Reexamination of PQ Standards Mark Stephens EPRI NYSDPS

Re-examination of PQ Standards, Mark Stephens – EPRI NYSDPS Power Quality Technical Conference December 12, 2013

Original EPRI R&D Project (1996 -2000) • In 1996 EPRI Began “Task 24: Power Quality in the Semiconductor Industry” • Original Participants: Original Task 24 Testing NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 2

Original R&D Project Timeline Task 24 Research and Testing June 97 33 Tools Tested By EPRI PEAC Nov 1998 April 99 Workshop 4 Tempe Mid 1996 April 97 Workshop 1 Dallas June 98 Workshop 3 Dallas Feb 1999 T 24 Final Report Sept 97 Workshop 2 Tempe Sept 99 SEMI F 47 Published SEMI Task Force Standards Work Oct 97 Approval for Task Force By SEMI June 99 Test Methodology SEMI F 42 Published NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 3 June 2000 SEMI F 42 And SEMI F 47 Revisions

SEMI F 47 -0706 Test Levels Relevant for Single and Two Phase Voltage Sags Only – does not address Three-Phase Events. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 4

The Question Remains… • Previous EPRI studies have shown that SEMI F 47 has made an impact on reducing fab downtime. • However, there are still PQ events above the SEMI F 47 requirements that are causing downtime in semiconductor fabs. • WHY? NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 5

New EPRI Research Project to Re-Examine SEMI F 47 • Understand electric service quality requirements. • Identify post SEMI F 47 mechanisms for continued semiconductor processing equipment downtime through industry collaboration. • Determine key strategies and methodologies to further harden semiconductor equipment to power quality events. • Inform development of revised standards, equipment designs, power system requirements, and PQ testing methodologies for current and next generation semiconductor tools. • Improve productivity through minimizing production downtime. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 6

Example Utility PQ Data at Semiconductor Plant • 72 Sag Events • 38 Diff. Days • 16 Events affected production, – Only 1 clearly below SEMI F 47! 2 Nominal Voltage (x 100%) 1, 5 1 SEMI F 47 Voltage Sag 0, 5 0 0, 01 0, 1 1 10 1000 Cycles NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 7 10000

Objectives • The key objectives of this project are: – Determine the characteristics of power quality events still causing semiconductor plant process downtime – Take a new look at the sensitivities in the process equipment – Determine any required adjustments to equipment design or standards to further reduce voltage sag induced losses by the semiconductor industry. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 8

Benefits (1) • Utility Collaborators will gain information about the characterization of a broad spectrum of equipment and fabrication tools, which will enable them to evaluate industry expectations in the context of the service capabilities of electric power systems. • Utility Host-Collaborators will work closely with partnering Semiconductor Site-Hosts to gain specific understanding of the needs and expectations of their customer. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 9

Benefits (2) • Semiconductor Site-Hosts will gain an in-depth understanding of why their equipment is still sensitive to specific power quality events and learn specific measures that can be taken to further harden their current and future process systems. • Equipment and Tool Manufacturers will gain a better understanding of why their equipment may still be vulnerable to power quality issues and methods for improving their power quality robustness. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 10

Stakeholder Collaboration • Collaboration between stakeholders will be required. Electric Utilities* • Project approach is to carefully examine the cause of PQ related downtime Tool Suppliers/ OEMs Semi Fabs Site-Hosts System Compatibility • The outcome of basic R&D will inform processes and standards to further improve system compatibility SEMI & Consultants SEMATECH EPRI *Utilities can participate as Collaborators or Host-Collaborators. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 11

Project Approach and Summary (1) • This project will methodically investigate the potential factors that continue to cause semiconductor plant downtime in a post-SEMI F 47 environment. • These factors may include: – (1) three-phase voltage sag events – (2) repeated voltage sag events due to recloser operations – (3) events outside SEMI F 47 scope of magnitude, duration, phase-shift or point-on-wave – (4) voltage sag testing methods utilized during certification, tool software, design, or configuration differences – (5) electrical system differences between the certification environment and the actual fab environment. – (6) Equipment not certified to SEMI F 47 NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 12

Project Approach and Summary (2) • Working through EPRI, Utility Host-Collaborators will partner with a specific Semiconductor Site-Hosts to investigate SEMI F 47 requirements against specific site electrical environments and identify potential gaps against tool shutdown data. • SEMI F 47 certification documents will be reviewed for the sensitive tool sets and characterization of this equipment will be accomplished through additional analysis and testing. • Specific design and mitigation strategies including costbenefit will be determined and documented. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 13

Project Approach and Summary (3) • The general, non-confidential findings from these partnerships will help inform the overall collaborative effort through workshops, white papers, and the final report. • Utility Collaborators will participate in the overall effort through participation in project web-meeting and workshops, inclusion in engagement with SEMI and SEMATECH, and involvement in any proposed standards modifications. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 14

Research Project Tasks Task 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Activity Plan Document the Electrical System Environment of the Semiconductor fab. Identify gaps and opportunities for creating a successor to the current SEMI F 47 standard that addresses the electrical environment for known equipment shutdowns. Review SEMI F 47 certification documents for sensitive tool sets. Determine the susceptibility of the current tool sets. Indentify effective design and mitigation strategies. Perform Cost-Benefit Analysis Provide Technical Materials for Host Site (Reporting, Findings, Recommendations) Write White Paper on Initial Findings. Collaborative Workshop No. 1 Propose modification to existing SEMI F 47 performance criteria as necessary as identified by the previous tasks. Develop and Validate Test Methods to Confirm that Tools Meet proposed Power Quality and Ride-Through Requirements. Suggest Revisions to SEMI 49 “Review Guide for Semiconductor Factory Voltage Sag Immunity” Suggest Revisions to SEMI F 50 “Performance Guidelines for Power Supplied to Semiconductor Factories by utilities” Perform a gap analysis to determine the potential reduction in power quality related costs. Collaborative Workshop No. 2 Technical Transfer Documentation and Reporting NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 15 Specific R&D Activities at each Host-Site Collaborative Activities

Technical Deep Dive Slides…

Potential Causes for Downtime (1) Three-Phase voltage sag events. – SEMI F 47 does not require the OEM to test their equipment against three phase voltage sags events – CIGRE study (C 4. 110) has shown that three-phase (Type III) events can make up to 20 percent of the events that occur Sites Voltage range Number of Single-phase sites (Type I) MV & HV Sites 6 k. V – 230 k. V 969 Sites <1 KV 206 NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 17 Two-phase (Type II) Three-phase (Type III) 27% 53% 20% 64% 25% 11%

Potential Causes for Downtime (2) Repeat voltage sag events – Reclosing to restore operations. – SEMI F 47 does not require testing for back to back events – Sag events can be multiple events in close intervals based on fault clearing NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 18

Potential Causes for Downtime (3) Event magnitude, duration, phase-shift, and point-on-wave are outside of SEMI F 47 scope – SEMI F 47 calls for testing at 0 degree point-on-wave. • Actual events can occur at various points-on-wave – Possible Phase-shifts are based on set of test vectors. • Actual events can be different – Minimum magnitudes and durations are: • 50% Vnom, 200 ms • 70% Vnom, 500 ms • 80% Vnom, 1 Second NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 19

Potential Causes for Downtime (4) Voltage sag testing methods utilized during certification. – There are three types of phase-to-phase voltage sag test vectors that are allowed for SEMI F 47 testing. – These three types of vectors result in various voltage magnitudes and phase shifts. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 20

Potential Causes for Downtime (5) Tool Software, design, or configuration differences – The tool configuration and software during SEMI F 47 testing may have been slightly different than the equipment that is installed at the semiconductor fab’s facility. Graphic Ref: SEMI F 49 “Guide for Semiconductor Factory Voltage Sag Immunity”” NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 21

Potential Causes for Downtime (6) Electrical System Differences – During SEMI F 47 certification testing, the equipment is tested against the nominal voltage available at the site. – The available nominal and peak voltage at the point of connection at the semiconductor fab can change due to changes to the system configuration. The Lower the Pre-Sag Voltage, the Less Energy (and ride-through time) Is Available in the Components with Capacitors NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 22

Potential Causes for Downtime (7) Installed Equipment In the Fab is not Compliant to SEMI F 47 • Equipment in the factory may not actually be compliant to power quality standards such as SEMI F 47. • In procurement, tool capabilities and delivery schedule may trump testing for SEMI F 47 compliance REF: Impact of SEMI F 47 on Utilities and Their Customers, EPRI, Palo Alto, CA: 2004. 1002284. NYSDPS Power Quality Technical Conference 12/12/2013 © 2013 Electric Power Research Institute, Inc. All rights reserved. 23