PROGNOSTÂŽ for LDPE
Online Condition Monitoring and Machine Protection
e s in p a r a
lle
8
hin
l
3
IEC61508:2010
mac
Automated Diagnoses and Messaging
Performance Optimization
Monitoring in LDPE service Monitoring of machines in LDPE (Low Density Polyethylene) service demands a highly adaptable and reliable technology. PROGNOST®-NT is the leading system for monitoring Booster/Primary and Secondary Compressors in LDPE plants. PROGNOST®-Predictor monitors Extruders and complex Gearboxes with vibration and provides an early warning of upcoming bearing failures. Learn how these PROGNOST® technologies enable operators to deploy successfully a predictive maintenance strategy and operate machines safely and reliably. PROGNOST monitors hundreds of machines in the LDPE industry at all major production companies.
New LDPE plants 2005 – 2018 started – up with PROGNOST® systems As of January 2019
3
Compressor Monitoring Typical Sensor Positions (Example: Hyper compressor) 1 Trigger switch 10 Acceleration vibration sensors (Crosshead) 20 Horizontal/vertical plunger position sensors 10 Strain Gauge rings for high pressure p-V diagrams DCS values e.g. temperatures (bearing, valves), line pressures 10 Acceleration vibration sensors (Cylinder) 2 frame vibration sensors (optional)
(Example: Booster/Primary Compressor) 1 Trigger switch 6 Acceleration vibration sensors (Crosshead) 6 Rod position sensors 12 Dynamic pressure transducers DCS values e.g. temperatures (bearing, valves), line pressures 6 Acceleration vibration sensors (Cylinder) 2 frame vibration sensors (optional)
Typical PROGNOSTÂŽ-NT sensor signals p-V diagrams of a Hyper compressor measured by strain gauge rings to monitor gas dynamics e.g. to analyze valve action with different gas compositions, check pulsation levels etc.. 3583 bar a
TDC
BDC
3000 2500 2000 1500 1000 500 0 0
20
40
60
80
100
120
140 160 180 Plunger stroke [mm]
200
220
240
260
280
300
320
PROGNOST Systems GmbH
4
Protection Analyses Visualizes the Machine Protection analyses performed by PROGNOST®-SILver, e.g. online signals, trends and safety limit violations.
Ring buffer Transient data recording allows maintenance technicians to replay a safety shutdown, alert, or machine start-up by examining a gapless recording of all signals in an uncompressed format – revolution by revolution. The ring buffer offers the possibility of subsequent analyses; the time frame of seven minutes before and three minutes after the ALERT, SHUTDOWN or UNSAFE alarm can be closely evaluated using all recorded time signals from all dynamic sensors and process values in the PROGNOST®-NT system.
Reliable machine protection hardware is only one side of the story. Answers as to “why” a safety incident occured is the next step further. PROGNOST®-NT Protection Analyses is a module to visualize and save online and trend data to provide all information required for precise root cause analysis.
Peak-to-Peak analysis of the piston rod position (operational run out) Displacement signals providing piston rod position are divided into 8 segments of 45° crank angle each for every revolution to detect critical mechanical conditions of the piston rod.
60 m il
BDC
TDC
TDC
45 30 15 0 0
15
30
45
60
75
90
10 5
12 0
13 5
15 0
16 5
18 0
19 5
21 0
degree crank angle
22 5
24 0
25 5
27 0
28 5
30 0
31 5
33 0
34 5
36 0
R D S tage 1, P eak-P eak over 8 S eg. .. 45 d egrees
m il 40 30 20 10 0
1 0
15
2 30
45
60
3 75
90
10 5
4 12 0
13 5
15 0
5 16 5
18 0
19 5
6 21 0
22 5
24 0
7 25 5
27 0
28 5
8 30 0
31 5
33 0
34 5
36 0
degree crank angle
The illustration shows two piston rod position signals: purple = normal position (“good condition”) green = strong fluctuations close to BDC (bottom dead center) due to liquid carryover inside the compression chamber
5
PROGNOST® plunger position monitoring Erratic plunger run out can be monitored with a PROGNOST® orbit visualization of the horizontal and vertical plunger run out and allows to set a threshold to a maximum deflection based on orbit or horizontal and vertical direction.
135
108
81
60
48
54
36
27
135
24
108
81
54
27
27 µm
54
81
108
53 4627
12
60
48
36
24
12
12 µm
24
36
48
137
60
54
12
81
4624
108
135
36 Machine C2202 C2202 C2202
48
Machine C2202
Measuring Point RD C2S hor
Dataname Roddrop Orbit
C2202 Online Orbit2.onl 09.08.2002 16:49:42 , 180.0 U/min
TAG-Name
Measuring Point RD C1S hor RD C2S hor RD C3S hor
Dataname Roddrop Orbit Roddrop Orbit Roddrop Orbit
TAG-Name
C2202 Online Orbit2.onl 09.08.2002 16:49:42 , 180.0 U/min
Value 46.6 0.93 7.9
Unit µm µm µm PROGNOST Systems GmbH
Value Unit 0.93 µm PROGNOST Systems GmbH
Plunger run out
Sensor arrangements High pressure packing flange
horizontal
Eddy current sensors
vertical
Plunger Surface Temp
6
135
Early Failure Detection •
Avoidance of costly damages through identification of mechanical defects at an early stage Information instead of data: clear text messages with local and functional clarity
•
With its dedicated analyses for reciprocating machinery, PROGNOST®-NT detects developing damages at an early stage, providing users with valuable lead time for proactive compressor operation management and efficient maintenance planning. Automatic detection of operating conditions PROGNOST®-NT recognizes changing machine operating conditions and automatically switches to corresponding, pre-defined threshold sets to avoid false warnings caused by changing load conditions.
Automated threshold setting Using the automated threshold setting, the system can easily be configured for new operating conditions to guarantee high quality of warnings. Pattern recognition with fully integrated diagnostic database All PROGNOST®-NT users benefit from experiences of more than 4.25 million recip operating hours annually and more than 20 years of diagnostic experience. All major failure modes are integrated within a failure pattern database and can be diagnosed automatically providing clear text messages including failure type and location of the failing component. The diagnostic capabilities are a blend of the most reliable and useful soft computing disciplines – from Fuzzy Logic to Rule Based.
Early Failure Detection on Hyper pressure cylinders (Correlation of machine functions to crank angle)
TDC (Top Dead Center) = Discharge Valve closes 45° after TDC = Suction Valve opens BDC (Bottom Dead Center) = Suction Valve closes
m/s²
V Cyl. 2F, Absolute maximum via 36 Seg. .. 10 degrees
240° after TDC = Discharge Valve opens
50
45
40
Segmentation of online vibration signal with 1st and 2nd warning threshold.
35
30
25
36 segments, each 10° crank angle, automatically monitor functions such as valve action and mechanical integrity of compressor drive gears to detect developing failures in early stages.
20
15
10
5 0
1
2
0
3
4 30
Machine 60-C1 002 2nd St Online 01.04.
5
6
7 60
8
9
10 90
Measuring Point V Cyl. 2F
11
12
13
120
14
15
16
150
Data name Absolute maximum
17
18
19
20
21
22
180 210 ° CA TAG-Name 60VT 10339
2008 13:00:01, 60-C-1002 2nd Stage - Absolute maximum via 36 Seg. .. , 200,2 1/min
23
24
25
240
26
27
28
270
29
30
31
300
32
33
34
330
35
36 360
Value Unit 0 - 2000 m/s² PROGNOST Systems GmbH
7
Early Failure Detection on Booster/Primary compressors (Correlation of machine functions to crank angle)
TDC (Top Dead Center) = Discharge Valve closes 45° after TDC = Suction Valve opens 45°
BDC (Bottom Dead Center) = Suction Valve closes 240° after TDC = Discharge Valve opens
Segmentation of online vibration signal with 1st and 2nd warning threshold. 36 segments, each 10° crank angle, automatically monitor functions such as valve action and mechanical integrity of compressor drive gears to detect developing failures in early stages.
Monitoring based on operating condition Adaptive condition monitoring based on operating condition e.g. due to changing suction/discharge pressures to avoid false alarms because of changing operating conditions
Automatic detection of the operating conditions and threshold adjustment
Vibration in „operating condition 1“ (2200 bar)
8
Vibration in „operating condition 2“ (2600 bar)
Gearbox and Extruder Monitoring PROGNOSTÂŽ-Predictor Monitor Critical Assets PROGNOSTÂŽ-Predictor is an automated condition monitoring system for rotating equipment that provides early fault detection for Predictive Maintenance by utilizing advanced diagnostics. Complete Component Diagnostics Specific analyses monitor multiple component types, including sleeve and rolling element bearings, gears, motors and shafts providing health information and historical data. PROGNOSTÂŽ-Predictor method uses the Confidence Factor
calculations to correctly analyze narrow band vibration signatures, allowing the system to identify magnitude and development of component damage at a very early stage while minimizing false alarms. Early Warning Early detection of machine faults allows scheduling of repairs well in advance, ensuring that all necessary parts and labor are available. Reduce or eliminate unscheduled outages and realize the benefits of an effective Predictive Maintenance approach.
Operators can check the status of each internal machine component through a series of graphical pictures. Components in an alarm state change color to indicate Warning, Alert, and Alarm status.
9
Gear Vibration Imaging Gear Vibration Imaging provides the ultimate method of detecting gear faults. Synchronous waveform averaging and further filtering methods isolate the vibration frequencies produced by each gear tooth as it passes through mesh, while eliminating noise and vibration generated by other mechanical components. Statistical pattern recognition methods are used to identify
and quantify these excess vibrations and alarm when the developing faults are detected. This method is superior to standard spectral analysis techniques because the energy from the fault is localized to a particular region in the vibration image. Gear Vibration Imaging detects developing gear faults much earlier than spectral methods.
Gear Vibration Imaging isolates the vibration generated by each gear in the gearbox while reject much interfering vibration. Damage to an individual tooth will produce localized high amplitude peaks.
Confidence Factor Confidence Factor is a pattern recognition technique that quantifies the similarity of the measured peaks to the expected fault peak frequencies and pattern. The Confidence Factor provides a strong indication of whether a measurement actually represents a real fault or damage. If
the Confidence Factor is low, it most likely means that the specific fault is not present, but if the Confidence Factor is high, then it is a safe bet that the fault indicated is real. The results of the banding calculation and the Confidence Factor are graphed in a Confidence Factor plot.
This plot shows a high confidence factor and high band amplitude which result in an alarm. All bands show matching peaks inside.
10
PROGNOST Systems GmbH Daimlerstr. 10 48432 Rheine Germany +49 (0)59 71 - 8 08 19 0 +49 (0)59 71 - 8 08 19 42 info@prognost.com
www.prognost.com
PROGNOST Systems, Inc. 1018 Hercules Ave. Houston, TX 77058 USA +1 - 281 - 480 - 9300 +1 - 281 - 480 - 9302 infousa@prognost.com