substantially lower cost Lowers our domestic carbon footprint Keeps government / regulators from demanding use of alternative gases

Alternative Gases

Alternative Dielectric Gases vs SF6 Alternative Gases currently tested cannot be used in existing GIE Do not have the ideal characteristics (Arc quenching AND dielectric) compared to SF6 More difficult to handle / on-site mixing requirements Multiple alternatives are currently being tested by different OEM’s Cost Require different handling and test equipment Much lower GWP Only benefit of alternative gases currently tested is the much lower GWP

Alternative Gases vs SF6 – Future Developments Environmental Regulators extremely interested in alternatives to SF6 Collectively (GIE OEM’s / Recyclers / Service Providers / Users) our industry needs to do a better job eliminating SF6 emissions – including Phantom Emissions Ensures that a potential switch to alternative gases will be decided by the market (i.e. SF6 Users / Electric Utilities) based on a number of factors (Economics / Ease of handling AND environmental impact) and NOT simply due to environmental regulations

Lessons Learned Despite most low hanging (GIE with known large leaks replaced or repaired) fruit picked, most utilities still report emission rates > 1% Anecdotal evidence suggests > 90% of GIE currently installed contains a different amount of SF6 (deviation > 1%) compared to nameplate Accidental SF6 emissions due to personnel not understanding residual pressure / blank-off pressure still extremely common Entities routinely report “Phantom Emissions” due to tracking problems Entities reporting should immediately establish TMM (True Mass Monitoring) for all their GIE True Mass Monitoring = Measuring and recording the SF6 weight (Mass) for every gas event with 1% or better accuracy for the complete life-cycle of the GIE.

Thank you for your attention! Lukas Rothlisberger CEO DILO Company, Inc. www.dilo.com www.dilodirect.com