ABB PA Life Cycle Services. ServicePort Access to Efficiency Performance Services

Transcription

1 ABB PA Life Cycle Services ServicePort Access to Efficiency Performance Services

2 Lifecycle Services Development Charter (agenda) Develop services and tools that address customer needs Identify Value Value Packaged Services Packaged Services Increase Efficiency Increase Productivity Leverage Knowledge Built on Expertise Sales Productivity Responsive Service Development Standard Efficient Service Tools Maximize Systems Optimize Processes ABB Group September 26, 2012 Slide 2

3 Lifecycle Services Development Charter (agenda) Develop services and tools that address customer needs Identify Value Identify Value Packaged Services Packaged Services Increase Efficiency Increase Productivity Leverage Knowledge Built on Expertise Sales Productivity Responsive Service Development Standard Efficient Service Tools Maximize Systems Optimize Processes ABB Group September 26, 2012 Slide 3

4 Fingerprint Services follow a proven process Diagnose (Fingerprints) Measure performance gap Forecast Return on Investment Deliver action plan Implement (HandsOn) Fix performance gap Define monitor plan Sustain (Scan/Track) Manage performance gap Schedule maintenance Define condition triggers Maintain to conditions 1 Proactive and collaborative Service 2 Diagnose (Fingerprints) 3 Implement (HandsOn) Sustain (Scan/Track) Increased performance ABB Group September 26, 2012 Slide 4

5 Scalable Advanced Services delivery

6 Scalable Service Delivery Technical Tools Troubleshooting/Implementation Fingerprint Stand alone 0 Best Option 100 ServicePort Channels - Continuous monitoring 1. Service Solution Loop Performance, Loop Tuning, Batch Performance 2. Service delivery Fingerprint, Remote analysis, Continuous monitoring 3. Fingerprint is a light application with no hardware using technical tools 4. Technical Tool usage for Troubleshooting and implementation 5. ServicePort provides continuous monitoring, requires ServicePort Channels

7 ABB Fingerprint Services: packaged deliverables for control systems and customer processes ABB Group September 26, 2012 Slide 7 Common Fingerprint: Boiler, Paper Machine, OMC, Alarm/Security QCS (VPA, Shift Day, Standardize) Loop Performance Transition Analysis Batch Analysis Tuning

8 Loop Performance Services Data Collection DL300 Loop Performance Fingerprint Loop Analyzer Signal Analyzer Loop Tuning LoopTune Plant Wide Disturbance Analysis Loop Analyzer Batch Analysis Sequence Analyzer High Speed data analysis HSD500/AGP400 Loop Performance Monitoring Solution ServicePort Loop Performance Channel Diagnose Loop Performance Management Implement Continuous Improvement at Maximum efficiency Problem ABB Group September 26, 2012 Slide 8

9 LoopScan makes process troubleshooting efficient Few units Easy to identify issues Many Hard units to find THE problem. Impractical to find with simple tools Important to avoiding false positives. ABB Group September 26, 2012 Slide 9 Finding a needle in a hay stack

10 Loop performance degradation Control loops will degrade in performance Half life of process controllers Given: 100 PID loops all tuned at once. Then: within 6 months, 50 of these loops will degrade in performance. 100 Simple PID utilization 15% 30% Manual Operation Output Out of range Increasing Variability Improving process Months 35% 25% PID Controllers are designed to: Regulate the process Reduce product instability Improve operations However, ABB is finding that: PID loops are not being maintained PID loops have degraded PID loops are standing in the way of production and performance.

11 Loop Performance Services manage performance gap Performance potential Application Process ABB Group September 26, 2012 Slide 11 Diagnose Implement Sustain Time

12 Loop Performance Services manage performance gap 1) Fingerprint Diagnose Gap Goal Preset Present Performance potential Application Process 1 ABB Group September 26, 2012 Slide 12 Diagnose Implement Sustain Time

13 Loop Performance Services manage performance gap 1) Fingerprint 2) Implementation Diagnose Gap Implement Goal Preset Present Performance potential Application Process 2 1 ABB Group September 26, 2012 Slide 13 Diagnose Implement Sustain Time

14 Loop Performance Services manage performance gap 1) Fingerprint 2) Implementation 3) Scan Diagnose Gap Implement Sustain Goal Preset Present Periodic Evaluation Periods Performance potential Application Process 2 3 Manage performance gap 1 ABB Group September 26, 2012 Slide 14 Diagnose Implement Sustain Time

15 Loop Performance Services manage performance gap Performance potential Application Process 1) Fingerprint 2) Implementation 3) Scan 4) Track 1 Diagnose Goal Gap Preset Present 2 Implement Sustain Periodic Evaluation Periods 4 Manage performance gap Alert ABB Group September 26, 2012 Slide 15 Diagnose Implement Sustain Time

16 ServicePort Performance Services reactive to proactive Event-driven action Reactive On Site On Site Scheduling Collection Analysis Resolution Proactive On Site Remote On Site Collection Analysis Resolution Fingerprint On-Site On Site Data Pool Collection Remote Analysis On Site Resolution Periodic Remote Enabled Service Modules On Site Data Pool KPI Trending On Site Collection Analysis Resolution Short Lead time Scan Services On Site Data Pool Condition Monitoring Remote Collection Analysis Resolution No Lead time Track Services Event Trigger

17 ABB Fingerprint finds gaps, develops customer ROI Action Report Report Gap Analysis ROI Forecast Action Plan ROI Benchmark Findings Plan Interpret Analyze View Performance Evaluation Standard Methodology Analysis Expertise Performance Visualization People Process Tools Get Problem ABB Group September 26, 2012 Slide 17 Data Collection/Testing 12 to 24 hours at 5-second data Controller parameters Customer interview: process area and loop criticality definitions. OPC Collection Tools

18 Goal: Continuous Improvement Action Plan Tuning Process Loops Prioritize and Categorize Actuator Implement Signal Stand Alone Tools Signal Analyzer LoopTune Loop Analyzer Continuous Tools ServicePort Logic

19 ABB Group September 26, 2012 Slide 19 Optimization Services: Performance Migration Maximum Off Specification Goal Optimal: Mechanical Constraints Sustain: Improvement Projects Implementation: Corrective Action Diagnose: Current Reduction in Variability = Less Raw Material Usage Improved energy consumption Wider Operating Window Increases in production, quality, and product purity Faster troubleshooting time Advanced Automation Solutions

20 QCS Performance Fingerprint Production Capability Thruput Reel Speed Lost Time Reel Variability Total Distribution Trends Settings Control Utilization Auto/Manual Grade change Fingerprint Inputs: Shift/Day Report VPA reel Report Standardize Report Check Sample Reports Tuning/Setup Parameters Sensor Stability Standarize Check Sample Calibrate Sample ABB Analysis Tools VPA200 CU100 SA100 MD400 Excel ReportPro ABB Group September 26, 2012 Slide 20 Non Invasive One week of scope

21 QCS Performance Fingerprint QCS Performance Fingerprint Production Capability Reel Statistics Control Utilization Sensor Stability Tuning/Setup Cluster Well defined decision tree for optimal solution definition Non Invasive ABB Group September 26, 2012 Slide 21 Packaged Service Modules Implementation Solutions

22 QCS Performance Fingerprint Production: Variability and Lost Time Range 100 t/d If variability can be reduced, Then the reduction translates into a production increase. Lost time contribution: Grade Change, Sheet Break, start up, quality problems. Average 3 hr/day ABB Group September 26, 2012 Slide 22 Control Utilization tool reading Shift/Day Info

23 System Reel Report 2-Sigma s? Basis Wt 1 LBS Quality Analysis ABB Inc. Target Actual Efficiency RES MDL CD TOT Moisture 1 PCT Target Actual Efficiency RES MDL CD TOT

24 2 Sigma Percent Percent QCS Performance Fingerprint: Variability Distribution Sigma as % of Process [Grammatura] Master_85_90 Master_100_120 Master_140_180 Master_200 Master_230_240 Master_280 Master_300_320 Master_350_360 Master_400 Grammatura Sigma as % of Process [Umidita Pope] Master_85_90 Master_100_120 Master_140_180 Master_200 Master_230_240 Master_280 Master_300_320 Master_350_360 Master_400 Umidità CD MDL MDS TOT Master_85_ Master_100_ Master_140_ Master_ Master_230_ Master_ Master_300_ Master_350_ Master_ CD MDL MDS T OT Master_85_ Master_100_ Master_140_ Master_ Master_230_ Master_ Master_300_ Master_350_ Master_ MDL represents the frequency band between 1 minute and 45 minutes. MDL is 3 to 5 times higher than expected in all measurements Sigma [Ceneri] M aster_85_90 M aster_100_120 M aster_140_180 M aster_200 M aster_230_240 M aster_280 M aster_300_320 M aster_350_360 M aster_400 A ll data in participating groups Ceneri CD MDL MDS TO T ABB Group September 26, 2012 Slide 24 M aster_85_ M aster_100_ M aster_140_ M aster_ M aster_230_ M aster_ M aster_300_ M aster_350_ M aster_ A ll data in participating groups

25 ABB Group September 26, 2012 Slide 25 QCS Performance Fingerprint: Control Utilization Machine 0 Controller Utilization By Month CassaAfflusso Ceneri Grammatura PastaalSecco Umidita1 Umidita2 VelocitaCoordinata Goal >95% 0 Month Oct 2010 Nov 2010 Dec 2010 Jan 2011 Feb 2011 Mar 2011 CassaAfflusso Ceneri Grammatura PastaalSecco Umidita Umidita VelocitaCoordinata

26 ABB Group September 26, 2012 Slide 26 QCS Performance Fingerprint: Tuning Number KPI s Setup/Tuning Validation Limit Check Cross Check Tuning Check Weight Reel Moisture Size Moisture Ash Configuration Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Name DW01_Control MT01_Control MT02_Control AC01_Control Description Reel Bone Dry Reel Moisture Size Moisture Reel Ash MV and SP Units (cu) lbs/ream % % % Output Units (cu) % % % % Number of Decimals Range Maximum (cu) Range Minimum (cu) Maximum Setpoint (cu) Minimum Setpoint (cu) Setpoint Ramp (customer/sec) Deadband (fraction) Normal Control Error (fraction) Control Error Sp Change (fraction) Control Model Filter (sec) s s m40s Filter for Control (1=MDMS, 2=Filt MDMS, 3=FWPM, 4=Advanced) Filter for Display (1=MDMS, 2=Filt MDMS, 3=FWPM, 4=Advanced) Extended Sheet Break Time (sec) 20s s s s Minimum Control Speed (cu) Extended Control Speed Time (sec) 20s s s s Download Auto Mode Restore FALSE 0 0 FALSE 0 0 FALSE 0 0 FALSE 0 0 Modeling Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Process Response (proc/actr) Process Response Gain (unitless) Process Response Final Calculated Process Time Constant (sec) 52s m00s m10s m00s Transport Distance (distance) Actuator DeadTime (sec) 1m00s m00s m10s m30s Closed-loop Time Constant (TAU) Max Actuator Change (actr/sec)) Max Actuator Change in AGC mode (actr/sec) Feed Forward (stock flow) Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Value High Limit Low Limit Enable TRUE 1 0 TRUE 1 1 FALSE 0 0 Process Response Process Resp Gain Process Time Constant 50s s s Transport Distance (Feet) Actuator Deadtime (Sec) 1m00s m00s m00s Filter Time 0s d0h0s s 20 0 Error Check (frac of SP) Advanced Controls Speed Auto grade MD Tuning Numbers Scan Level Modeling Feed forward Level 1 CD Tuning Numbers Tuning Modeling Additional Checks Control Model Filter Matches Process Time Constant Stock Feed Forward Time Constant matches Stock to Weight Time Constant Ratio of Feedforward time constant to primary process time constant is not greater than 2.5 Feed forward filter is only used when feed forward is tuned as a lead action

27 Standarize Problem Sensors Top Pressure Status Bot Pressure Sensor Zero Sensor Noise Sensor Max QCS Performance Fingerprint: Sensor Stability 1 Standardize Reports Ensure reliable measurements Check Sample Reports Ensures repeatable measurements Calibrate Sample Ensures exact process measurements 1 Standardize P roblem s Frame 1:Caliper - Caliper Standardize Sensor Max Sensor Noise Sensor Zero Bot Pressure Status Top Pressure Air Column Stdz Brightness Stdz Color Stdz Moisture Stdz Weight Stdz Ash Stdz Caliper Stdz Formation Stdz Opacity Stdz Thu 15 Thu22 Thu 1 Sun 8 Sun 15 Sun22 Sun 1 Sun 8 Sun 15 Sun22 Sun Jan 2009 ABB Group

28 1 Harmony Fingerprint uses Harmony Analyzer System Issues Loading GMI Suppression of Exception Reports Process Control Unit Overload Control way Problem Evidence TMax Exception Period Collision Error Counters Firmware Jumpers Control System Performance Fingerprint: Proprietary analysis tools Defined Scope Trained Engineers ABB Group Group September 26, 2012 Slide 28 26, 2012 Slide 28

29 Harmony Diagnostic Tools Overview Harmony Diagnostic Tools: semapi Harmony Direct (OPC Server) DL300 (Datalogger) HPA200 (Analyzer) Prerequisites: CIU (ICT03/13) on central loop Local bridges consist of IIT03 w/ B.4 firmware 1 Fingerprint can contain up to 3 loop, 50 PCUs per loop ABB Aug 13, 2012 Slide 29

30 Analyzing Diagnostic Data Harmony Performance Analyzer (HPA200) The Harmony Performance Analyzer is used to analyze the collected data from the DL300 and provide valuable information using graphs and tables that could help improve system performance. ABB Aug 13, 2012 Slide 30

31 Harmony Analyzer (HPA200) System Layout The Harmony Analyzer provides a system topology layout. Each scanned loop becomes available for viewing. All PCUs, CIUs, and Bridge nodes become visible in proper node order. ABB Aug 13, 2012 Slide 31

32 Harmony Analyzer (HPA200) Firmware Audit A firmware audit can be conducted using the HPA200. Each module s firmware revision is compared to the latest released revisi on. The HPA200 highlights modules with outdated firmware. Major revision changes are shown in red and minor changes are highlighted in yellow. ABB Aug 13, 2012 Slide 32

33 Harmony Control System Fingerprint Report The report summarizes the system and any issues found. Each issue is given a recommended plan of action for an engineer to tackle. Auto-generated reporting will eventually become an added feature to the Harmony Analyzer. ABB Aug 13, 2012 Slide 33

34 ServicePort Scan solution ABB Group September 26, 2012 Slide 34

35 Return On Investment Steady State -Variability Reduction Target Shift Quality Improvement Less Broke as related to variability Improved Lab test Sheet Break reduction Disturbance Rejection Transient Grade/Shade Change Sheet Break recovery time MD Sheet Break recovery time CD Reduced Start Up time OS -35

36 Sigma Profile COV as a percent of total Fingerprint: Paper Machine Evaluates: On control Performance of: Total Head, Thick Stock Flow, Thick Stock Consistency, Machine Chest Level Provides information on: Controllable Energy Stock Approach performance Tuning Quality Oscillation sources Provides information on: Controllable Energy Mechanical Pulsations of vibrations Benchmark of machine stability Stock Approach Stability Poor regulation Offset from setpoint Good regulation CD Product Variability Fingerprint: Weight and Moisture Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Actual Decade 6 History (VPA) Profiling Capability Actual vs. Calc. Profiles for 1 inch Zones Calculated weight moisture Machine Response Cyclic response Slow response Fast response Evaluates: Cyclic content of Weight and Moisture in: Cross Direction and Machine Direction from High Frequency up 500Hz down to Low Frequency of 5 hours. Provides information on: Start up time Grade Change recovery Disturbance reduction Sheet break recovery Responsiveness Evaluates: Automatic and Manual mode operation of: Weight, Moisture, and Total Head. Evaluates: On control Performance and model for: Weight, Moisture, and Caliper Zones Current Potential Provides information on: Will CD control improve the profile? Is current CD control optimized? ABB Inc.

37 OS -37 MO11 m1k BW11 m1k MW11 m100 BW11 m100 1MT1CTRL:MV /AI Channel/1ME1.CW11MV:VALUE Overnight Power Spectrum Comparison 17 Min 9.8 Min Weight Spectral Overlay Moisture Scanning 0.5Hz sample 5.7 Hrs duration Frequency (Cycle/Time) Sensor m100 Power Spectrum Comparison Weight Hz Moisture Single Point 100Hz sample 34 min duration Frequency (Cycle/Time) Sensor M1k Power Spectrum Comparison 60 Hz Weight Hz Moisture Single Point 1000Hz sample 3.4 min duration Frequency (Cycle/Time)

38 OS -38 Comparison PM Fingerprint: Product Variability Fingerprint Comparison - % Total COV Weight Moisture Provides a summary of the cyclic energy in the sheet over a corresponding frequency band. Includes Cross direction, Weight and moisture up to 100 hertz QCS DCS Mechanical CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 Decade 6 Weight Moisture

39 OS -39 Comparison Sigma P r o f i l e COV as a percent of total PM Fingerprint: Machine Response Provides an index for the regulatory capability of the control application. Fiber Line Stability Poor regulation Product Variability Fingerprint: Weight and Moisture weight 40 moisture 20 0 CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Decade 6 Machine Response Cyclic response Machine Response Index Total = Total Head Weight Moisture Offset from setpoint Good regulation Profiling Capability Ac tu a l v s. C a lc. Pr o file s fo r 1 in c h Z o n e s A c tu a l C a l c u l a te d Slow response Fast response Z o n e s Current Potential Total Head Weight Moisture Acceptable Performance

40 OS -40 Comparison O u t p u t Sigma P r o f i l e S P a n d M V COV as a percent of total PM Fingerprint: Machine Response Provides an index for the regulatory capability of the control application. Machine Response Index Total = Automatic Mode Closed Loop Test SP Change Total Head Weight Moisture Fiber Line Stability Poor regulation T o ta l H e a d : B o t to m T D H Offset from setpoint Good regulation Product Variability Fingerprint: Weight and Moisture weight moisture Manual Mode CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Decade 6 Profiling Capability Ac tu a l v s. C a lc. Pr o file s fo r 1 in c h Z o n e s A c tu a l C a l c u l a te d Machine Response Open Loop Test Cyclic response Output Change Slow response Fast response Z o n e s Current Potential Also, visible signal conditioning problems present Total Head Weight Moisture D a t a P o in ts, T s = 5 s e c Acceptable Performance

41 OS -41 Comparison Sigma P r o f i l e COV as a percent of total PM Fingerprint: Stock Approach Stability Indicates the stability of the fiber line. A high index suggests problems related to: process, control, mixing, etc. Fiber Line Stability Poor regulation Product Variability Fingerprint: Weight and Moisture weight 40 moisture 20 0 CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Decade 6 Machine Response Cyclic response Offset from setpoint Slow response Fiber Stability Index Total = Good regulation Profiling Capability Ac tu a l v s. C a lc. Pr o file s fo r 1 in c h Z o n e s A c tu a l C a l c u l a te d Fast response -2-3 Total Head Chest Level Stock Flow Consistency Z o n e s Current Potential Total Head Chest Level Stock Flow Consistency Index points to Machine chest level as the primary problem Acceptable Performance

42 OS -42 Comparison Consistency Sigma P r o f i l e Stock Flow Chest Level Total Head COV as a percent of total PM Fingerprint: Stock Approach Stability Indicates the stability of the fiber line. A high index suggests problems related to: process, control, mixing, etc Fiber Stability Trends P oints = 1681 Product Variability Fiber Line Stability Poor regulation Fingerprint: Weight and Moisture weight 40 moisture 20 0 CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Decade 6 Machine Response Cyclic response 60 Offset from setpoint Slow response 7 Fiber Stability Index Total = Total Head Chest Level Stock Flow Consistency 4.1 Good regulation A c tu a l Profiling Capability Ac tu a l v s. C a lc. Pr o file s fo r 1 in c h Z o n e s C a l c u l a te d Z o n e s Current Potential Fast response Total Head Chest Level Stock Flow Consistency raw data Raw Data Trend (~5 hours) shows that chest level impacts consistency and stock flow. Index points to Machine chest level as the primary problem Acceptable Performance

43 OS -43 Power (variance) Sigma P r o f i l e Profile COV as a percent of total PM Fingerprint: Profile Capability Actual vs. Calc. Profiles for 7.14 inch Zones Actual Calculated Product Variability Fingerprint: Weight and Moisture Sigma Original 0.54 Capability % Potential Fiber Line Stability Poor regulation Offset from setpoint Good regulation CD Decade 1 Decade 2 Decade 3 Decade 4 Decade 5 COV Weight = 5.17 COV Moisture = Decade 6 Profiling Capability Ac tu a l v s. C a lc. Pr o file s fo r 1 in c h Z o n e s A c tu a l C a l c u l a te d weight moisture Machine Response Cyclic response Slow response Fast response Zones Calc. Power Spectrum Point Width = Z o n e s Current Potential Actual Calculated Power Spectrum Current and Forecast Using process data, setup information, controller type, ABB can predict what the capability of a current or new set of actuators can do Frequency (Cycle/Distance)

44 OS -44 Fingerprint Block Diagram - Schedule Paper Machine Production Data Process Data Analysis VPA MR SS PC PV Power Point per area RO I Daily Activity List Exit Meeting: Benchmark, Initial findings and recommendations Final Analysis Final Reports: Executive, Technical, and Implementation Plan Week 1 Week 2

45 Moisture2 Moisture 1 Basis Weight High Frequency Results Amplitude Spectrum Comparison P oints = Single Point Frame data collected at 100 Hz and 1000 Hz for complete high frequency picture Provides insight into pulsation and vibration in both weight and moisture A B C D E F G A B C -1 Cycles/Point D E F G X1000 Hz Amplitude Frequency Period BW MT1 MT2 diameter length Potential Source Machine Chest Level? Top Felt Wire? ? After Section Dryer After Section Dryer fan pump? (Yankee speed = 3049fpm) ABB Inc.

46 OS -46 Bottom Chest Level PM1_AC450B_UDMISC_LT20-310_I_abbsvr_AccuRay_O bje Top Chest Level PM1_AC450B_UDMISC_LI20-331_I_abbsvr_AccuRay_O bje Machine Chest Level Control Logic Manual Operation Raw Data New Control Control Logic Added, Improvement > 90% Manual Operation New Control Sample Number, Ts = 5, Total Samples = 7304 Diagnose: Fiber Stability Index High due to poor Machine chest regulation. Implementation: Found that the poor Machine chest regulation was a control logic problem. The logic was corrected, the loops were re-tuned, and the variability dropped by over 90% Sustain: Teach operators how to use new control logic, changed the SOP to make sure control stayed in range.

47 OS -47 Weight Before vs. After Tuning Changes Before Improvement 45% After Conditioned Weight 8 Moisture 7 45% Stock Flow % Sample Number, Ts = 5, Total Samples = 8653 Additional 20 to 30% potential with improved stock flow regulation

48 OS -48 Headbox 2 Before Vs After Tuning Before After 07PIC2201 G: 6 to 2.2, Ti 11 to 13 07LIC2202 G: 1.4 to 0.6

49 OS -49 Base Stock AI Setup Change Raw Data 07FIC101:MV 07FIC101:POUT 07FIC101:WSP 07ZI100.AI:VALUE Sample Number, Ts = 5, Total Samples = 3242 MV Deadband changed from 0.4% to 0.0%

50 Oc c urrenc e MV OU T 26-PIC-0334.M V PIC-0334.OUT Oscillation Strange Bad Valve Process Out Of Range No Signal Integral Loop Performance Fingerprint Provides information on: Process and interactions Product Variability Raw material supply Evaluates: Historical Data - Process and control loop interactions. Identifies true source of process disturbances. Evaluates: Bump Tests - Performance of the DCS control loops within two categories including control, process, and signal conditioning issues. Root Cause Analysis Process Stability Process Stability Index 26-PIC-0334 Loop Tuning Evaluation Evaluates: Overnight data - along with the current DCS current loop tuning parameters are used to benchmark DCS control performance. Pressure Control Tuning Proportional Operations interview to define loop priority Provides information on: Over and under control Output Oscillation Offset and out of range loops Limit cycle and manual mode Control loop tuning validity Provides information on: Process upset response Bad Valves Over filtering Quantized signals Noisy signals

51 Control Goals Disturbances Set Point - Controller FCE Error Output Input Process Sensor Actual Process Measured Process Signal Conditioning Controller Process Signal Conditioning Disturbances Set Point - Error Controller Output Output FCE Input Process Actual Process Sensor Actual Process Reduce error: SP-PV Insure Output impacts Process Measured Process Minimize difference between actual and measured process CONTROL PROCESS SIGNAL CONDITION C1: Static Output P1: FCE Out of Range S1: Dead Signal C2: Over tune P2: FCE Size S2: MV Out of Range C3: Slow tune P3: FCE Broke S3: Quantization C4: FCE travel P4: Calibration / FCE Leakage S4: Compression C5: Offset P5: Intermittent Disturbance S5: Excessive Noise C6: Error Deadband P6: Persistent oscillatory Dist. S6: Spikes C7: Setpoint Oscillations P7: Questionable S7: Step Out C8: Controller update rate slow S8: Over filter C9: Questionable S9: Sampling rate S10: Questionable

52 ABB Group September 26, 2012 Slide 52 Terms Signal Conditioning Process Measured Process Sensor Actual Process Output FCE Disturbances Signal Conditioning: Getting the measurement to the controller with as true a signal as possible! (~15 items) Sample time, Filtering, Quantization, deadbands, compression, compensation, calibration, saturation, decimal points, Dead Signals, Spikes, Outliers, non-gaussian noise, Noise bursts, square root extraction. Process: Ensuring that the FCE to Actual Process (Transfer Function) is repeatable and predictable. (~21 items) Disturbances that are more powerful than the FCE, Process nature changes, process nonlinearity (sideways tank), FCE failure: Stiction, Backlash, hysteresis, resolution (on time, pulse width), valve size, valve type, DP drop across valve, cavitation, vena contracta, valve positioners, cams, electronic cams, VFD, process out of range. Control: Deals with keeping the error near or at zero. Means is we want the MV to track the SP. (~25 items) Tuning parameters, algorithm type, Setup parameters related to hi and low setpoint, output, and PV ranges, number of decimal points, Controller execution rate, Control deadbands, Additional degrees of freedom such as non linearity compensation options: gain scheduling, beta factor, adaptive control, out of range: wind up, loop is off(manual), offset from setpoint, rate of change of output, rate of change of setpoint, filtering. Input Process Process Sensor Actual Process Set Point Control - Error Controller Output

53 KPI Determination Mathematical Formulations Solution Surface Loop Performance Data Setpoint Measured Value Output Mode (Optional) Setup Tuning Parameters Diagnoses KPI Rules

54 KPI Navigation Overall rating for KPI Categories KPI sorted by number of problems KPI sorted by severity Trend Visualization

55 Loop Analyzer Classification Sorting on KPI severity Loop Performance KPI s Control Process Signal Conditioning Supports Loop Performance Fingerprint Output Out Of Range KPI Reporting Control Signal Conditioning Process Process Out Of Range Oscillating Output Manual Over Control Slow Control Oscillating Setpoint Offset Deadband Noisy Spikes Quantized Over Filtered Oscillation Strange Bad Valve PIC21- LIC21- LIC21-1MT1.CT LIC21- LIC21- PDIC11- LIC21- FIC30-1 LIC RL FIC101ST 1095 FIC FIC TDH LIC LIC21- LIC21- FIC21- LIC21- LIC21- FIC30-2 PIC PIC FIC PL1.CTRL FIC LIC LIC PIC21- FIC21- LIC21- FIC21- FIC21-3 LIC B FIC FIC MT1.CTRL 1154 FIC FIC21- CIC21- PIC21- FIC21- NIC30- FIC21- PIC21-4 LIC :MV FIC FIC21- PIC21- LIC11- RIC21- FIC21- FIC21-5 LIC LIC BP_SIZE FIC21- LIC21- PIC21- FIC21-6 PIC LIC FIC101ST 1103B LIC LIC21- FIC21-1PL1.CTR PIC21-7 LIC L 1145 PIC PIC21- FIC21- FIC21- FIC21-8 PIC A PIC FIC LIC LIC BP_SIZE 1PL1.CTRL FIC21- CIC21-10 PIC RIC FIC21-11 LIC PIC LIC PDIC11-1SP.CTRL: 12 RIC LIC MV FIC21- LIC21-13 BP_SIZE LIC FIC LIC LIC LIC FIC DW1.CT RL 17 1MT1.CT RL FIC21-

56 Loop Performance Fingerprint Loop Analyzer Performance Calculations Navigation options Diagnosis results in tree Loops Ranked for selected Diagnosis Navigate by clicking on bars ABB Inc. September 26, 2012 Slide 56 Diagnosis results in frame for selected loop

57 Loop Performance Fingerprint Loop Analyzer More Plots Power Spectrum Plot types Histogram Quick comparison plots, any plot type, one tag frozen on top half ABB Inc. September 26, 2012 Slide 57 Group Trend

58 Loop Performance Fingerprint Final Control Element Problem This is a flow controller that is the inner loop of a cascade. Exhibits classic stiction Controller output ramps up and down in triangular pattern Process variable moves in square wave ABB Inc. September 26, 2012 Slide 58

59 Loop Analyzer on Compressed Data

60 Compressed Data resulting in wrong classification

61 ABB Inc. September 26, 2012 Slide 61 Loop Performance Fingerprint Report The report highlights some loops, as shown in the previous slides and summarizes the results in tables. This table is for the TFE Synthesis section of the plant.

62 DP1 Loop Performance Fingerprint LI2 Plant wide Disturbance Analysis DP2 Disturbances in chemical plants act on many process variables TI1 TI2 TC1 LC1 LC1 DP1 DP1 TI2 TI3 LC1 DP1 LI2 DP2 TI6 DP1 TI7 LI2 LI2 LC1 DP2 DP2 DP1 TI1 TI1 LC1 TI2 LI2 TI2 DP1 DP2 TI5 DP2 TI4 LI2 TI1 LC1 LI2 TI1 TI2 TI1 DP2 TI2 TI1 Disturbances can propagate counter TI2 flow because of recycle and thermal integration ABB Inc. September 26, 2012 Slide 62

63 DP1 Loop Performance Fingerprint LI2 Plant wide Disturbance Analysis DP2 Disturbances in chemical plants act on many process variables TI1 TI2 TC1 LC1 LC1 DP1 DP1 TI2 TI3 LC1 DP1 LI2 DP2 TI6 DP1 TI7 LI2 LI2 LC1 DP2 DP2 DP1 TI1 TI1 LC1 TI2 LI2 TI2 DP1 DP2 TI5 DP2 TI4 LI2 TI1 LC1 LI2 TI1 TI2 TI1 DP2 TI2 TI1 Disturbances can propagate counter TI2 flow because of recycle and thermal integration ABB Inc. September 26, 2012 Slide 63

64 DP1 Loop Performance Fingerprint LI2 Plant wide Disturbance Analysis DP2 Disturbances in chemical plants act on many process variables TI1 TI2 TC1 LC1 LC1 DP1 DP1 TI2 TI3 LC1 DP1 LI2 DP2 TI6 DP1 TI7 LI2 LI2 LC1 DP2 DP2 DP1 TI1 TI1 LC1 TI2 LI2 TI2 DP1 DP2 TI5 DP2 TI4 LI2 TI1 LC1 LI2 TI1 TI2 TI1 DP2 TI2 TI1 Disturbances can propagate counter TI2 flow because of recycle and thermal integration ABB Inc. September 26, 2012 Slide 64

65 Loop Performance Fingerprint PCA Cluster Example Find signals with similar patterns, probably due to disturbances Not looking for oscillating signals Here all signals are in two columns that are adjacent ABB Inc. September 26, 2012 Slide 65

66 ABB Inc. September 26, 2012 Slide 66 Implementation Example PID Loop Parameter Example Sweet Spot Questionable Range The tuning documented during Fingerprint Improved tuning after implementation phase

67 Signal Analysis

68 LoopTune: Implementation improvements Visualization and Setup Analysis Standard Reporting LoopTune Identification Collection Tuning and Simulation ABB Group September 26, 2012 Slide 68

69 LoopTune Supports Self Regulating Non-Self Regulating Auto Model Identification Results stored as a LoopTune channel in ServicePort Provides long term process model tracking.

70 LoopTune: Tuning/Simulation Supports 800xA Controllers Harmony/Infi90 Controllers Mod300 Controllers Generic industry standard controllers Plug in modules for custom controllers

71 LoopTune: Loop Tuning Report Automatically generates Loop Tuning Reports

72 Example of improved control loop performance Much Better Manual Operation Original Tuning (Unstable) New Tuning ABB Group September 26, 2012 Slide 72

73 AGP400: High Speed Data Collection Portable or Continuous P/V Kit ABB Group September 26, 2012 Slide 73

74 Sequence Analysis Transition Time Deviation Error KPIs Prediction Accuracy ABB Group September 26, 2012 Slide 74 Multiple Changes Custom Trend Visualization Detailed Analysis Standard KPI

75 Batch Analysis Overview of Methodology Batch Process Define Event Marking Perform Aggregate Statistics and KPI s Data View

76 Segment Definitions Sequence Definition options: leading edge, falling edge, time trigger, user defined, external triggers, logical conditions Sequence 3 Sequence 2 Sequence 1

77 Introducing Performance Channels Goal: Maintain improved performance level Adjust service operating procedures Improve standard operating procedures Remote Capable monitoring ABB Group Group September 26, 2012 Slide 77 26, 2012 Slide 77 Specifics are a function of the Implement phase Periodic monitoring of key process indicators utilizing local or remote expertise

78 Oc c urrenc e MV OU T PIC-0334.M V Process Stability Index 26-PIC-0334.OUT 26-PIC Oscillation Strange Bad Valve Process Out Of Range No Signal Integral Corrective Action 3 Scan/Track: Remote Enabled Delivery Options Local Diagnose Implement Delivery Sustain Remote Enabled Packaged Services (Fingerprint) Periodic (LoopScan) Continuous (LoopTRACK) Root Cause Analysis Loop Tuning Evaluation Root Cause Analysis KPI Alarm Pressure Control Tuning Phone Process Stability Proportional Cluster Analysis KPI Notify Trigger Message Loop Stability KPI Local Delivery Troubleshooting Automated Reporting Event Reporting ABB Group September 26, 2012 Slide 78

79 3 ABB s Advanced Process Control Methodology Stabilize Process - Loop Performance Fingerprint Process Interaction Matrix Identification - APC Fingerprint APC System Solution (MIMO) Process Interaction Matrix Identification - APC SCAN Monitor Performance LoopSCAN Service Pre Study + Implementation Quick Customer Value Project delivery improved Installation and commissioning Proven Approach Service Periodic Service No results erosion Continuous Improvement ABB Group September 26, 2012 Slide 79

80 Technical Tool Architecture Service Applications Driver Optimization Tools View Analyze Interpret Report Fingerprint Report Unique Solutions System Hardware Software Control Process Get Structured Data Sort Data Segments KPIs Notify Common Infrastructure Alarm Phone Trigger Message Track Report DataPro Event History: KPI + Data Scan Report ABB Group September 26, 2012 Slide 80

81 LCS Service Key: ABB Applications Standard Apps: Boiler Fingerprint Alarm/Event Fingerprint System Performance Loop Analysis Transition Analysis Batch Analysis Tuning APC Fingerprint Technical Tools DL300 Data Logger Data file User Application User Interface Analyzer Engine KPI File Fingerprint Report OPC Establish Value Proof Statement Diagnostic Service Tools Stand alone User Driven File based Implementation Corrective Action: Tuning, Software, Hardware, etc. Stand alone Technical Tools

82 LCS Service Key: ABB Applications Standard Apps: Boiler Fingerprint Alarm/Event Fingerprint System Performance Loop Analysis Transition Analysis Batch Analysis Tuning APC Fingerprint Technical Tools DL300 Data Logger Data file User Application User Interface Analyzer Engine KPI File Fingerprint Report OPC Establish Value Proof Statement ServicePort Diagnostic Service Tools Stand alone User Driven File based Implementation Corrective Action: Tuning, Software, Hardware, etc. Scan Services Regular Track Services Monitoring Stand alone Technical Tools

83 ABB ServicePort Access to Efficiency! The ABB ServicePort is the secure portal through which customers access configuration tools, diagnostic applications, improvement activities, performance-sustaining troubleshooting, and scanning software that deploys agreed actions. ABB can connect to any system through ServicePort, which resides at the customer site, and implement fixes to diagnosed problems. Regular delivery of Scan and Track services can be done safely and securely with ABB s ServicePort. ABB Group Group September 26, 2012 Slide 83 26, 2012 Slide 83

84 ServicePort topology Service Capabilities Event Notification Control Tuning Optimization Services Support Services Software Support System Health Check Remote Troubleshooting Remote Firewall Secure Access Local Customer-Defined Access ServicePro LoopScan/Track DriveScan/Track HoistScan/Track System Scan/Track Troubleshooting Services APC Scan/Track Services Secure Tunnel Customer-Defined Access Engineering Stations ABB Group September 26, 2012 Slide 84 Operator Stations Drives OCS Historian Instruments/ Actuators

85 Performance Channel Guide concept ABB Group September 26, 2012 Slide 85

86 ServicePort: Data Flows 800xA Standard OPC OPC Direct OCS 800xA Standard OPC Raw HSD500 OPC Data PDA Channel Alarm Channel PDA Analyzer Alarm Analyzer ABB OPC Data Logger (DL300) Loop Analyzer Loop Tune Batch Analyzer Loop Channel Loop Tune Batch Channel ServicePort Explorer

87 ServicePort Explorer Define/Search for Triggers ServicePort Explorer Channels Process Platform System Drives... Event 1 Event 2 Event n KPI1 Engine KPI2 Engine KPIn Engine ABB Group September 26, 2012 Slide 87 Explorer: View/Scan/Track Service Channels

88 Severity Service Channel Components Trend for Experts View Expert level. Visual validation. No automatic triggers. Data arranged by asset requirements. SCAN Math Function KPI(s) Determination of a KPI that is proportional to expert evaluation. Defines good and bad performance. TRACK Rules Goal Gap> rule= Alarm Present Continuous evaluations of KPI calculations and reporting. Tailored to customer requirements of asset criticality.

89 ServicePort Explorer Dave View 1. Select ServicePort Displays navigation Menu 2. Expand Performance Analysis menu item 3. Expand Channel 10.1 View 4. Select DataView. This will cause the display that is shown to appear 5. Next Select a loop name and a time event. The SP, MV, and Out along with loop KPI s will be displayed

90 ServicePort Explorer Scan: Event and Time Views In this case, the Signal Conditioning KPI s indicate a large number of violations. 1. Select ServicePort Displays navigation Menu 2. Expand Performance Analysis menu item 3. Expand Channel 10.2 Scan 4. Select EventView. This will cause the display that is shown to appear 5. Next Select a time event. The number of loops with problems related to control, signal conditioning, or Process for this time event will be shown. 6. KPI Level 1 and Level 2 Pareto charts can be drilled into from the top drop down menu. 7. Historical Trends of KPI s per event can be evaluated by selecting Time Based View

91 ServicePort Explorer Scan: Level 1 Event view In this case, data compression is the cause of the signal KPI violations. This causes analysis results to be less accurate

92 ServicePort Data View

93 ServicePort Explorer Improved Navigation Much Faster presentation of information.

94 ServicePort Explorer Track (Near Future Release) User can pick the KPI of interest User builds Notification Rules. Rules are logical expressions comparing KPI s to thresholds of interest. Violations will be stored in an event list or sent via for support

95 KPI Tracking Sustain: Scan KPIs to ensure improvement Q1 Q2 Q3 Q4 Production increase! Continuous Improvement Variability decrease! Kpi1 Kpi2 1 Kpi3 Kpi4 ABB Group September 26, 2012 Slide 95 Delivery Schedule

96 LoopScan Performance Service Report Statistical Evaluation of Historical KPI s performed twice a year Ensures stability of KPI s Reduces the risk of false positives Keep up to date with process Crucial to ensure continuous improvement

97 Service Delivery Single Channel Delivery model January December Calibration Fingerprint/Training Regular Performance Calculations/evaluations/Daily Usage SCAN Service Report: SCAN Analysis /report/implementation planning/ Notification Validation On Demand Service: Remote troubleshooting assistance/track Triggered

98 Channel Offerings Industrial Automation Loop Performance Batch Performance Transition Performance Plant wide disturbance analysis LoopTuning OEE APC Interactions Platform Diagnostic Services Harmony Performance 800xA Performance Alarm and Event Security Drives GMD Low Voltage Minerals Hoist Performance Dragline View Options: View Scan - Track Pharma Oil and Gas Chemical Pulp and Paper VPA Control Utilization Standardize Check Sample Calibrate Sample Machine Direction Grade Change Transition Batch Sheetbreak recovery Startup efficiency Profile performance Profile transition profile recovery Profile tuning Machine Direction QCS Tuning Product Variability Machine Response Profile Capability

99 Channel Offerings Industrial Automation Loop Performance Batch Performance Transition Performance Plant wide disturbance analysis LoopTuning OEE APC Interactions Platform Diagnostic Services Harmony Performance 800xA Performance Alarm and Event Security Drives GMD Low Voltage Minerals Hoist Performance Dragline View Options: View Scan - Track Pharma Oil and Gas Chemical Pulp and Paper VPA Control Utilization Standardize Check Sample Calibrate Sample Machine Direction Grade Customized Service Change Transition from Batch Sheetbreak recovery Proven Solutions Startup efficiency Profile performance Profile transition profile recovery Profile tuning Machine Direction QCS Tuning Product Variability Machine Response Profile Capability

100 Service Delivery Multi Channel Delivery model January December Time Event Channel Process/Equipment Calibration Fingerprint/Training Periodic Performance Calculations/evaluations: Time based, Event based Scheduled SCANService: SCAN Analysis /report/implementation planning/ Notification Validation On Demand Service: Remote troubleshooting assistance

101 ABB Group September 26, 2012 Slide 101 ServicePort Pilot sites SCA Eerbeek, NL Gasco Habshan, AE Glatfelter Chillicothe, OH, US ABB Finland (development system) IP Augusta, GA, US ISAB Powerplant, IT Kinross Mining, BR Lion Copolymer Geismar, LA, US Sappi Nijmegen, NL UPM Changshu, Jiangsu, CN GP Big Island, VA, US APP Tjiwi Kima, Sidoarjo, ID Boise International Falls, MN, US Domtar Ashdown, AR, US NOVA Chemical Painesville, OH, US

102 Trends ServicePort - Benefit to Customer Directly impacts: Production Production Platform Stability Maintenance Expense ABB Apps ServicePort provides: Proven Solutions, Immediate access to experts, Customizable Applications

103 Advanced services: increasing customer loyalty and expanding service scope Example: Chillicothe paper, Ohio, USA Customer challenge: poor maintenance practices, poor system performance and low quality after market parts used by a third party negatively effecting the customers own product quality. ABB solution: A QCS performance Fingerprint, ServicePro contract management tool were utilized to identify the issues and to show best practices and tools. Optimization services like continuous performance tracking with ServicePort were included to create customer value beyond traditional maintenance contracts. Customer benefits: Value-add analysis which provided the customer with increased automation efficiency beyond the original scope of the audit.

104 Advanced services: increasing customer loyalty and expanding service scope Example: Lion Copolymer, Louisiana, USA Customer challenge: A three-line chemical polymers plant makes multiple grade changes. Automation performance has declined as product line expanded into specialty chemicals. ABB solution: LoopPerformance and Transition Fingerprints packaged with Implementation Services for Tuning and Corrective Action. ServicePort for sustaining production enhancements. Customer benefits: Estimates of over $1Million per year in increased production and reduced off-spec waste byproducts. ABB benefit: 1 st Chemicals Industry contract in the USA with Transition Fingerprint and ServicePort. 1 st Fingerprint plus Implementation service contract for Chemicals Industry growth initiative. Enhanced competitive position against 3 rd party with black box APC claims.

105 Customer Case Study Pulp mill, Arkansas, USA Customer need Replacement of knowledge that left the company Expert support for legacy control system Fast responses to production issues Site: Pulp Mill, AR, USA Unit: Pulp Mill Issues: Customer lost an important engineer INFI 90 system required upgrade Self-maintenance no longer possible Agreement: 1-yr agreement + resident engineer ABB s response ServicePort for onsite control system support Delivers ABB expert resources remotely Provides tech support Performs non-invasive checks on the operating system Temporary resident field service engineer Customer Benefit Replaces lost expertise Solves tech support issues immediately Identifies problems before they occur Provides diagnostics to troubleshoot disturbances Delivers ABB Optimization Solutions Reduces costs Need-identification-to-resolution within 45 days ABB Group

106 ServicePort: Partnering for Success Customer Benefit Use of same technical tools that ABB experts use. Regular updating of software. Training/support/access to world class subject matter experts. Continuous Monitoring of customer defined assets. Optimize time to solution as well as resource planning. Periodic Performance Scan reports to ensure continuous process improvement, reduced false positives, improved key performance indicators, adaptation to conditions or process changes.

107 ABB Group September 26, 2012 Slide 107 ServicePort + Channels ServicePort Base Unit Calibration Fingerprint Start-up Services Process Channels 2 scan reports per year On Demand Process Support Equipment Channels 2 scan reports per year On Demand Process Support Workbench Troubleshooting/Implementation Tools: DataLogger, Signal Analyzer, LoopTune, Loop Analyzer, Sequence Analyzer Remote Access Platform

108 Loop Performance Services Portfolio Saleable Service Selection Fingerprint Implementation WorkBench Tools Data Logger Signal Analyzer Loop Tune Loop Analyzer ServicePort Base Unit Process Channel Remote Access Platform Secure Data Tunnel Training ABB Group September 26, 2012 Slide 108

109