Automated Machinery
Maintenance
Bill Powell
Tony Burnett
Agenda
Industry Trends & Challenges
Machinery Health Strategy
Automating Decision Support
Key Points
The Game of Plant Reliability - Planning
Question: Can we make it to the next outage?
Challenge: Track fault get us to Nov 20
OUTAGE PLAN
Action: Alert Maintenance Planner 1. Change Bearing $$$
2. Align Motor $$$
3. Install New Coupling $$$
Planned Outage Planned Outage
Failure
Outage
Business Metrics
Areas that are impacted most by Predictive Decision Support
Key Trends in Industry
Meet plant OUTPUT requirements
Eliminate unplanned failures
Control COSTS
Smart skill set utilization O&M COST
Annual Maintenance Cost
Optimize work planning O&M COST
Analyze MTBF / MTBM
% Unplanned Maintenance
Work Order Compliance
Control outage times
Utilize CMMS / Maint. Planning
Program Performance
1988 2001 Ideal
Reactive 55% 55% 10%
Preventive 30% 31% 25-35%
Predictive 10% 12% 45-55%
Proactive 5% 2% Balance
Reliability Magazines 2002 report
Over 11 years, plants are still REACTIVE.
World Class organizations focus on PREDICTIVE
Cost Statistics of the HPI Segments
Skill Sets
75% or operations retiring in the next 15 years
Gray matter is leaving the plant!
Costs
20-30% of costs are related to maintenance
Mechanical Reliability
3-7% Capacity Impact in loss/slowdown in production of key
process equipment due to unplanned down
Largest is due to mechanical equipment (40%)
Predictive Maintenance program shown to impact reductions to
unplanned mechanical shutdowns by ~30%
* HPI = Hydrocarbon Process Industries
Operational Challenges of Today
Our target is ZERO unplanned downtime!
Maximize Equipment Availability & Reliability
Plan all maintenance - HOW?
We are trying to be competitive today with a plant that is
typically more than 40 years old - and so are our competitors.
Extend Machinery Life & Rebuilds
We are running our equipment beyond its design capacity to
handle the variety of materials that we must process
Running equipment beyond rated capacity
Increased Throughput, but without RISK?
Answer These Tough Questions...
What maintenance does the machine need during the next planned
shutdown?
Are the spare parts in my inventory?
Is my spares inventory too large?
Can the equipment run beyond the next scheduled outage?
Do I know when cavitation is occurring?
Fact: 40-60% of machine problems related to lack of feedback to
operators*
Automation systems provide feedback (i.e., Cavitation) back to Operator
* Ron Moore book
Answer These Tough Questions...
Do you know the # of failures ranked by equipment type?
Rotating Equipment can be segmented into
Pumps / Fans / Compressors
Pumps can be segmented into
Centrifugal / Recip / Rotary / special effect
What are your most problematic machines?
Are they the bottlenecks in your production facility?
These machines = downtime cost for entire plant.
Automated monitoring strategy would be ideal for these cases
Answer These Tough Questions...
What are your Key Performance Indicators (KPI)?
# failures, MTBF/MTBM
Cost of Repairs, Avg. Repair Cost,
Maintenance costs by Equipment Group
Quantified Lost Production Opportunity (LPO)
Ex.: ExxonMobil reduced maintenance costs 20% corporate-
wide by considering these categorical cost concepts
Answer These Tough Questions...
Have you identified your most critical equipment?
One facility has identified that
~ 6 machines can bring them to 0 output MOST CRITICAL
~ 30 machines can reduce output to 40-60% ESSENTIAL
Point: Some output is better than no output!
If production is $1M / day, then 50% output loss = $500K / day
Likely comes from O&M Budget
Ex.: TN Eastman has 27,000 pieces of equipment
3% (or ~750) need online-predictive & protection (AMI 1)
26% (~7000) need online and walk-around (AMI 2 or 3)
What percentage would your operations group tell you?
* AMI = Asset Management Index (1-5, 1 is most critical)
Conclusions
The market is forcing major industrial facilities to adopt
new methods for the sake of profitability
Equipment is getting oldRunning beyond design speeds
Planning is everything
But we dont have unlimited man-power industry is challenged!
Trained skill sets are starting to retiring
How do you capture this knowledge?
Do I have MTBF statistics?
CMMS is becoming important tool for O&M strategies
Identify the machines that reduce greatest percentage of
output reduction
These are most critical! Automation is easily justifiable.
Agenda
Industry Trends & Challenges
Machinery Health Strategy
Automating Decision Support
Key Points
Machinery Health Strategy
DESIGN
Business Plant Failure Performance
Objective Assessment Defense Quantification
Identification and Planning
Benchmarking
IMPLEMENTATION
REVIEW
Technology Expertise Performance
Deployment Optimization Improvement
Measurement
Planning
Work and Analysis
Process
Optimization
Continuous
Improvement
Asset Optimization
Technology
To provide decision support
Expertise
Qualified personnel with
current knowledge
Work Processes
To focus resources on
priorities
Agenda
Industry Trends & Challenges
Machinery Health Strategy
Automating Decision Support
Key Points
Automated Monitoring Concepts
Detect Detailed Machinery
Problems
Unbalance, Misalignment, Looseness
Looseness, Shaft Cracks, Oil Whirl, Problem!
Phase, Rubs, Gear and Bearing
Problems
Signal Processing Flow
Transducer
Waveform
Amplitude
Time
FFT
Spectrum
Amplitude
Data Collector/Analyzer Frequency
Time Waveform
Amplitude
Heavy Spot
+
0 Time
360 degrees
-
Rotation
1 revolution
3600 rpm = 3600 cycles per minute
60 Hz = 60 cycles per second
1 order = one times turning speed
Time Waveform
Amplitude
+
0 Time
-
1000 rpm 1 revolution
4 blades = vibration occurs 4 times per revolution
4 x 1000 rpm = vibration occurs at 4000 cycles per minute
= 4000 cpm
Time Waveform
Amplitude
+
12 tooth
gear 0 Time
-
1 revolution
1000 rpm
12 teeth are meshing every revolution of the gear
12 x 1000 rpm = vibration occurs at 12,000 cycles per minute
= 2,000 cpm = 200 Hz
+
0 Time
0 Time
+
0 Time
-
Complex Time Waveform
Time Waveform contains all the different
frequencies mixed together
+
0 Time
-
Complex Time Waveform
Time Waveform contains all the different
frequencies mixed together
We are now entering the Frequency Domain
FFT - Fast Fourier Transform
Separates individual frequencies
Detects how much vibration at each
frequency
TIME WAVEFORM
AMPLITUDE VS TIME
Amplitude
Time
Amplitude
Time
Amplitude
Amplitude
Frequency
0 Time
- 1x
Frequency
0 Time
- 4x Frequency
+ Time
0
- 12x Frequency
Predefined Spectrum Analysis Bands
1xRPM - BALANCE
1.8
2xRPM - ALIGNMENT
1.5
1.2 3-5xRPM - LOOSENESS
0.9
ANTI-FRICTION BEARINGS & GEARMESH
0.6
5-25xRPM 25-65xRPM
0.3
20000
5000 10000 15000
Frequency Hz
Bearing Fault Frequencies
Function of the Geometry of the Bearing
Outer Race
(BPFO)
Inner Race
(BPFI)
Ball Spin
(BSF)
Cage
(FTF)
ROLLER BEARING EXAMPLE
Frequency Band Alarming and Trending
Sub-
Harmonic 1X 2X Bearing Bearing Gears Bearing
Amplitude
1x 2x 25-60 x
Trend of Trend of
Balance Bearings
Alarm
.5
Alarm
in/sec
Warning .1 Warning
in/sec
Time Time
(Days) (Days)
Automated Monitoring Concepts
Detect Detailed Machinery
Problems
Unbalance, Misalignment, Looseness
Looseness, Shaft Cracks, Oil Whirl, Problem!
Phase, Rubs, Gear and Bearing
Problems
Pass Information to DCS Plant
Information Systems via OPC
Use existing plant LAN Ethernet
infrastructure
Like Process Control for the Vibration World
PLC
PLC
Controllers/Servers
PLC Network Hub,
Network Hub,
Router, or Switch
Router, or Switch
PLC
Reliability Engineering Maintenance Office
Customer Provided Customer Provided
Windows 95/NT Windows 95/NT
=
Workstation with CSIs =
Workstation with CSIs
Online WATCH S/W Online WATCH S/W
loaded. loaded.
Control Room
Generalto Specific
Graphical Interfaces - Overall
2301 Compressor - Asset Reliability System - CSI 4500 Series Online RBMCONSULTANTTM
Graphical Interfaces - Specific Faults
Monitored Fault Value
Overall Vibration 3.204 mils
Oil Whirl/Whip 0.506 mils
Rubs 0.112 mils
Unbalance 1.886 mils
1x Peak 0.231 mils
1x Phase 72 degrees
Misalignment 0.107 mils
2x Peak 1.886 mils
2x Phase (cracked shaft) 35 degree
Looseness 0.231 mils
Non-Rotational 0.501 mils
Automated Monitoring Concepts
Detect Detailed Machinery Problems
Unbalance, Misalignment, Looseness, Shaft
Cracks, Oil Whirl, Phase, Rubs, Gear and
Bearing Problems Looseness
Problem!
Pass Information to DCS Plant
Information Systems via OPC
Use existing plant LAN Ethernet
infrastructure
Confirm mechanical conditions will reach
planned shutdown, i.e., plant capacity
target
Fix what is broke before failure
Failure
Outage
Automation Technology Best Practices
Data-to-Information
Assess condition/faults of machine in field
Report results not just data
Accuracy of automated analysis
Combine analysis with machine operating condition
Addresses false alarming
Report-upon-exception
Keep skilled analysts focused on problems
Addresses the data generator issue
Report-upon-Exception not just data
- - - - - - - HI HI - - - - - - - - - - - - - CRITICAL - - - - - - - - - - - - - - - - - - - - - - - - - DEADBAND
--
Deadband (hysterysis) - limits annoyance alarms
- - - - - - - - HI - - - - - - - - - - - - - - URGENT - - - - - - - - - - - - - - - - - - - - - - - - - - -DEADBAND
-
ALARM TYPES RATE-OF-CHANGE URGENCY ACTIONS
Fault HI HI (ROC) Critical Relay Activate
Caution HI
Caution LO Urgent Relay Activate
Fault LO LO Notify Relay Activate
D Rate ROC Normal OPERATE
D Amplitude ABS EPS RELAYS
D Time Report
!
Exception
D Amplitude Report (ABSOLUTE EPLISON)
Report on Exception
TIME HISTORY
System reports only when health CHANGES
Benefits of Reporting only Exceptions
Too Much Data! Automation should handle data rejection so
111111111111111111111111111111111111111
111111111111111111111111111111111111111
Reliability Group only has to handle EXCEPTIONS
111111111111111111111111111111111111111
111111111111111111111111111111111111111
111111111111111111111111111111111111111 Allows valuable human experience to
11111111111111111111111111111111O111111
111111111111111111111111111111111111111
11111111O111111111111111111111111111111
be applied to analyzing problems
111111111111111111111111111111111111111
111111111111111111111111111111111111111
111111111111111111111111111111111111111
Logic helps to ID
111111111111111111111111111111111111111
111111111111111111111111111111111111111
canned observations
111111111111111111111111111111111111111
111111111111111111111111111111111111111
versus
111111111111111111111O11111111111111111
111111111111111111111111111111111111111
reporting numbers
11111111O111111111111111111111111111111
111111111111111111111111111111111111111
111111111111111111111111111111111111111 Result is
- Eliminate repetitive work
- Easy for end user
- Efficient analysis only on problems
- Optimize labor effort
Agenda
Industry Trends & Challenges
Machinery Health Strategy
Automating Decision Support
Key Points
Using Web Technologies to Enable
Remote Monitoring & Analysis
Provides customers with cost-effective access to
information and expertise not previously
available in their own plants
AOS Server
HQ
Process
Plant
Emerson
Service
Process
Professional
Plant
Reliability
KEY: Engineer
Subsidiary
Submit Data
Access to Reports via Internet Link
Asset Management Integration
Asset
Enterprise Optimization
Interface World Wide Web
Ethernet WAN
Operation Optimization Asset Management
Control and GRA
Application
Decision CMMS Device Equipment Condition Efficiency
Process Control Process
Support Management Diagnostic Monitoring Monitoring
(DeltaV)
Ethernet LAN
Control Bus (FieldBus, HART,
Connectivity ProfiBus), Ethernet LAN, modem
Device I & C monitoring, measurement, and
regulation
Assets Motors, pumps, fans, boilers, turbines Periodic Monitoring
motor control centers, transformers
Emerson Asset Optimization Architecture phase1
AOweb
Asset Optimization Server
Data Collector
AO Web AO Web AO Web
AMSweb Services
RBMweb Services e-fficiency Services
AMS RBMware e-fficiency
Using Advances in Web Technology
e-fficiency
AMSweb
Asset Optimization
web server
Link Relevant Information
Embed links to other sources of
information
Key Points
Automation & Decision Support Tools help the
plant meet OUTPUT & COST targets by
1. Optimize Reliability Groups Time / Effort
2. Capture Plant Knowledge
3. Reduce O&M Costs by Managing Reliability
