Introduction to analytics in PI System

Transcription

1 Introduction to analytics in PI System Landry Khounlavong Product Marketing Manager

2 Intro to PI System Track Landscape Crash Course Asset Monitoring Analytics PI System & the greater technology landscape Get the most out of PI World Intro to condition based maintenance with the PI System Intro to analytics in PI System

3 Analytics give many choices 3

4 Planning to start with big data? 5 gotchas! 1. Gaps in data collection 2. Poor data quality 3. Low # of modes 4. Unknown relationships 5. Unable to influence decisions 4

5 Where do I start? or Where are the quick wins? 5

6 ROI Small projects propel success for a large project Must-have project characteristics Subject matter expert included Sufficient business value Related to larger initiative 3y time 6

7 Building blocks of an analytics journey Collect measurements in one place Calculate physics-based decisions Contextualize system view of data Model enterprise-level comparisons Combine profitability analysis Apply algorithm augmented decisions 7

8 80-5.Net Volume GE01_A_DT GE04_OSAsset1.Problem s QI-122 FI-151 AlarmTest.Input.Float32.1 DC.SJ.PUE TI-102 DC.Zero DY-108 DC.SJ.C1.Z3.R3.PDU1.PF GE01_A_DT _Wind Speed 02F102.1HRAVG BGT001 AQUA2-SI005.PV GE02_Energy PI-111 facility_output 0_ENG_MODE_STS FI-151 AQUA2-TI-201.PV BGE003 FI T100 AF_FLOW3 02:T103 AT401 ACEDemo.Unit1.Output FI T103.Q TI-178 B352_W778 0_CMP_SVLV_PCT DC.Z1R 02F104 CD:F161 0_CMP_HDR 94:BW.R TI-101 F723_E Temp Compressor-439.Feed Rate DC.CH.DCE FIC- Boiler-334.Feed Rate 02F102.1HR AVG BGT001 PI-111 facility_output FIC-144 GE04_Status 02F100 fasttag 02:F101.C AQUA2-SI005.PV GE02_Energy 0_ENG_MODE_STS FI-151 TI T103.Q DC.Z1R asset1_output 0_CMP_HDR_SUC_PRS Active Meters FI-111 GE01_A_DT aso AF_NOISE DM-05:BW.R PI-115 DC.SJ.PUE fic1001.c GE02_OT FI-101 bf5e1d1d- 39c9-4b5b-b3d3- FI-101 Volume trigger AC09.Power GE05_Energy C1:14AT5 AC03.Air Flow FeedBin.Cmt Boiler Cold Reheat Pressure DC.SJ.SiteRealTim Anacortes Refinery.Alkylation.Asset B737_FG117 DC.TimeLoad eitload.pr FT9001 Problems B210_FG005.KPIExcursion D-110.Tank Pressure.PV AC04.Air FI-101 bf5e1d1d-39c9- GE04_DT QI-121 GE03_V_WIN 4-36.NetVolume Flow 02T100 03LBB02CT b5b-b3d3-c2ce05fa3a26 DM-05:BW.R DC.Rk07R AT401 DC.NY.Actual.PWR.day.Tot 0_CLR_FINAL_OUT_B_TDC.Srv06R GE04_Energy AlarmTest.Input.Float32.10 MP F506_E _Clear Sky AC Power TI-121 FT9001 Global GE01_DT _Wind Speed QI-109 Horiz GE01_DT FAC.OAK.Power-Kh-Val.PV GE01_DT Cooling Fan-711.Feed Rate DC.Srv01R Boiler-209.Fuel Gas AQUA2-SI005.PV GE02_Energy Boiler- AlarmTest.Input.Float32.1 Flow DC.Srv01R 0_ENG_MODE_STS 125.Fuel Gas 94:GRDIDX.Tr 1-13.Net Volume B045_FG :210 GE01_DT FI-121 AF_FLOW3 0_ENG_MODE_STS GE03_Q DY-131:166 GE01_TD AT401 02T103.Q DC.Z1R 0_CM P_ HDR _SU C _ PRS FI-101 bf5e1d1d-39c9-4b5b-b3d3-c2ce05f Net BGE003 FI T100 AF_FLOW3 02:T103 ACEDemo.Unit1.Output TI-178 B352_W778 0_CMP_SVLV_PCT 02F104 CD:F161 c2ce05fa3a26 AT401 DM-05:BW.R 0_CLR_FINAL_OUT_B _TMP F506_E _Clear Sky Global Horiz GE01_DT DY :F101.C DY-131 Volume FI-151 DM-05:BW.R 0_ENG_AUX_STS DC.SJ.C1.Z1.R1.Rk06.S2. O03.PWR QI-111 TIC-181 FinalProductBin.On 94:GRDIDX.ProdID Boiler-209.Fuel Gas Flow AF_NOISE D-110.Tank Pressure.P V Boiler Feed Pump #1 1-8.Net Volume 03LBA32CT0 AT401 % CO2 GE05_ES T 1-16.Net Volume CB1992_MS 0_CMP_FLOW_TOTAL 01-2 DC.SJ.ITLoad.P WR GE02_Energy FT9001 TI-145 fic1001.c FR5001 fasttag FR2001 FT9001 FeedBin.Cmt TI _Wind DC.Zone1.Number GE04_OS Speed FT9001 DailyTrigger FrqPrbCost_ER 45-2.Net Volume FT9001 FR5001 DC.C2Z1.Pwr.Rippl FT9001 AF_NOISE GE01_CON e GE01_A_DT AlarmTest.Input. Float32.1 DC.SJ.SiteRealTim eitload.pr FT9001 GE01_DT _Wind Speed QI-109 DM-05:BW.R GE01_DT Cooling Fan DC.SJ.C1.Z1.R1.R k06.s2.o03.pwr GE01_A_DT QI-111 FinalProductBin.On fic1001.c 94:GRDIDX.Pr 02:F101.C odid Boiler- FR2001 TIC Fuel Gas Aso AT401 Flow DC.Srv01R Map raw data to the physical world Weather Conditions Relative Humidity: 34% Current Temp: 85 F High: 92 Low: 57 F Wind: 8 mph/n Crude Desalter Operating Pressure: 110 psi Charge Rate: 14 gph Mix Valve Pressure: 8 psi Water Rate: 8% Crude Furnace Draft Pressure: -0.5 WC Stack Temp: 316 F Oxygen: 2.5% Outlet Temp: 840 F Cold Oil Velocity: 6 ft/sec 8

9 Make raw data decision-ready Total Power Power Status Power Status Power Status Unit % Uptime Overall Efficiency Both variables are potential features!! 9

10 Continuous streams have discrete events Downtime Excursion Batch Event Attribute Value Name Ex Start 15-Dec :35:02 End 15-Dec :47:26 What happened during the excursion event? Asset Excursion Type Fuel Gas Flow.Avg mykpi.max Boiler-352 High Violation k sft3/h bbl/d 10

11 Simplify Data Analysis Compare Assets Compare Events Discover Event Interrelationships Name Max EF EF EF EF Running XYZ Downtime 11

12 Customer realizes value from Event Frames 90% time reduction on intelligence gathering $350k investment saving on offshore system solution 12

13 What you can analyze depends on the granularity of your data 13

14 Time-series data is not naturally aligned remember this when you want to combine this data with business data! 14

15 PI Integrators simplify joining time-series data with business data 1. Create View 4. Publish No code required! 2. Select Data 3. Augment and Filter 15

16 Customer realizes value from PI Integrators New assets included and disposed of assets removed, automatically Allowing Data & Analytics team to do more analytics and less data preparation Over 200,000 calculations and 1,521 hard-coded cells

17 Building blocks of an analytics journey Collect measurements in one place Calculate physics-based decisions Contextualize system view of data Model enterprise-level comparisons Combine profitability analysis Apply algorithm augmented decisions 17

18 Data Sources Collect Store & Enhance Deliver Notifications PI Integrators SAP HANA Business Analytics Event Frames PI System Access Line of Business Systems PI Interfaces & PI Connectors Asset Analytics Asset Framework PI Cloud Services Visualisation Tools PI Cloud Connect PI Vision Data Archive PI DataLink PI ProcessBook PI Server PI WebParts 18

19 Takeaways Start small to de-risk analytics projects PI System offers many ways to accelerates analytics projects Data Scientist Investigate how PI Integrators can save you time preparing data Engineer/Operator See how PI Vision can give a full view of your processes and systems General Manager Speak to our customers about the value they gain from PI System and checkout industry tracks on Wednesday 19

20 Some related presentations tomorrow 9:00a Derek Shuen from Goldcorp: Real Time Data Critical to Mining results after only 6 months of implementation have surpassed expectations 2:30p Rick Smith from International Paper: Feeding the Machine Learning Monster 9:45a David Bartolo from AGL Energy: AGL Energy s Real-Time Data Journey Continues developing its own solutions for thermodynamic performance optimisation 3:15p Oscar Smith, Brian Morel from EQT: The Evolution of PI System at EQT 20

21 Copyright 2018 OSIsoft, LLC