Too Many Alarms: Where Do I Begin? Standards Certification Education & Training Publishing Conferences & Exhibits Speakers: Kevin Patel, PE, MBA Signature Automation, LLC 2013 ISA Water / Wastewater and Automatic Controls Symposium August 6-8, 2013 Orlando, Florida, USA
Presenter 10 years of experience in integrating and programming instrumentation and control (I&C) systems, SCADA systems including PLCs, RTUs, HMIs and networking primarily for water and wastewater treatment facilities. Bachelor of Science in Computer Engineering from Texas A&M University Master in Business Administration from University of Texas at Dallas Current member of the ISA101, ISA105, ISA106 and ISA18 2
Presentation Outline Introduction Alarming Problems Alarm Management Benchmarking Case Study Statistics Alarm Design Methods Getting Started Summary
Introduction Alarm Systems Far More Powerful Today No Guidance is Given During Development Not Enough Information for Alarm Response Reliance on Unmanaged Alarm Systems Causing Safety Problems Development of ISA18.2, Technical Reports http://www.pegaenerji.com
Alarming Problems Nuisance Alarms Most common problem Alarms that generate in excess Acknowledged without consideration of validity mind-trix.com
Alarming Problems Stale Alarms Remain in alarm for long periods of time Acknowledged, then ignored No action taken during maintenance
Alarming Problems Flood Alarms Surge of many alarms at once Prevents appropriate response Delays reaction time
Alarming Problems Unclear Alarms Generate with unclear information for action Delays response time Eventually may become stale meuniversitylive.com
Alarm Management Benchmarking Alarm Performance Metrics Average annunciated rate per operator Peak annunciated per operator Alarm floods Frequently occurring alarms Chattering/fleeting alarms Stale Alarms Priority Distribution
Case Study Statistics Month 1 Initial Analysis of Site 4 days of data 193 Tags 209,083 Alarms ~363 Alarms Every 10 minutes
Case Study Statistics Month 1 Rank Top 20 Alarms Alarm Instances % of Total Day 1 Day 2 Day 3 Day 4 1 D110_002_220PLC02YA 37716 18.04% 9541 9355 9374 9446 2 D110_037_960PLC37YA 37694 18.03% 9548 9334 9338 9474 3 H110_002_220VRTEX04FIUR 26771 12.80% 6719 6658 6753 6641 4 V110_002_220VRTEX01FIMDI 23500 11.24% 5871 5860 5874 5895 5 H110_002_220VRTEX01FIUR 23427 11.20% 5855 5837 5858 5877 6 H110_002_220INFPS02FIUR 16706 7.99% 4215 4215 4046 4230 7 V110_002_220VRTEX03FIMDI 10351 4.95% 5908 4303 64 76 8 H110_002_220VRTEX03FIUR 10315 4.93% 5887 4287 63 78 9 H110_002_220PSHSF01AIUR 7396 3.54% 1568 2446 2189 1193 10 A-46-208 2660 1.27% 34 1151 804 671 11 V110_038_JT-4-90-1MDI 1928 0.92% 0 0 0 1928 12 FIT310_001_05_FI_UR 1574 0.75% 1574 0 0 0 13 H110_002_220PSLDG02FIUR 1315 0.63% 227 276 338 474 14 VLV240_2_63_ZI_OR 1179 0.56% 248 166 646 119 15 VLV240_2_63_ZI_BAD 1178 0.56% 247 165 646 120 16 D_110_010964BIOSB06YA 1152 0.55% 288 288 288 288 17 PLC952_073_201_MOD_FAIL 730 0.35% 730 0 0 0 18 V110_048_461TMZN3A2AIMDI 647 0.31% 262 123 194 68 19 MS10_0M_COMMFAIL 593 0.28% 126 175 145 147 20 PLC01-1SLOT7-FAULT 506 0.24% 253 125 64 64 Total 207338 99.17% 59101 54764 46684 46789
Case Study Statistics Month 1
Alarm Management Lifecycle Detailed Design Stage Basic Alarm Design HMI Design Philosophy Design of Advanced Alarming Techniques Identification Rationalization Detailed Design Management of Change Audit Implementation Operation Monitoring & Assessment Maintenance
Alarm Management Detailed Design Basic Alarm Design Alarm setpoints Alarm deadbands Alarm on/off delays HMI Design Where are alarm changes made HMI display requirements Alarm visual and audible indications Alarm logging Advanced Alarm Design Designed suppression Alarm redirection
Alarm Management Basic Alarm Design 60 58 56 54 52 50 48 46 44 42 40 Signal Type Deadband (Percent of Operating Range) Delay Time (On or Off) Flow Rate ~5% ~15 Seconds Level ~5% ~60 seconds Pressure ~2% ~15 seconds Temperature ~1% ~60 seconds
Case Study Statistics Month 1 Top 2 points were PLC heartbeats (Disabled) 6 points were under-range alarms chattering (On Delay Timer was Added) 4 points were duplicate alarms (Disabled) Others were left as they seemed to be anomalies. Will re-check on next analysis to see if they continue to be a problem. 12 of the top 20 alarms were addressed.
Case Study Statistics Month 2 Rank Top 20 Alarms Alarm Instances % of Total Day 1 Day 2 Day 3 Day 4 Day 5 1 M-4-55-1-AL 7184 14.14% 1424 1440 1440 1440 1440 2 M-4-55-2-AL 7184 14.14% 1424 1440 1440 1440 1440 3 M-4-55-3-AL 7184 14.14% 1424 1440 1440 1440 1440 4 M-4-55-4-AL 7184 14.14% 1424 1440 1440 1440 1440 5 M-4-55-5-AL 7184 14.14% 1424 1440 1440 1440 1440 6 M-4-55-6-AL 7184 14.14% 1424 1440 1440 1440 1440 7 FIT310_001_05_FI_UR 1926 3.79% 698 803 8 0 417 8 VLV240_2_63_ZI_BAD 874 1.72% 275 76 124 126 273 9 VLV240_2_63_ZI_OR 874 1.72% 275 76 124 126 273 10 MS10_0M_COMMFAIL 669 1.32% 158 182 140 129 60 11 PMP7334_256_01_MM 302 0.59% 90 43 32 62 75 12 PMP7334_256_02_MM 301 0.59% 89 43 32 61 76 13 V110_048_461TMZN3A2AIMDI 297 0.58% 0 0 0 130 167 14 LSLL-4-56-2 281 0.55% 0 0 50 185 46 15 PLC12-IPF 278 0.55% 18 103 55 70 32 16 MS11-7E_COMM_01 227 0.45% 32 41 34 44 76 17 PLC952_073_201_MOD_FAIL 215 0.42% 0 0 2 0 213 18 M-30-410-AL 200 0.39% 0 200 0 0 0 19 H110_002_220PSHSF01AIUR 118 0.23% 5 6 39 25 43 20 FP-6-252-1-8 108 0.21% 10 32 26 17 23 Total 49774 97.96% 10194 10245 9306 9615 10414
Case Study Statistics Month 2
Case Study Conclusions Find the problem alarms and tackle them individually immediately. Small amount of changes can make a HUGE difference. Most systems have built-in capabilities to provide alarm suppression and alarm time delay. Take advantage of these features. Provide continuous monitoring and assessment to fine tune your alarm system.
Getting Started Time to get started is NOW! Execute alarm management on new or existing alarm systems For existing systems, it is best to start with a benchmark More than a one day venture
Thank You! Contact: Kevin Patel, P.E. 469-248-6840 knpatel@sig-auto.com