US20240036011A1 - System and method for corrosion and erosion monitoring of pipes and vessels - Google Patents
System and method for corrosion and erosion monitoring of pipes and vessels Download PDFInfo
- Publication number
- US20240036011A1 US20240036011A1 US18/265,433 US202018265433A US2024036011A1 US 20240036011 A1 US20240036011 A1 US 20240036011A1 US 202018265433 A US202018265433 A US 202018265433A US 2024036011 A1 US2024036011 A1 US 2024036011A1
- Authority
- US
- United States
- Prior art keywords
- groupid
- wall thickness
- probe
- monitoring
- probe assemblies
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 153
- 230000007797 corrosion Effects 0.000 title claims abstract description 92
- 238000005260 corrosion Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 90
- 230000003628 erosive effect Effects 0.000 title abstract description 24
- 238000005259 measurement Methods 0.000 claims abstract description 138
- 239000000523 sample Substances 0.000 claims description 161
- 230000000712 assembly Effects 0.000 claims description 86
- 238000000429 assembly Methods 0.000 claims description 86
- 230000006870 function Effects 0.000 claims description 52
- 238000013528 artificial neural network Methods 0.000 claims description 42
- 238000007689 inspection Methods 0.000 claims description 38
- 238000005457 optimization Methods 0.000 claims description 37
- 238000012549 training Methods 0.000 claims description 13
- 230000015654 memory Effects 0.000 claims description 8
- 239000000463 material Substances 0.000 claims description 2
- 230000036541 health Effects 0.000 abstract description 7
- 238000012545 processing Methods 0.000 description 42
- 238000012517 data analytics Methods 0.000 description 15
- 230000000875 corresponding effect Effects 0.000 description 13
- 238000004891 communication Methods 0.000 description 12
- 238000010801 machine learning Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 8
- 238000012986 modification Methods 0.000 description 8
- 230000004048 modification Effects 0.000 description 8
- 230000035945 sensitivity Effects 0.000 description 7
- 230000009471 action Effects 0.000 description 6
- 238000013459 approach Methods 0.000 description 6
- 239000012530 fluid Substances 0.000 description 6
- 238000012800 visualization Methods 0.000 description 6
- 238000010187 selection method Methods 0.000 description 5
- 230000003321 amplification Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000004590 computer program Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000001815 facial effect Effects 0.000 description 4
- 238000003199 nucleic acid amplification method Methods 0.000 description 4
- 230000002829 reductive effect Effects 0.000 description 4
- 230000001413 cellular effect Effects 0.000 description 3
- 238000003066 decision tree Methods 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000005284 excitation Effects 0.000 description 3
- 230000000670 limiting effect Effects 0.000 description 3
- 239000007788 liquid Substances 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- HNNQYHFROJDYHQ-UHFFFAOYSA-N 3-(4-ethylcyclohexyl)propanoic acid 3-(3-ethylcyclopentyl)propanoic acid Chemical compound CCC1CCC(CCC(O)=O)C1.CCC1CCC(CCC(O)=O)CC1 HNNQYHFROJDYHQ-UHFFFAOYSA-N 0.000 description 2
- 239000002253 acid Substances 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000013079 data visualisation Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- IHPYMWDTONKSCO-UHFFFAOYSA-N 2,2'-piperazine-1,4-diylbisethanesulfonic acid Chemical compound OS(=O)(=O)CCN1CCN(CCS(O)(=O)=O)CC1 IHPYMWDTONKSCO-UHFFFAOYSA-N 0.000 description 1
- 101100458289 Drosophila melanogaster msps gene Proteins 0.000 description 1
- 239000007990 PIPES buffer Substances 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000003990 capacitor Substances 0.000 description 1
- 238000012993 chemical processing Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000002592 echocardiography Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000001747 exhibiting effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000007620 mathematical function Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000000135 prohibitive effect Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000005201 scrubbing Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 231100000331 toxic Toxicity 0.000 description 1
- 230000002588 toxic effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N17/00—Investigating resistance of materials to the weather, to corrosion, or to light
- G01N17/006—Investigating resistance of materials to the weather, to corrosion, or to light of metals
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/22—Details, e.g. general constructional or apparatus details
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/34—Generating the ultrasonic, sonic or infrasonic waves, e.g. electronic circuits specially adapted therefor
- G01N29/348—Generating the ultrasonic, sonic or infrasonic waves, e.g. electronic circuits specially adapted therefor with frequency characteristics, e.g. single frequency signals, chirp signals
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B17/00—Measuring arrangements characterised by the use of infrasonic, sonic or ultrasonic vibrations
- G01B17/02—Measuring arrangements characterised by the use of infrasonic, sonic or ultrasonic vibrations for measuring thickness
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N17/00—Investigating resistance of materials to the weather, to corrosion, or to light
- G01N17/04—Corrosion probes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
- G01N29/043—Analysing solids in the interior, e.g. by shear waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
- G01N29/07—Analysing solids by measuring propagation velocity or propagation time of acoustic waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
- G01N29/11—Analysing solids by measuring attenuation of acoustic waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/22—Details, e.g. general constructional or apparatus details
- G01N29/223—Supports, positioning or alignment in fixed situation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/34—Generating the ultrasonic, sonic or infrasonic waves, e.g. electronic circuits specially adapted therefor
- G01N29/341—Generating the ultrasonic, sonic or infrasonic waves, e.g. electronic circuits specially adapted therefor with time characteristics
- G01N29/343—Generating the ultrasonic, sonic or infrasonic waves, e.g. electronic circuits specially adapted therefor with time characteristics pulse waves, e.g. particular sequence of pulses, bursts
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4409—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by comparison
- G01N29/4418—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by comparison with a model, e.g. best-fit, regression analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4481—Neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/48—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by amplitude comparison
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/01—Indexing codes associated with the measuring variable
- G01N2291/011—Velocity or travel time
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/01—Indexing codes associated with the measuring variable
- G01N2291/015—Attenuation, scattering
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/023—Solids
- G01N2291/0234—Metals, e.g. steel
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/025—Change of phase or condition
- G01N2291/0258—Structural degradation, e.g. fatigue of composites, ageing of oils
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/028—Material parameters
- G01N2291/02854—Length, thickness
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/04—Wave modes and trajectories
- G01N2291/042—Wave modes
- G01N2291/0421—Longitudinal waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/04—Wave modes and trajectories
- G01N2291/042—Wave modes
- G01N2291/0422—Shear waves, transverse waves, horizontally polarised waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/04—Wave modes and trajectories
- G01N2291/042—Wave modes
- G01N2291/0423—Surface waves, e.g. Rayleigh waves, Love waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/04—Wave modes and trajectories
- G01N2291/042—Wave modes
- G01N2291/0425—Parallel to the surface, e.g. creep waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/04—Wave modes and trajectories
- G01N2291/044—Internal reflections (echoes), e.g. on walls or defects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/26—Scanned objects
- G01N2291/263—Surfaces
- G01N2291/2634—Surfaces cylindrical from outside
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
- G01N29/06—Visualisation of the interior, e.g. acoustic microscopy
- G01N29/0609—Display arrangements, e.g. colour displays
- G01N29/0645—Display representation or displayed parameters, e.g. A-, B- or C-Scan
Definitions
- This disclosure relates to the field of corrosion and erosion monitoring of pipes and vessels.
- this disclosure relates to a corrosion and/or erosion monitoring system comprising mechanical components, hardware, software, analytics, and/or a combination thereof.
- the mechanical components and hardware may comprise one or more ultrasonic transducers, base units, gateways, and/or combination thereof.
- the system may further comprise a software platform for remote monitoring.
- the system may further comprise, in some embodiments, analytics tools for front-end services and back-end services for remote monitoring and/or diagnostics.
- this disclosure may relate to a system and method for corrosion and erosion monitoring of pipes and vessels, where the system/method combines ultrasonic thickness monitoring using longitudinal waves with ultrasonic area monitoring using one or more guided waves, whereby representative thickness measurements are complemented by an area monitoring feature to detect localized corrosion/erosion in between representative thickness measurement locations.
- a system and method for optimized asset health monitoring that includes an analytics solution is disclosed.
- ultrasonic transducers for ultrasonically monitoring the condition and integrity of structural assets, including pipes and pressure vessels, such as those used in the oil and gas and power generation industries.
- corrosion and erosion monitoring systems and techniques incorporating/using ultrasonic transducers are known to include thickness monitoring at a location and area monitoring (also known as guided wave inspection).
- these two systems and techniques are typically separate from one another.
- internal corrosion of piping systems is also sometimes monitored using radiographic (RT) thickness testing, in addition to ultrasonic (UT) testing, to measure wall thicknesses for selected components at prescribed intervals, over the life of the system.
- RT radiographic
- UT ultrasonic
- Thickness monitoring ultrasonic transducers and systems utilizing same typically measure a thickness of a pipe/vessel wall at the spot where the ultrasonic transducer is provided—in other words, it does not provide any information regarding the thickness of the pipe/vessel wall at locations surrounding the exact spot where the ultrasonic transducer is provided. As such, if corrosion/erosion is occurring at a location other than where the ultrasonic transducer is provided, it is likely that the corrosion/erosion will not be detected, unless thickness monitoring is accompanied by ultrasonic transducer mapping. Of course, ultrasonic transducer mapping increases the inspection cost. These ultrasonic transducers and systems are, however, beneficially permanently installed on pipes/vessels.
- area monitoring ultrasonic transducers and systems utilizing same typically measure the thickness of a pipe/vessel wall across a larger area of the pipe/vessel wall, which area being measured is typically beyond the location where the thickness monitoring ultrasonic transducers are provided on the pipe/vessel.
- Such area monitoring ultrasonic transducers and systems utilizing same will typically develop a thickness map of the pipe/vessel wall across the area being measured.
- a generated thickness map is beneficial, but at present, such guided wave inspection is extremely complex as general hardware in that segment generates ten to twenty different guided wave modes, and the high number of wave modes and the complex analysis negatively impacts the confidence in the inspection results.
- guided wave inspection is typically not permanently installed on pipes and vessels. Additionally, highly localized corrosion cannot be reliably detected with temporarily installed guided wave systems as described in API 574 (API 574, Inspection practices for piping system components, 4 th edition, 2016).
- piping systems might transport the fluids to one or more tanks and/or chemical processing unit.
- Some piping systems handle dedicated fluids at prescribed temperatures and/or pressures; these piping systems may transfer highly corrosive fluids at elevated temperatures and pressures.
- a system of one or more computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination of them installed on the system that in operation causes or cause the system to perform the actions.
- One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.
- One general aspect includes a method for down-selecting from among probe assemblies installed on a piping system. The method also includes setting a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter for a model, before training the model.
- the method also includes grouping, by the model executing on a processor, a first set of the probe assemblies based at least on historical pipe wall thickness measurements collected from the probe assemblies installed on the piping system over a period of time.
- the method also includes assigning a unique groupID to each set of probe assemblies.
- the method also includes selecting, by the model after training the model, an optimization function from among a plurality of optimization functions for the model.
- the method also includes identifying, by the model, a single probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system.
- the method also includes sending, by a thickness monitoring controller associated with the piping system, a pipe wall thickness measurement of the single probe assembly from each groupID for inspection.
- Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
- Implementations may include one or more of the following features.
- the method may include one or more steps to, during the inspection, disregard all remaining probe assemblies in each groupID except the single probe assembly from each groupID.
- the grouping of the first set of the probe assemblies is further based at least on inspection information provided to the system and historical pipe wall thickness measurements collected over a period of time from the probe assemblies installed on the piping system.
- the piping system may include a tank, and where a first probe assembly of the probe assemblies is configured to measure a wall thickness of the tank.
- the method may also include steps for storing, in computer memory communicatively coupled to the processor, historical pipe wall thickness measurements collected over an extended period of time from the probe assemblies installed on the piping system; and for training, by the processor, the model with at least the historical pipe wall thickness measurements stored in the computer memory.
- One general aspect includes a system for detecting general corrosion (e.g., a lack of localized corrosion) to a plurality of components that transport materials across a distance.
- the system may also include a plurality of probe assemblies affixed to one or more of the components, where the probe assemblies may include at least a thickness monitoring ultrasonic transducer and an area monitoring ultrasonic transducer configured to detect corrosion (e.g., general corrosion and/or localized corrosion) to the components.
- the system may also include a data store configured to store historical wall thickness measurements collected over a period of time from measurements performed by the probe assemblies.
- the system may also include a model trained on the historical wall thickness measurements in the data store and with hyperparamters may include a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter.
- the system may also include a monitoring apparatus may include a processor and a memory storing computer-executable instructions that, when executed by the processor, cause the system to perform steps that may also include: grouping, based on the model, a first set of the probe assemblies; assigning a unique groupid to each set of probe assemblies; selecting, based on the model, an optimization function from among a plurality of optimization functions; identifying, based on the model and selected optimization function, a probe assembly corresponding to each groupid for wall thickness monitoring of the components; and sending, by a thickness monitoring controller associated with the components, a wall thickness measurement of the probe assembly from each groupid for inspection.
- system may output a list of the unique identifiers corresponding to any groupID in lieu of sending the wall thickness measurement for inspection.
- An inspector may receive the system's output and react accordingly, as discussed in various embodiments disclosed herein.
- Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
- Implementations may include one or more of the following features.
- the system where the probe assembly identified from each groupID, may include more than one probe assembly of the plurality of probe assemblies, and where the memory of the monitoring apparatus stores computer-executable instructions that, when executed by the processor, cause the system to perform steps that may include: during the inspection, disregarding all remaining probe assemblies in each groupID except the more than one probe assembly from each groupID; and validating that the wall thickness measurements of the more than one probe assembly from each groupID is general corrosion and not localized corrosion.
- the wall thickness measurement of the probe assembly from a first groupID may include a thickness of a wall of a pipe component at the probe assembly.
- the wall thickness measurement of the probe assembly from a first groupID may include a thickness of a wall of a tank component at the probe assembly.
- the method may include validating that the pipe wall thickness measurement of the single probe assembly is general corrosion (e.g., a lack of localized corrosion) by: (i) generating a probability plot of all pipe wall thickness measurements associated with the piping system, (ii) grouping the plotted pipe wall thickness measurements by nominal thickness, and (iii) identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm the generalized corrosion (e.g., lack of localized corrosion).
- general corrosion e.g., a lack of localized corrosion
- the pipe wall thickness monitoring may include steps for, by the probe assemblies, analyzing the original wall thicknesses, wall thickness loss over time, calibration error, and measurement location repeatability error. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.
- Implementations may include one or more of the following features.
- the method may further include steps for validating that the pipe wall thickness measurement of the single probe assembly is general corrosion (e.g., a lack of localized corrosion) by: generating a probability plot of all pipe wall thickness measurements associated with the piping system, grouping the plotted pipe wall thickness measurements by nominal thickness, and identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm the general corrosion (e.g., the lack of localized corrosion).
- FIG. 1 is an illustration of the system for corrosion/erosion monitoring
- FIG. 2 is an illustration of a thickness monitoring controller and a piezo assembly of the system of FIG. 1 ;
- FIG. 3 is an illustration of the thickness monitoring controller of FIG. 2 ;
- FIG. 4 is an illustration of a switch assembly forming part of the piezo assembly of FIG. 2 ;
- FIG. 5 is an illustration of the piezo assembly of FIG. 2 ;
- FIG. 6 , FIG. 7 , and FIG. 8 are illustrations of the method for corrosion/erosion monitoring
- FIG. 9 , FIG. 10 , FIG. 11 , and FIG. 12 are illustrations to display the signal modulation
- FIG. 13 A and FIG. 13 B are drawings of one illustrative piping with installed MUT sensors in accordance with one or more aspects of the features disclosed herein;
- FIG. 14 is an illustrative network architecture of an industrial facility in accordance with various aspects of the disclosure.
- FIG. 15 is an illustrative diagram of probe assembly groupings in one embodiment of the disclosure.
- FIG. 16 A , FIG. 16 B , and FIG. 16 C (collectively referred to as “ FIG. 16 ”) illustrate plots on a graph.
- FIG. 16 A is a graph illustrating probability plot of measurement values for validating general corrosion in contrast to localized corrosion.
- FIG. 16 B is a graph charting level of risk against TMLs in accordance with various aspect disclosed herein.
- FIG. 16 C illustrates a shift in the curve depicting the level of risk against TMLs after down-selection in accordance with various aspect disclosed herein;
- FIG. 17 is a graph plot of illustrating cumulative thickness distribution for tubes with naphthenic acid corrosion
- FIG. 18 A is a corrosion sensor analytics graph illustrating TML measurements by date in one embodiment of the disclosure.
- FIG. 18 B is another corrosion sensor analytics graph illustrating TML measurements by date as in FIG. 8 A , but with a higher grouping sensitivity setting;
- FIG. 18 C is yet another corrosion sensor analytics graph illustrating TML measurements by date as in FIG. 8 A , but with an even higher grouping sensitivity setting;
- FIG. 19 A and FIG. 19 B are graphs in accordance with one or more aspects of the disclosure.
- FIG. 20 A and FIG. 20 B are also graphs in accordance with one or more aspects of the disclosure.
- FIG. 21 shows an illustrative artificial neural network configured to operate in collaboration with systems, methods, and algorithms disclosed herein;
- FIG. 22 is a flowchart showing illustrative steps of a method performed in accordance with some embodiments disclosed herein;
- FIG. 23 is an illustration of a simplified pipe and instrumentation diagram (PID) corresponding to an illustrative corrosion/erosion monitoring system, as illustrated in FIG. 1 , in accordance with some embodiments disclosed herein.
- PID pipe and instrumentation diagram
- the monitoring system may comprise a software platform for remote monitoring and analytics of historical measurements collected by a plurality of sensors affixed to the pipes and components.
- the monitoring system may include analytics tools for monitoring, diagnostics, and/or prediction of localized corrosion and/or general corrosion.
- the thickness monitoring locations may be optimized to, among other things, reduce the number of measurement locations without compromising risk—i.e., down-selecting. Through down-selecting, by strategically reducing the number of probe assemblies that need to be sampled during an inspection, the amount of time/cost of an inspection is reduced while simultaneously maintaining (or even reducing) the risk profile of the industrial facility, as explained in this disclosure.
- FIG. 1 and FIG. 2 A system 100 for monitoring corrosion and erosion of pipes/vessels is illustrated in FIG. 1 and FIG. 2 .
- the system 100 includes a data analytics and visualization platform 110 , an optional gateway 120 , a thickness monitoring controller 130 , a thickness monitoring ultrasonic transducer 140 that is used for standardization purposes, and at least one probe assembly 150 .
- Each probe assembly 150 includes a switch assembly 160 , at least one thickness monitoring ultrasonic transducer 170 , and at least one area monitoring ultrasonic transducer 180 .
- the data analytics and visualization platform 110 includes a data analytics portion 112 and a visualization portion 114 .
- the data analytics portion 112 is typically a cloud-based powered software that is configured to receive signals, typically wirelessly, from one or both of the gateway 120 or the thickness monitoring controller 130 . These signals are analyzed by the data analytics portion 112 to translate them into visuals for display on the visualization portion 114 .
- the visualization portion 114 may be any suitable device, e.g., a computer monitor, a tablet, a phone, etc., that are of a type that will aid an individual monitoring the platform 110 in understanding the information regarding corrosion/erosion identified by the system 100 . The individual may also be able to change the images/information on the visualization portion 114 by providing further inputs to the software.
- the gateway 120 may be provided to receive signals, typically wirelessly, from the thickness monitoring controller 130 , and to send such signals, typically wirelessly, to the platform 110 .
- the thickness monitoring controllers 130 would use, e.g., the XBee protocol, to communicate with the gateway 120 .
- the gateway 120 could be installed at a higher location to establish cellular connection and the thickness monitoring controller 130 would submit data to the gateway 120 using, for example, the XBee protocol.
- the thickness monitoring controller 130 includes a modem 131 , a microprocessor 132 , a pulser 133 , an analog-to-digital converter (ADC) 134 , an adjustable gain amplifier 135 , a transmit channel 136 , and a receive channel 137 .
- the modem 131 is configured to communicate with one or both of the platform 110 and the gateway 120 .
- the modem 131 may use any appropriate communication option, including, but not limited to XBee 915 MHz and LTE-M/NB.
- the modem 131 is configured to communicate with the microprocessor 132 .
- the microprocessor 132 may be any type of microprocessor which will provide the desired functions.
- One such microprocessor 132 is the LPC4370 that is manufactured and sold by NXP Semiconductors.
- the microprocessor 132 is configured to communicate with both the pulser 133 and the ADC 134 .
- the pulser 133 is preferably a high voltage pulser capacitor.
- the ADC 134 is preferably a 16-bit, 2 msps (million samples per second), but other ADC types may also be provided as appropriate.
- the ADC 134 is configured to communicate with the adjustable gain amplifier 135 (sometimes also commonly known as a variable gain amplifier).
- the adjustable gain amplifier 135 preferably has a decibel range of 26-54 dB and a frequency range of 10 kHZ to 300 kHz, but other ranges may also be provided as appropriate.
- the pulser 133 is configured to communicate with the transmit channel 136 to transmit signals to the transmit channel 136 .
- the adjustable gain amplifier 135 is configured to communicate with the receive channel 137 to receive signals from the receive channel 137 .
- the thickness monitoring controller 130 is preferably configured to accommodate a desired number of amplitude scans (“A-scans”) (or waveform displays). In the embodiments illustrated, the controller 130 is configured to accommodate sixteen A-scans (one from the thickness measurement ultrasonic transducer 140 and five each from the three different probe assemblies 150 ).
- controller 130 can be configured to accommodate more or less than sixteen A-scans as appropriate.
- the thickness monitoring ultrasonic transducer 140 is configured to receive signals from the transmit channel 136 of the thickness monitoring controller 130 and is further configured to transmit signals to the receive channel 137 of the thickness monitoring controller 130 .
- the thickness monitoring ultrasonic transducer 140 is used for standardization purposes and, thus, functions to calibrate the measurement system when a group of ultrasonic transducers are utilized (in this instance, the at least one thickness monitoring ultrasonic transducer 170 , and the at least one area monitoring ultrasonic transducer 180 ).
- the standardization thickness monitoring ultrasonic transducer 140 works to ensure that the system 100 always performs the same way and functions properly, which is required by industrial standards.
- the standardization thickness monitoring ultrasonic transducer 140 is configured to perform a single A-scan.
- the thickness monitoring ultrasonic transducer 140 is typically placed on a standardization block or a thickness calibrated metal piece to serve as a standardization transducer.
- the system 100 includes three different/distinct probe assemblies 150 A, 150 B, 150 C (each also referred to as probe assembly 150 ).
- the number of probe assemblies 150 provided in the system 100 can be less than three (e.g., one or two) or can be more than three (e.g., four, five, etc.), as appropriate.
- minor variations/modifications may need to be made to the system 100 as would be understood by one of ordinary skill in the art.
- each probe assembly 150 includes a switch assembly 160 .
- the switch assembly 160 includes a power supply 161 , a transmit switch 162 , a microcontroller 163 , a memory 164 , a receive switch 165 , an amplifier 166 , and an optional resistance temperature detector (RTD) interface 167 .
- the power supply 161 is in communication with the transmit channel 136 of the thickness monitoring controller 130 .
- the transmit switch 162 is in communication with the transmit channel 136 of the thickness monitoring controller 130 .
- the transmit switch 162 preferably has five “switch” channels 162 a , 162 b , 162 c , 162 d , 162 e , the purpose and function of each will be discussed herein.
- the microcontroller 163 is in communication with the transmit channel 136 of the thickness monitoring controller 130 , the transmit switch 162 , the memory 164 , and the receive switch 165 .
- the microcontroller 163 may be any type of microcontroller which will provide the desired functions.
- One such microcontroller 163 is the PIC18 that is manufactured and sold by Microchip Technology.
- the memory 164 is preferably a non-volatile memory.
- the receive switch 165 preferably has four “switch” channels 165 a , 165 b , 165 c , 165 d , the purpose and function of each will be discussed hereinbelow.
- the amplifier 166 is in communication with the receive channel 137 of the thickness monitoring controller 130 and the receive switch 165 .
- the amplifier 166 preferably has an amplification of 26 to 48 dB and a frequency range of 10 kHz to 300 kHz, but other levels/ranges may also be provided as appropriate.
- the amplifier 166 is also preferably a two-stage amplifier, where 26 dB amplification is provided for a single stage option and 48 dB amplification is provided for a two-stage option, which can be selectable by populating or depopulating components on an amplification board.
- the optional RTD interface 167 is provided if the at least one thickness monitoring ultrasonic transducer 170 incorporates an RTD 171 (as discussed below). In the illustrated embodiment, each switch assembly 160 is instructed by controller 130 to collect five A-scans (one from the thickness monitoring ultrasonic transducer 170 and one from each of the four area monitoring ultrasonic transducers 180 ).
- each probe assembly 150 includes at least one thickness monitoring ultrasonic transducer 170 . As illustrated in FIG. 1 , each probe assembly 150 includes one thickness monitoring ultrasonic transducer 170 . Depending on the system 100 and the probe assembly 150 , the number of thickness monitoring ultrasonic transducers 170 provided in each probe assembly 150 can be more than one (e.g., two, three, four, etc.), as appropriate. Depending on the number of thickness monitoring ultrasonic transducers 170 provided in each probe assembly 150 , minor variations/modifications may need to be made to the probe assembly 150 and/or system 100 as would be understood by one of ordinary skill in the art.
- Each thickness monitoring ultrasonic transducer 170 may optionally have an RTD 171 associated therewith to measure the temperature of the pipe/vessel at or near where the thickness measurement is occurring.
- Each thickness monitoring ultrasonic transducer 170 is in communication with the fifth “switch” channel 162 e of the transmit switch 162 and, if the thickness monitoring ultrasonic transducer 170 includes the RTD 171 , is also in communication with the RTD interface 167 .
- the thickness monitoring ultrasonic transducer 170 (as well as the 140 ) operates by generating high frequency ultrasonic waves (e.g., 5 MHz). These ultrasonic waves are commonly referred to as longitudinal waves (LW) and, as such, the thickness monitoring ultrasonic transducers 170 may also be referred to as LW transducers. In the illustrated embodiment, each thickness monitoring ultrasonic transducer 170 is configured to perform a single A-scan. Unlike the thickness monitoring ultrasonic transducer 140 , the thickness monitoring ultrasonic transducer 170 is not placed on a standardization block or a thickness calibrated metal piece, but rather is placed on the pipe/vessel to measure the thickness of the pipe/vessel at the location where it is installed.
- LW longitudinal waves
- each thickness monitoring ultrasonic transducer 170 is configured to perform a single A-scan.
- the thickness monitoring ultrasonic transducer 170 is not placed on a standardization block or a thickness calibrated metal piece, but rather is placed on the pipe
- each probe assembly 150 includes at least one area monitoring ultrasonic transducer 180 .
- each probe assembly 150 includes four area monitoring ultrasonic transducers 180 A, 180 B, 180 C, 180 D (each also referred to as area monitoring ultrasonic transducer 180 ).
- the number of area monitoring ultrasonic transducers 180 provided in each probe assembly 150 can be less than four (e.g., one, two or three) or more than four (e.g., five, six, etc.), as appropriate.
- the first area monitoring ultrasonic transducer 180 A is in communication with the first “switch” channel 162 a of the transmit switch 162 and the first “switch” channel 165 a of the receive switch 165 .
- the second area monitoring ultrasonic transducer 180 B is in communication with the second “switch” channel 162 b of the transmit switch 162 and the second “switch” channel 165 b of the receive switch 165 .
- the third area monitoring ultrasonic transducer 180 C is in communication with the third “switch” channel 162 c of the transmit switch 162 and the third “switch” channel 165 c of the receive switch 165 .
- the fourth area monitoring ultrasonic transducer 180 D is in communication with the fourth “switch” channel 162 d of the transmit switch 162 and the fourth “switch” channel 165 d of the receive switch 165 .
- the probe assembly 150 may comprise a thickness transducer 170 and a set of area transducers 180 individually wired to switch/preamp assembly 160 .
- thickness and area transducers 170 , 180 can be combined in a single, larger probe wired via a single multiconductor cable into switch/preamp assembly 160 .
- it also can be a set of larger probes (thickness+2 area, area+area etc.)
- the area monitoring ultrasonic transducers 180 operate by generating low frequency ultrasonic waves (e.g., 50 kHz to 500 kHz). These ultrasonic waves are commonly referred to as guided waves (GW) and, as such, the area monitoring ultrasonic transducers 180 may also be referred to as GW transducers.
- GW guided waves
- each area monitoring ultrasonic transducer 180 is configured to perform a single A-scan.
- the GW transducers 180 are preferably in the form of piezo patch transducers, but may alternatively be in other forms, such as, for instance, face-shear piezo elements.
- the GW transducers 180 A, 180 B, 180 C, 180 D are positioned in a rectangular configuration around the LW transducer 170 , where GW transducer 180 A is positioned above and to the left of LW transducer 170 , GW transducer 180 B is positioned below and to the left of LW transducer 170 , GW transducer 180 C is positioned below and to the right of LW transducer 170 , and GW transducer 180 D is positioned above and to the right of LW transducer 170 .
- a straight line from GW transducer 180 A to GW transducer 180 B is parallel to a straight line from GW transducer 180 C to GW transducer 180 D
- a straight line from GW transducer 180 A to GW transducer 180 D is parallel to a straight line from GW transducer 180 B to GW transducer 180 C.
- a straight line from GW transducer 180 A to GW transducer 180 C intersects LW transducer 170
- a straight line from GW transducer 180 B to GW transducer 180 D intersects LW transducer 170 , such that an “X-shape” configuration is provided.
- the system 100 when associated with a pipe/vessel, may be utilized to measure the corrosion/erosion of the pipe/vessel.
- one method 200 of measuring the corrosion/erosion of the pipe/vessel is described below and illustrated in FIG. 6 , FIG. 7 , and FIG. 8 .
- the method 200 includes the step 205 of manually measuring the actual longitudinal velocity and the temperature of the pipe/vessel to be inspected.
- the method 200 includes the step 210 of manually measuring the actual guided wave velocity and the temperature of the pipe/vessel to be inspected.
- the method 200 includes the step 215 of performing a thickness standardization measurement with the standardization thickness monitoring ultrasonic transducer 140 and the RTD 171 (it is to be understood that, like the thickness monitoring ultrasonic transducer 170 , the standardization thickness monitoring ultrasonic transducer 140 could also optionally incorporate the RTD 171 ).
- the method 200 includes the step 220 of performing measurements using the probe assembly 150 A.
- Step 220 includes the sub-step 220 a of performing a thickness measurement with the thickness monitoring ultrasonic transducer 170 and the RTD 171 .
- Step 220 includes the sub-step 220 b of performing an area thickness monitoring with the area monitoring ultrasonic transducers 180 A, 180 B, 180 C, 180 D at a first frequency.
- Sub-step 220 b includes the sub-step 220 b 1 of performing axial scanning whereby area monitoring ultrasonic transducer 180 A is excited and data is recorded with area monitoring ultrasonic transducer 180 B. The measurement taken in sub-step 220 b 1 is repeated as often as specified in configuration setting and average A-scans.
- Sub-step 220 b includes the sub-step 220 b 2 of performing axial scanning whereby area monitoring ultrasonic transducer 180 C is excited and data is recorded with area monitoring ultrasonic transducer 180 D. The measurement taken in sub-step 220 b 2 is repeated as often as specified in configuration setting and average A-scans.
- Sub-step 220 b includes the sub-step 220 b 3 of performing circumferential scanning whereby area monitoring ultrasonic transducer 180 A is excited and data is recorded with area monitoring ultrasonic transducer 180 D. The measurement taken in sub-step 220 b 3 is repeated as often as specified in configuration setting and average A-scans.
- Sub-step 220 b includes the sub-step 220 b 4 of performing circumferential scanning whereby area monitoring ultrasonic transducer 180 C is excited and data is recorded with area monitoring ultrasonic transducer 180 C.
- the measurement taken in sub-step 220 b 4 is repeated as often as specified in configuration setting and average A-scans.
- channels 162 a , 162 c (which are associated with GW transducers 180 A, 180 C) act as guided wave transmit channels while channels 162 b , 162 d (which are associated with GW transducers 180 B, 180 D) act as guided wave receive channels.
- the receive path further goes via the amplifier 166 to the receive channel 137 of the thickness monitoring controller 130 .
- Step 220 includes the sub-step 220 c of repeating sub-step 220 b at a second frequency, which second frequency is different from the first frequency.
- Step 220 includes the sub-step 220 d of repeating sub-step 220 b at a third frequency, which third frequency is different from both the first frequency and the second frequency.
- the method 200 includes the step 225 , which comprises repeating step 220 to perform measurements using the probe assembly 150 B.
- the method 200 includes the step 230 , which comprises repeating step 220 to perform measurements using the probe assembly 150 C.
- the method 200 combines ultrasonic thickness monitoring using longitudinal waves with ultrasonic area monitoring using guided waves and, in a preferred embodiment, just one special non-dispersive shear wave mode (SH 0 or T(0,1)).
- the method 200 takes representative thickness measurements, rather than trying to develop a thickness map, which will be complemented by an area monitoring feature to detect localized corrosion/erosion in-between representative thickness measurement locations.
- the system 100 utilizes new electronics which use a single circuitry to deliver two distinctive, different excitation signals, e.g., high frequency ultrasonic waves for thickness monitoring (5 MHz) and low frequency ultrasonic waves for area monitoring (50-500 kHz), from two different types of ultrasonic transducers, e.g., LW transducer 170 and GW transducers 180 .
- Each excitation signal needs to be generated and processed differently.
- pulser 133 of the controller 130 is a digital switch capable of delivering only predetermined fixed voltage levels: high voltage, low voltage and zero voltage. High and low voltage levels are normally adjustable in a range of 5V to 90V and ⁇ 5V to ⁇ 90V but different voltage levels are permissible as well.
- Microprocessor 132 signals pulser 133 to output to transmit channel 136 one of the fixed voltage levels: ex. high voltage for a specified period of time.
- Example of a pulse used to excite LW transducer 170 processor 130 instructs pulser 133 to output 0V, then high voltage for a period of 100 ns, then low voltage for a period of 100 ns, then 0V. Described sequence would generate bipolar square wave of 5 MHz frequency suitable to excite LW transducer 170 .
- GW transducers 180 different frequencies and signals amplitudes are required.
- waveforms needed to excite GW transducers 180 can have rather complex shapes like, ex: 5 cycle sinusoid wave superimposed on Hanning window signal (ex. half cycle cosine) shown as 330 in FIG. 12 that would allow for a smoother transition from no-signal to signal condition.
- suitable waveforms combination of a pulser 133 digital output shown as waveform 300 in FIG. 9 , FIG. 10 , FIG. 11 , and FIG. 12 in-series resistance of the transmit channel 136 and impedance of the GW transducer 180 are used.
- GW transducer 180 impedance in a frequency range used to generate GW waves is usually in majority composed of capacitance. This capacitance and mentioned in-series resistance of the transmit channel 136 form a low pass filter. Pulser 133 under instructions from the microprocessor 132 generates a high frequency (usually in range of tens of MHz) digital waveform 300 that when passed thru the transmit channel 136 and GW transducer 180 capacitance results in a different waveform 310 than originally outputted from the pulser 133 (as illustrated in FIG. 10 ).
- Varying high frequency digital waveforms from the pulser 133 can generate, once passed thru the transmit channel 136 in-series resistance and transducer 180 capacitance, a range of analog waveforms, ex: sinusoids without Hanning windows, shown as 320 in FIG. 11 or sinusoids with Hanning windows, shown as 330 in FIG. 12 , chirp (frequency changes during duration of the pulse), ramp-up, seesaw and other.
- sinusoids without Hanning windows shown as 320 in FIG. 11
- sinusoids with Hanning windows shown as 330 in FIG. 12
- chirp frequency changes during duration of the pulse
- a chirp signal can be used to excite multiple frequencies at the same time from a single channel
- Proper software filtering can decode the individual frequency response from a single A-scan.
- the time of flight and the amplitude of the echo reflected at a defect on the pipe/vessel can be evaluated. More specifically, by sending excitation signals from GW transducer 180 C and receiving by GW transducers 180 B, 180 D, the reflection echo will be earlier in time trace in the GW transducer 180 B, 180 D that is closer to the damage, e.g., GW transducer 180 B if the damage is to the left of both GW transducers 180 B, 180 D, or GW transducer 180 D if the damage is to the right of both GW transducers 180 B, 180 D, where GW transducers 180 B, 180 D are positioned as illustrated in FIG. 1 and FIG. 5 .
- Defects as pittings or corrosion/erosion patches usually increase in size over time. Therefore, the amplitude of the echoes reflected at the defects will increase over time. Permanently installed systems therefore allow one to monitor the change of amplitude next to the time-of-flight. Monitoring changes in A-Scans after for example baseline subtraction and digital filtering reduces the complexity of the analysis and increases confidence in the inspection results. Next to baseline subtraction additional digital signal processing tools or machine learning algorithms can be used for feature extraction or pattern recognition which additionally increase confidence levels and help to detect changes earlier in time.
- FIG. 23 illustrates a simplified pipe and instrumentation diagram (PID) corresponding to an illustrative corrosion/erosion monitoring system, as illustrated in FIG. 1 , in accordance with some embodiments disclosed herein.
- the simplified PID 2300 includes numerous probe assemblies depicted as circles numbered nineteen to eighty-four. For example, three different/distinct probe assemblies 150 A, 150 B, 150 C are illustrated. Of course, the number of probe assemblies in the PID 2300 can be any number, as appropriate.
- a human operator/inspector may focus the inspection on a down-selected list of TMLs, as explained herein.
- These down-selected TMLs may represent more efficient candidate measuring locations to capture general corrosion behavior of the entire asset, while still being able to inspect for localized corrosion. For example, substantial amount of time/energy and cost may be saved by down-selecting the number of TMLs so that only those probe assemblies with the highest probability of detecting localized corrosion are examined by the human operator/inspector. Rather than checking all of probe assemblies nineteen to eighty-four, or even randomly checking less than all of probe assemblies nineteen to eighty-four, the down-selected TMLs are a more optimal identification of which TMLs to measure. In some examples, the inspector may use a handheld or other manual device to measure wall thickness at the numbered locations on the simplified PID 2300 .
- a rig or harness of sorts may be pre-installed at the numbered location on the simplified PID 2300 to allow the inspector to measure wall thickness at each thickness measuring location.
- the inspector may be an automated machine that takes measurements at the down-selected TMLs at particular time intervals. Even in an automated measuring system, down-selecting TMLs is advantageous because it reduces the amount of processing power and network bandwidth consumed by measurement data generated by a measuring device at each numbered location on the simplified PID 2300 . For example, some large industrial facilities may have thousands upon thousands of probe assemblies that could result in a prohibitive amount of generated data.
- a human operator may indicate as much so that any model can be updated to reflect the new wall thickness values.
- supervised human input into a machine learning or neural network which is executing in a digital analytics platform, may refine its alerts and model accordingly.
- FIG. 13 A illustrates an illustrative piping with sensors 1301 installed on the pipe in accordance with one or more aspects of the features disclosed herein.
- the pipe may have a flow of liquid in the direction depicted by the arrows.
- one approach may be to inspect and take measurements from each and every sensor 1 to 6 depicted in FIG. 13 A .
- a random selection of sensors may be inspected and measured.
- the plurality of thickness monitoring locations (TMLs) shown at each sensor 1 to 6 may be intelligently considered and a smaller/narrower set of TMLs may be down-selected for inspection.
- TMLs thickness monitoring locations
- the TMLs may be grouped based on one or more criteria in the process of down-selecting the TMLs.
- the down-selecting criteria may, in one simplified example, identify and exclude those sensors (e.g., sensors 1 and 3 ) that historically measured only general corrosion in its area.
- sensors e.g., sensors 1 and 3
- grouping to down-select some sensors as being superfluous to the assessment of the health of the mechanical component.
- some prior systems attempted to reduce risk by adding more TMLs and inspections of those TMLs.
- an RBI approach may also use a model that takes into consideration other criteria such as the type of fluid being transported in the piping system, the temperature inside and outside of the pipes/components, elbow/configuration of the piping components, and other criteria.
- the measurements at an elbow may be weighted to be more likely to be selected as part of down-selecting in a group because historically, the locations near an elbow in piping is a place that will have more turbulence and friction, thus a possibility of higher corrosion and acidity.
- probe assemblies 1302 may comprise a tethered device that captures accurate spot measurements of thickness of components.
- probe assemblies may comprise a tethered device that captures accurate spot measurements and area monitoring.
- the device in FIG. 13 B or comparable devices may be used to capture area monitoring of the thickness of a pipe component.
- the probe assembly may comprise a wireless device that captures accurate spot measurements without necessarily being in direct contact with a piping component that requires thickness monitoring.
- the probe assemblies may comprise one or more of thickness monitoring ultrasonic transducers, area monitoring ultrasonic transducers, and/or a combination thereof that are configured to validate general corrosion (e.g., confirm no detection of localized corrosion) in the piping system.
- FIG. 13 B is a drawing of an illustrative piping with installed sensors.
- the sensors 1302 may be any of various types of sensors configured to measure a thickness of the piping at or near the vicinity of the point of its installation on the pipe.
- the sensors 1302 are typically installed in a permanent location and remains affixed to the pipe for an extended period of time (e.g., for the lifespan of that circuit of the piping, for over five years, for over three years, or other period of time).
- the sensors 1302 displayed in FIG. 13 B are installed to the outside of the piping and tethered with wires, in some examples in accordance with one or more aspects of the disclosure, the sensors may be untethered and wirelessly communicate data to one or more wireless receiver/transceiver devices.
- the sensors displayed in FIG. 13 B are illustrated in a straight linear pattern along the longitude of the pipe, the disclosure contemplates sensors installed in any of several different patterns.
- the density of installed sensors may be based on the direction of gravity and the type of substance being transported in the piping.
- the piping in FIG. 13 B is transporting a liquid along the length of pipe from the left to the right when the bottom of the pipe is the portion of the pipe on which sensor 1302 is installed.
- the sensors installed on the piping may be distributed around the circumference of the piping taking into consideration that climate conditions (e.g., rain, hail, sun) may expose portions of the pipe to greater possibility of deterioration while internal conditions in the piping (e.g., more liquid contacts the bottom of the pipe than the top of the pipe) may expose inner portions of the pipe to greater possibility of deterioration.
- climate conditions e.g., rain, hail, sun
- internal conditions in the piping e.g., more liquid contacts the bottom of the pipe than the top of the pipe
- FIG. 14 is an illustrative network architecture of an industrial facility with sensors, communication components, and other components in accordance with various aspects of the disclosure.
- the data analytics platform 112 may be communicatively coupled over a network, such as a local area network 1408 , to one or more networked components.
- the data analytics platform 112 may output to a visualization platform 114 for generation of one or more of the illustrative graphs included herein.
- a monitoring system may comprise the software platform 112 to remotely monitor and analyze historical measurements collected by a plurality of sensors affixed to the pipes and components.
- the monitoring system may include analytics tools for monitoring, diagnostics, and/or prediction of areas that are candidates for localized corrosion (e.g., because the system was unable to confirm general corrosion to the area).
- the TML may be optimized to, among other things, reduce the number of measurement locations without compromising risk—i.e., down-selecting.
- the data analytics platform 112 may trigger an alert to be generated at a remote alert device 1410 .
- the remote alert device 1410 may result in an immediate inspection of one or more components, or may result in particular piping components being prioritized for a subsequent inspection of the facility.
- the data may be stored in a data store 1406 that is communicatively coupled and accessible to the data analytics platform 112 .
- the data may be stored in computer memory 1404 , however, the amount of computer memory required may be high.
- a model 1412 such as a machine learning artificial neural network, may be stored at the computer memory 1404 for execution by a processor 1402 , while historical data and other data may be stored at a data store 1406 .
- the data store may be moved into the platform 112 although it is shown for illustrative purposes as communicating over the local area network 1408 with the platform 112 .
- FIG. 15 is an illustrative diagram of a plurality of sensor (e.g., probe assembly) groupings in one embodiment of the disclosure.
- Each probe assembly may be assigned a unique TML identifier (TML ID) as illustrated in FIG. 15 .
- TML ID may be any unique letter, character, or other identifier that uniquely identifies each TML (i.e., probe assembly).
- the thick-lined rectangular box around select TML ID numbers shows probe assembly groupings.
- the system has grouped probe assemblies 4 , 5 , 6 , and 7 into one grouping based on or more rules.
- 1504 on Mar.
- the graphical representation of the data stored in the computer memory 1404 shows that the system 1400 has adjusted the grouping to include/exclude one or more TMLs.
- the model may recommend that the probe assembly corresponding to TML ID number four should no longer be a part of the groupID corresponding to the thick-lined rectangular box in 1504 .
- the one or more probe assembly down-selected for that groupID may also change.
- the graphical depiction shows that the system 1400 has further adjusted the grouping to now group probe assemblies 5 and 6 into a first groupID and probe assemblies 7 and 8 to a different/separate second groupID.
- the down-selecting and risk profile will change for the overall system 100 .
- the grouping of TMLs into a groupID may be done in one of several different methods.
- the initial grouping for each circuit of components at a facility may be based on the measurement data level. For every date on which measurements were taken by the probe assemblies, a new group may be triggered if a probe assembly satisfies any of the following conditions: (i) if the probe assembly is the first TML of the circuit; (ii) if the (absolute) difference between the measurement value and the preceding TML's measurement value is greater than about 0.5 to about 3.0 standard deviation of all measurements for that date, then the value of this parameter may be reduced for more conservative grouping, or increased for more aggressive grouping; (iii) if the TML's nominal wall thickness measurement is different as compared to the preceding TML's nominal wall thickness measurement; or if the TML has only one measurement historically (across all dates).
- the grouping of TMLs may be done in a multi-step process.
- a first step all measurements taken in a group of connected components (e.g., a circuit) on a particular date (or any other predefined time period—e.g., within a one-hour window of time, within the same week, or other) may be compared to determine how many pairs (or tuples) were measured on the particular date.
- any TML pairs that have less than a predetermined percentage e.g., 70%, 80%, 60%, 75%, or other percent
- the minimum measurement value of all the TMLs may be identified and all TMLs that were paired in an earlier (e.g., first) step with that TML are assigned to the same groupID.
- Other examples of rules for grouping the TMLs would be apparent to a person having skill in the art after review of the entirety disclosed herein.
- TML pairs that are grouped together at least a predetermined threshold percentage of times may be retained in the same group, but TMLs that do NOT meet this threshold may be individually assigned to separate groups using one or more rules.
- measurement dates that do not have sufficient TMLs may be dropped.
- the system 1400 may consider, in some examples, only those measurement dates which have at least a predetermined threshold percentage of the maximum number of TMLs for any date. TMLs that appear in dates that do not meet this threshold may be individually assigned to separate groups.
- the system 1400 may discard (e.g., drop) seemingly invalid measurements based on a lack of historical data, and proceed to re-group TMLs based one or more of the rules described herein.
- the thresholds used are hyperparameters that can be adjusted based on data set diversity and quality. This adjustment may occur at the end of the process upon data confirmation and validation. In one example, threshold percentage may be set to 75%, but with some TMLs the prior measurement might not have occurred in the past, many years.
- a hyper-grid may be generated and used to adjust the parameters and/or hyperparameters of the system 1400 .
- the threshold setting may be strongly correlated to how many TML measurements a system 1400 has collected for each TML ID. Thus, the threshold may be adjusted up or down based on how much data is made available to the system 1400 .
- FIG. 16 and FIG. 17 show graph plots of various data collected and/or analyzed by the system 1400 .
- FIG. 16 A shows a probability plot graph of measurements values (normal is 95%) where percentage is on the Y-axis and measurement value is on the X-axis.
- the system 1400 defaults to assuming that general corrosion has been detected, except when the plot shows that the tail is not running vertical, such as shown near the top of the graph in FIG. 16 A .
- the data analytics platform 112 may validate that the pipe wall thickness measurement of the probe assembly is general corrosion and not localized corrosion by performing one or more steps.
- the validating may be performed by generating a probability plot of all pipe wall thickness measurements associated with the piping system, then grouping the plotted pipe wall thickness measurements by nominal thickness, and identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm that the corrosion is likely not general corrosion. Meanwhile, where the graph shows a linear relationship, then the TMLs corresponding to those data points in the graph are exhibiting general corrosion. This approach is an advancement over systems that may have used standard deviation to build normal probability plots. Moreover, the validating step adds further assurance that the system 1400 is accurately detecting general corrosion and acting accordingly to down-select the appropriate probe assemblies installed on the components in the facility.
- the system 1400 should not generate an alert (e.g., from device 1410 ) for general corrosion because general corrosion is pervasive and is typically not of primary interest during inspections. Rather, general corrosion is accounted for in the scheduling and planning for bulk replacement of components in a facility.
- FIG. 16 B and FIG. 16 C those graphs illustrate the relationship between a risk of mis-identifying general corrosion and the quantity of thickness measurement locations (TMLs).
- TMLs thickness measurement locations
- FIG. 16 B shows that the level of risk charted against the quantity of probe assemblies (i.e., TMLs) decreases as more TMLs are added.
- FIG. 16 C illustrates a shift in the curve depicting the level of risk charted against the quantity of TMLs after down-selection.
- FIG. 16 B and FIG. 16 C are described in more detail below in conjunction with the method steps illustrated in the flowchart of FIG. 22 .
- FIG. 17 is a graph illustrating cumulative thickness distribution for tubes with naphthenic acid corrosion in existing systems known in the art.
- FIG. 18 A is a corrosion sensor analytics graph illustrating TML measurements by date for a specific circuit ID (or asset ID).
- the X-axis corresponds to TML identifiers.
- the probe assemblies installed on a piping system may be assigned identifiers in a sequential or otherwise ordered sequence along the circuit formed by the piping system.
- Each TML might have an ID that shows its position upstream or downstream on the pipe. Other data cleaning and/or scrubbing of the TMLs based on positional data may be performed to harmonize/standardize the measured data for analysis.
- Each TML might be assigned a nominal thickness from when the pipe was first installed.
- One or more publicly available databases e.g., Meridian database
- measurements may be taken over a period of time so that historical data spanning at least a few years (i.e., an extended period of time) may be stored and analyzed.
- almost twenty-five years of wall thickness measurement data is stored, analyzed, and plotted in FIG. 18 A .
- Graph plot 1802 in FIG. 18 A corresponds to measurements taken on 2015 Aug. 3.
- the other plots in the graph correspond to thickness measurements taken for each TML on the corresponding date spanning back almost twenty-five years (e.g., an extended period of time).
- the data analytics platform 112 may set one or more hyperparameter for the model 1412 corresponding to the graph plotted in FIG. 18 A .
- a hyperparameter is typically set before the training/learning process begins on a model; in contrast, the values of other parameters are derived through training of the model.
- FIG. 18 A a graphical user interface for adjusting the grouping_sensitivity hyperparameter is displayed at the top.
- the visual platform 114 may include a graphical tool/slider through which the hyperparameter may be adjusted.
- the grouping_sensitivity hyperparameter is shown set to a “standard” setting.
- FIG. 18 B which shows another illustration of the model 1412
- the grouping_sensitivity hyperparameter is shown set to a “medium” setting.
- the number of groups is only sixty-one in FIG. 18 B instead of ninety-seven groups in FIG. 18 A .
- the graph plotted 1812 in FIG. 18 B is slightly different than the graph 1802 in FIG. 18 A due to the change in hyperparameter settings and TML selection methods.
- the graph plotted 1822 in FIG. 18 C is even more different from FIG. 18 A and FIG. 18 B .
- the number of groups is about seventy-five while the total number of TMLs remains constant at one hundred fifty-five.
- the grouping_sensitivity hyperparameter refers to the sensitivity or aggressiveness of TML grouping, and may be applied at the initial grouping stage.
- a TML may be assigned to a new group when the (absolute) difference between the measurement value and the preceding TML's measurement value is greater than 1 standard deviation (SD) of all measurements for that date.
- SD standard deviation
- This threshold can be adjusted for more conservative or aggressive grouping. A threshold less than 1 SD will result in the grouping being more sensitive to changes in measurements and will lead to a more conservative grouping. On the other hand, a threshold greater than 1 SD will cause the grouping being less sensitive to changes in measurements and will lead to more aggressive grouping (e.g., higher grouping ratios).
- five different grouping sensitivities may be implemented, as shown in FIG. 18 C , in decreasing sensitivity—from most conservative to most aggressive as follows: High (0.5 SD), Standard (1 SD), Medium (1.5 SD), Low (2 SD), and Very Low (3 SD).
- more or less than the aforementioned five groupings may be used to provide more granular or coarse sensitivity.
- the grouping_sensitivity hyperparameter is applied at the initial grouping stage, as is the case hyperparameters, all subsequent grouping steps may be re-run based on the initial grouping results—i.e., the entire grouping cycle is repeated five times, once for each of the five grouping sensitivity levels.
- FIG. 18 A , FIG. 18 B , and FIG. 18 C (collectively referred to as “ FIG. 18 ”) list a plurality of TML selection methods that may be applied to the measurement to optimize the grouping and plotting of the data points.
- FIG. 18 lists three optimization functions—namely median_TML_within_groupID, minimum_average_TML_within_groupID, and minimum_variation_from_mean—other optimization functions may be used in accordance with one or more aspects of the disclosure.
- a TML_position optimization function may be used where if one TML is to be selected, the TML at the center of the group is chosen.
- the optimization function may be a minimum_average_TML_within_groupID optimization function.
- minimum_average_TML_within_groupID method for deciding which TML(s) to pick from each group, the method selects the TML(s) having the lowest average measurement within each group (across dates).
- the system may calculate average measurement of each TML (across dates), rank TMLs in each group by (e.g., ascending) average measurement, and based on number of TMLs to be picked (n) from each group, pick first n TMLs.
- the median_TML_within_groupID optimization function is similar to the minimum_average_TML_within_groupID optimization function, but based on the median instead of the minimum average.
- the optimization function may be a minimum_variation_from_mean optimization function.
- a minimum_variation_from_mean method for deciding which TML(s) to pick from each group the method selects the TML(s) having the lowest average variation from the mean measurement of the group. For example, in one illustrative system using the minimum_variation_from_mean optimization function, the system may calculate the mean group measurement for each date. Then, for each TML, for each date calculate the absolute difference from mean, and for each TML, calculate the average variation (e.g., absolute difference from mean). Next, the minimum_variation_from_mean optimization function ranks TMLs in each group by (e.g., ascending) average variation, and based on number of TMLs to be picked (n) from each group, pick first n TMLs.
- the system 1400 determines a TML candidate selection method and a desired number of probe assemblies per group.
- the number of candidates per group may be another hyper-parameter.
- the system 1400 may use the greater of 1% or one for each TML group. If more than one TML candidate is to be selected, then the TML group may be divided into equally big sub-groups while preserving the TML ordering. Then, the system 1400 may apply a TML candidate selection method based on the one or more scenarios described herein.
- FIG. 19 A is a graph 1902 showing the measured TC thickness in millimeters of a component over time.
- the graph may show the values of temperature calibration (in Celsius), temperature coefficient (e.g., 1%), corrosion rate ST (in millimeters per year), corrosion rate LT (in millimeters per year), remaining life of the component (in years), remaining half life (also in years), and the actual thickness (in millimeters).
- FIG. 19 B is a rectified graph showing FSH (in percentage) values over thickness values (in millimeters or other units).
- the graph also illustrates the thickness range of Gate A 1904 and Gate B 1906 .
- the graph may chart the mV value as a substitute for FSH.
- the graph may be displayed as HF instead of rectified.
- FIG. 20 A is a graph 2002 of an acid battery facility showing the measured TC thickness in millimeters of a component over time.
- the graph may show the values of temperature calibration (in Celsius), temperature coefficient (e.g., 1%), corrosion rate ST (in millimeters per year), corrosion rate LT (in millimeters per year), remaining life of the component (in years), remaining half life (also in years), and the actual thickness (in millimeters).
- FIG. 20 B is a rectified graph of an acid battery facility showing FSH (in percentage) values over thickness values (in millimeters or other units). The graph also illustrates the thickness range of Gate A 2004 and Gate B 2006 . Alternatively, the graph may chart the mV value as a substitute for FSH. Moreover, in some examples, the graph may be displayed as HF instead of rectified.
- FIG. 21 illustrates a simplified example of an artificial neural network 2100 on which a machine learning algorithm may be executed.
- FIG. 21 is merely an example of nonlinear processing using an artificial neural network; other forms of nonlinear processing may be used to implement a machine learning algorithm in accordance with features described herein.
- each of input nodes is connected to a first set of processing nodes.
- the external source 2102 which is fed into input nodes may be the metrics from the results through the steps of the methods disclosed herein.
- Each of the first set of processing nodes is connected to each of a second set of processing nodes.
- Each of the second set of processing nodes is connected to each of output nodes. Though only two sets of processing nodes are shown, any number of processing nodes may be implemented. Similarly, though only four input nodes, five processing nodes, and two output nodes per set are shown in FIG. 21 , any number of nodes may be implemented per set. Data flows in FIG.
- data may be input into an input node, may flow through one or more processing nodes, and may be output by an output node.
- Input into the input nodes may originate from an external source 2102 .
- Output 2104 may be sent to a feedback system 2106 and/or to data store.
- the feedback system 2106 may send output to the input nodes for successive processing iterations with the same or different input data.
- the system may use machine learning to determine an output.
- the output may include a leak area boundary, a multi-sensor detection event, confidence values, and/or classification output.
- the system may use an appropriate machine learning model including xg-boosted decision trees, auto-encoders, perceptron, decision trees, support vector machines, regression, and/or a neural network.
- the neural network may be an appropriate type of neural network including a feed forward network, radial basis network, recurrent neural network, long/short term memory, gated recurrent unit, auto encoder, variational autoencoder, convolutional network, residual network, Kohonen network, and/or other type.
- the output data in the machine learning system may be represented as multi-dimensional arrays, an extension of two-dimensional tables (such as matrices) to data with higher dimensionality.
- the neural network may include an input layer, a number of intermediate layers, and an output layer. Each layer may have its own weights.
- the input layer may be configured to receive as input one or more feature vectors described herein.
- the intermediate layers may be convolutional layers, pooling layers, dense (fully connected) layers, and/or other types.
- the input layer may pass inputs to the intermediate layers.
- each intermediate layer may process the output from the previous layer and then pass output to the next intermediate layer.
- the output layer may be configured to output a classification or a real value.
- the layers in the neural network may use an activation function such as a sigmoid function, a Tanh function, a ReLu function, and/or other functions.
- the neural network may include a loss function.
- a loss function may, in some examples, measure a number of missed positives; alternatively, it may also measure a number of false positives.
- the loss function may be used to determine error when comparing an output value and a target value. For example, when training the neural network, the output of the output layer may be used as a prediction and may be compared with a target value of a training instance to determine an error. The error may be used to update weights in each layer of the neural network.
- the neural network may include a technique for updating the weights in one or more of the layers based on the error.
- the neural network may use gradient descent to update weights.
- the neural network may use an optimizer to update weights in each layer.
- the optimizer may use various techniques, or combination of techniques, to update weights in each layer.
- the neural network may include a mechanism to prevent overfitting—regularization (such as L1 or L2), dropout, and/or other techniques.
- the neural network may also increase the amount of training data used to prevent overfitting.
- FIG. 21 depicts nodes that may perform various types of processing, such as discrete computations, computer programs, and/or mathematical functions implemented by a computing device.
- the input nodes may comprise logical inputs of different data sources, such as one or more data servers.
- the processing nodes may comprise parallel processes executing on multiple servers in a data center.
- the output nodes may be the logical outputs that ultimately are stored in results data stores, such as the same or different data servers as for the input nodes.
- the nodes need not be distinct. For example, two nodes in any two sets may perform the exact same processing. The same node may be repeated for the same or different sets.
- Each of the nodes may be connected to one or more other nodes.
- the connections may connect the output of a node to the input of another node.
- a connection may be correlated with a weighting value. For example, one connection may be weighted as more important or significant than another, thereby influencing the degree of further processing as input traverses across the artificial neural network.
- Such connections may be modified such that the artificial neural network 2100 may learn and/or be dynamically reconfigured.
- nodes are depicted as having connections only to successive nodes in FIG. 21 , connections may be formed between any nodes.
- one processing node may be configured to send output to a previous processing node.
- Input received in the input nodes may be processed through processing nodes, such as the first set of processing nodes and the second set of processing nodes.
- the processing may result in output in output nodes.
- processing may comprise multiple steps or sequences.
- the first set of processing nodes may be a rough data filter
- the second set of processing nodes may be a more detailed data filter.
- the artificial neural network 2100 may be configured to effectuate decision-making.
- the artificial neural network 2100 may be configured to detect faces in photographs.
- the input nodes may be provided with a digital copy of a photograph.
- the first set of processing nodes may be each configured to perform specific steps to remove non-facial content, such as large contiguous sections of the color red.
- the second set of processing nodes may be each configured to look for rough approximations of faces, such as facial shapes and skin tones. Multiple subsequent sets may further refine this processing, each looking for further more specific tasks, with each node performing some form of processing which need not necessarily operate in the furtherance of that task.
- the artificial neural network 2100 may then predict the location on the face. The prediction may be correct or incorrect.
- the feedback system 2106 may be configured to determine whether or not the artificial neural network 2100 made a correct decision.
- Feedback may comprise an indication of a correct answer and/or an indication of an incorrect answer and/or a degree of correctness (e.g., a percentage).
- the feedback system 2106 may be configured to determine if the face was correctly identified and, if so, what percentage of the face was correctly identified.
- the feedback system may already know a correct answer, such that the feedback system may train the artificial neural network 2100 by indicating whether it made a correct decision.
- the feedback system may comprise human input, such as an administrator telling the artificial neural network 2100 whether it made a correct decision.
- the feedback system may provide feedback (e.g., an indication of whether the previous output was correct or incorrect) to the artificial neural network 2100 via input nodes or may transmit such information to one or more nodes.
- the feedback system may additionally or alternatively be coupled to the storage such that output is stored.
- the feedback system may not have correct answers at all, but instead base feedback on further processing: for example, the feedback system may comprise a system programmed to identify faces, such that the feedback allows the artificial neural network 2100 to compare its results to that of a manually programmed system.
- the artificial neural network 2100 may be dynamically modified to learn and provide better input. Based on, for example, previous input and output and feedback from the feedback system 2106 , the artificial neural network 2100 may modify itself. For example, processing in nodes may change and/or connections may be weighted differently. Following on the example provided previously, the facial prediction may have been incorrect because the photos provided to the algorithm were tinted in a manner which made all faces look red. As such, the node which excluded sections of photos containing large contiguous sections of the color red could be considered unreliable, and the connections to that node may be weighted significantly less. Additionally, or alternatively, the node may be reconfigured to process photos differently. The modifications may be predictions and/or guesses by the artificial neural network 2100 , such that the artificial neural network 2100 may vary its nodes and connections to test hypotheses.
- the artificial neural network 2100 need not have a set number of processing nodes or number of sets of processing nodes, but may increase or decrease its complexity. For example, the artificial neural network 2100 may determine that one or more processing nodes are unnecessary or should be repurposed, and either discard or reconfigure the processing nodes on that basis. As another example, the artificial neural network 2100 may determine that further processing of all or part of the input is required and add additional processing nodes and/or sets of processing nodes on that basis.
- the feedback provided by the feedback system 2106 may be mere reinforcement (e.g., providing an indication that output is correct or incorrect, awarding the machine learning algorithm a number of points, or the like) or may be specific (e.g., providing the correct output).
- the machine learning algorithm may be asked to detect faces in photographs.
- the feedback system may indicate a score (e.g., 75% accuracy, an indication that the guess was accurate, or the like) or a specific response (e.g., specifically identifying where the face was located).
- a human operator/inspector may focus the inspection on a down-selected list of TMLs.
- a human operator may indicate as much so that the model 1412 can be updated to reflect the new wall thickness values.
- a localized corrosion may be erroneously identified in the system 1400 , and supervised human input into a machine learning or neural network executing in the digital analytics platform 112 may refine its alerts and model accordingly.
- the artificial neural network 2100 may be supported or replaced by other forms of machine learning.
- one or more of the nodes of artificial neural network 2100 may implement a decision tree, associational rule set, logic programming, regression model, cluster analysis mechanisms, Bayesian network, propositional formulae, generative models, and/or other algorithms or forms of decision-making.
- the artificial neural network 2100 may effectuate deep learning.
- FIG. 22 is a flowchart showing illustrative steps of a method 2200 performed in accordance with some embodiments disclosed herein.
- the method 2200 may be performed by a system 1400 when computer-executable instructions, which are stored in a non-transitory computer-readable medium, are executed by a processor.
- the method 2200 may, among other things, down-select from among probe assemblies installed on a piping system in an industrial facility.
- the system and method for optimized asset health monitoring is improved because representative measurement locations are identified through down selection and the remaining probe assemblies can be disregarded during routine inspections of the piping system and other components in an industrial facility.
- the system is storing, in a computer memory 1406 communicatively coupled to the processor 1402 , historical pipe wall thickness measurements collected over a period of time from the probe assemblies 150 A installed on the piping system in the industrial facility.
- the data analytics platform 112 may set one or more hyperparameters, such as but not limited to a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, a group_size hyperparameter, and/or combination thereof.
- the system 1400 may begin training the model with at least the historical pipe wall thickness measurements stored in the computer memory 1406 and hyperparamter values stored in computer memory 1404 .
- the model stored in computer memory 1404 may group a first set of the probe assemblies from among the numerous probe assemblies installed on the piping system. As explained in this disclosure, such as with respect to FIG. 15 , several methods are provided by which the grouping may occur in the model.
- the data analytics platform 112 may assign a unique group identifier (groupID) to each set of probe assemblies.
- groupID may be any identifier that the system 1400 can use to uniquely refer to the group of probe assemblies.
- the data analytics platform 112 selects, based on at least the trained model, an optimization function for the operation of the system 1400 .
- an optimization function for the operation of the system 1400 .
- Numerous illustrative optimization functions are described in this disclosure, including but not limited to a median_TML_within_groupID optimization function, minimum_average_TML_within_groupID optimization function, minimum_variation_from_mean optimization function, and/or TML_position optimization function.
- the decision to select a specific optimization function causes subsequent identification and measurement steps to be effected.
- the system 1400 identifies, based on the model stored in the computer memory 1404 and selected optimization function, a probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system.
- the system 1400 may identify a single probe assembly for the entire groupID to be representative of the area being measured. In other examples, multiple probe assemblies may be identified to be representative of the groupID.
- the number of TMLs assigned to be down-selected from a group may be based on one or more rules.
- the 0.25 step value can be modified for adjusting the sensitivity of TML selection. Decreasing the 0.25 value leads to more TMLs being selected in each group (i.e., a conservative approach), and increasing the 0.25 value leads to less TMLs being selected in each group (i.e., an aggressive approach).
- the system 1400 may disregard all remaining probe assemblies in each groupID except the probe assembly identified from each groupID.
- the system may measure the wall thickness of each of the identified probe assemblies for each groupID, but exclude the other probe assemblies in the groupID.
- the system down-selects from among the plurality of probe assemblies installed on a piping system. At least one benefit of down-selecting the number of probe assemblies to use during an inspection is the time savings that results. For example, a human inspector that might have previously checked each probe assembly may now check measurements at a reduced number of probe assemblies without substantially increasing the risk of missing dangerous localized corrosion.
- the system 1400 may output a human-readable report listing those probe assemblies a human inspector should manually inspect for wall thickness measurements.
- the output may be ordered in any of various ways—e.g., based on highest risk of localized corrosion, based on geographic convenience from a known start position of the human inspector, or other order.
- the amount of risk when the amount of risk is graphed against the quantity of measurement locations taken, the amount of risk is asymptotic to a threshold minimum amount of risk 1602 regardless of the number of measurement locations increases.
- the delta change in risk increases at an increasing rate as shown by graph 1604 —stated another way, reducing the number of probe assemblies that are sampled can increase the risk to an unsafe amount.
- the system 1400 and method 2200 disclosed herein shifts the graph from an initial risk graph 1606 to a more favorable risk graph 1608 .
- the thickness monitoring controller 130 may receive and send the pipe wall thickness measurement of the probe assembly from each groupID for inspection.
- the thickness monitoring controller 130 may send the measurement data (and any other data) to the data store 1406 for historical recordkeeping and analytics, as well as to the data analytics platform 112 for analysis and generation of visualizations.
- the wall thickness measurements may show that a particular segment of pipe in the piping system is suffering from degradation other than general corrosion such that it rises to the level of dangerous, localized corrosion and must be replaced within a particular period of time.
- pipe wall thickness measurements may be taken at one or more of a pipe, tank, vessel, and/or pipeline at a facility.
Landscapes
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Acoustics & Sound (AREA)
- Engineering & Computer Science (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental & Geological Engineering (AREA)
- Environmental Sciences (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
- Length Measuring Devices Characterised By Use Of Acoustic Means (AREA)
Abstract
This disclosure relates to the field of corrosion and erosion monitoring of pipes and vessels. More specifically, this disclosure relates to a system and method for corrosion and erosion monitoring of pipes and vessels, where the system/method combines ultrasonic thickness monitoring using longitudinal waves with ultrasonic area monitoring using one or more guided waves, whereby representative thickness measurements are complemented by an area monitoring feature to detect localized corrosion/erosion in between representative thickness measurement locations. In another embodiment, a system and method for optimized asset health monitoring that includes an analytics solution is disclosed.
Description
- This application is related to U.S. Provisional Patent Application Ser. No. 62/982,751 with Attorney Docket No. MX-2020-PAT-0029-US-PRO, filed Feb. 28, 2020, with title “SYSTEM AND METHOD FOR CORROSION AND EROSION MONITORING OF PIPES AND VESSELS.” The aforementioned patent application is incorporated by reference in its entirety herein.
- This disclosure relates to the field of corrosion and erosion monitoring of pipes and vessels. Specifically, this disclosure relates to a corrosion and/or erosion monitoring system comprising mechanical components, hardware, software, analytics, and/or a combination thereof. In one embodiment, the mechanical components and hardware may comprise one or more ultrasonic transducers, base units, gateways, and/or combination thereof. The system may further comprise a software platform for remote monitoring. The system may further comprise, in some embodiments, analytics tools for front-end services and back-end services for remote monitoring and/or diagnostics. More specifically, in some embodiments, this disclosure may relate to a system and method for corrosion and erosion monitoring of pipes and vessels, where the system/method combines ultrasonic thickness monitoring using longitudinal waves with ultrasonic area monitoring using one or more guided waves, whereby representative thickness measurements are complemented by an area monitoring feature to detect localized corrosion/erosion in between representative thickness measurement locations. In another embodiment, a system and method for optimized asset health monitoring that includes an analytics solution is disclosed.
- The use of ultrasonic transducers for ultrasonically monitoring the condition and integrity of structural assets, including pipes and pressure vessels, such as those used in the oil and gas and power generation industries, is well-known. At present, corrosion and erosion monitoring systems and techniques incorporating/using ultrasonic transducers are known to include thickness monitoring at a location and area monitoring (also known as guided wave inspection). However, these two systems and techniques are typically separate from one another. Moreover, internal corrosion of piping systems is also sometimes monitored using radiographic (RT) thickness testing, in addition to ultrasonic (UT) testing, to measure wall thicknesses for selected components at prescribed intervals, over the life of the system.
- Thickness monitoring ultrasonic transducers and systems utilizing same typically measure a thickness of a pipe/vessel wall at the spot where the ultrasonic transducer is provided—in other words, it does not provide any information regarding the thickness of the pipe/vessel wall at locations surrounding the exact spot where the ultrasonic transducer is provided. As such, if corrosion/erosion is occurring at a location other than where the ultrasonic transducer is provided, it is likely that the corrosion/erosion will not be detected, unless thickness monitoring is accompanied by ultrasonic transducer mapping. Of course, ultrasonic transducer mapping increases the inspection cost. These ultrasonic transducers and systems are, however, beneficially permanently installed on pipes/vessels.
- Conversely, area monitoring ultrasonic transducers and systems utilizing same typically measure the thickness of a pipe/vessel wall across a larger area of the pipe/vessel wall, which area being measured is typically beyond the location where the thickness monitoring ultrasonic transducers are provided on the pipe/vessel. Such area monitoring ultrasonic transducers and systems utilizing same will typically develop a thickness map of the pipe/vessel wall across the area being measured. In theory, such a generated thickness map is beneficial, but at present, such guided wave inspection is extremely complex as general hardware in that segment generates ten to twenty different guided wave modes, and the high number of wave modes and the complex analysis negatively impacts the confidence in the inspection results. Further, guided wave inspection is typically not permanently installed on pipes and vessels. Additionally, highly localized corrosion cannot be reliably detected with temporarily installed guided wave systems as described in API 574 (API 574, Inspection practices for piping system components, 4th edition, 2016).
- In addition, existing permanently installed corrosion monitoring systems fail to use adequate data to determine the placement of sensors in an industrial facility, such as an oil refinery and petrochemical plant, that transport fluids using piping systems. The piping system might transport the fluids to one or more tanks and/or chemical processing unit. Some piping systems handle dedicated fluids at prescribed temperatures and/or pressures; these piping systems may transfer highly corrosive fluids at elevated temperatures and pressures.
- Moreover, many industrial facilities face health and safety concerns. They might transport fluids that may be flammable and/or toxic. As such, a failure in the piping system may cause leakage to the atmosphere and/or exposure to plant personnel. Moreover, some facilities operate with no scheduled shutdown for several years. Therefore, reliability of the piping system and its components is of importance.
- In addition to health and safety concerns, unplanned outages due to piping system failures are problematic from a business consequence standpoint. Given the potential safety, health, environmental, and business risks associated with piping failures, the condition of piping systems is monitored to accurately project their remaining life and determine safe repair or replacement dates.
- As a result of the foregoing, certain individuals would appreciate improvements in systems and methods for corrosion and erosion monitoring of pipes and vessels.
- In the following description of various illustrative embodiments, reference is made to the accompanying drawings, which form a part hereof, and in which is shown, by way of illustration, various embodiments in which aspects of the disclosure may be practiced. It is to be understood that other embodiments may be utilized, and structural and functional modifications may be made, without departing from the scope of the present disclosure. It is noted that various connections between elements are discussed in the following description. It is noted that these connections are general and, unless specified otherwise, may be direct or indirect, wired or wireless, and that the specification is not intended to be limiting in this respect.
- A system of one or more computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination of them installed on the system that in operation causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions. One general aspect includes a method for down-selecting from among probe assemblies installed on a piping system. The method also includes setting a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter for a model, before training the model. The method also includes grouping, by the model executing on a processor, a first set of the probe assemblies based at least on historical pipe wall thickness measurements collected from the probe assemblies installed on the piping system over a period of time. The method also includes assigning a unique groupID to each set of probe assemblies. The method also includes selecting, by the model after training the model, an optimization function from among a plurality of optimization functions for the model. The method also includes identifying, by the model, a single probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system. The method also includes sending, by a thickness monitoring controller associated with the piping system, a pipe wall thickness measurement of the single probe assembly from each groupID for inspection. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
- Implementations may include one or more of the following features. The method may include one or more steps to, during the inspection, disregard all remaining probe assemblies in each groupID except the single probe assembly from each groupID. The grouping of the first set of the probe assemblies is further based at least on inspection information provided to the system and historical pipe wall thickness measurements collected over a period of time from the probe assemblies installed on the piping system. The piping system may include a tank, and where a first probe assembly of the probe assemblies is configured to measure a wall thickness of the tank. The method may also include steps for storing, in computer memory communicatively coupled to the processor, historical pipe wall thickness measurements collected over an extended period of time from the probe assemblies installed on the piping system; and for training, by the processor, the model with at least the historical pipe wall thickness measurements stored in the computer memory. The model may include an artificial neural network. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.
- One general aspect includes a system for detecting general corrosion (e.g., a lack of localized corrosion) to a plurality of components that transport materials across a distance. The system may also include a plurality of probe assemblies affixed to one or more of the components, where the probe assemblies may include at least a thickness monitoring ultrasonic transducer and an area monitoring ultrasonic transducer configured to detect corrosion (e.g., general corrosion and/or localized corrosion) to the components. The system may also include a data store configured to store historical wall thickness measurements collected over a period of time from measurements performed by the probe assemblies. The system may also include a model trained on the historical wall thickness measurements in the data store and with hyperparamters may include a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter. The system may also include a monitoring apparatus may include a processor and a memory storing computer-executable instructions that, when executed by the processor, cause the system to perform steps that may also include: grouping, based on the model, a first set of the probe assemblies; assigning a unique groupid to each set of probe assemblies; selecting, based on the model, an optimization function from among a plurality of optimization functions; identifying, based on the model and selected optimization function, a probe assembly corresponding to each groupid for wall thickness monitoring of the components; and sending, by a thickness monitoring controller associated with the components, a wall thickness measurement of the probe assembly from each groupid for inspection. In another embodiment, the system may output a list of the unique identifiers corresponding to any groupID in lieu of sending the wall thickness measurement for inspection. An inspector may receive the system's output and react accordingly, as discussed in various embodiments disclosed herein. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
- Implementations may include one or more of the following features. The system, where the probe assembly identified from each groupID, may include more than one probe assembly of the plurality of probe assemblies, and where the memory of the monitoring apparatus stores computer-executable instructions that, when executed by the processor, cause the system to perform steps that may include: during the inspection, disregarding all remaining probe assemblies in each groupID except the more than one probe assembly from each groupID; and validating that the wall thickness measurements of the more than one probe assembly from each groupID is general corrosion and not localized corrosion. The wall thickness measurement of the probe assembly from a first groupID may include a thickness of a wall of a pipe component at the probe assembly. The wall thickness measurement of the probe assembly from a first groupID may include a thickness of a wall of a tank component at the probe assembly. The method may include validating that the pipe wall thickness measurement of the single probe assembly is general corrosion (e.g., a lack of localized corrosion) by: (i) generating a probability plot of all pipe wall thickness measurements associated with the piping system, (ii) grouping the plotted pipe wall thickness measurements by nominal thickness, and (iii) identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm the generalized corrosion (e.g., lack of localized corrosion). The pipe wall thickness monitoring may include steps for, by the probe assemblies, analyzing the original wall thicknesses, wall thickness loss over time, calibration error, and measurement location repeatability error. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.
- Implementations may include one or more of the following features. The method may further include steps for validating that the pipe wall thickness measurement of the single probe assembly is general corrosion (e.g., a lack of localized corrosion) by: generating a probability plot of all pipe wall thickness measurements associated with the piping system, grouping the plotted pipe wall thickness measurements by nominal thickness, and identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm the general corrosion (e.g., the lack of localized corrosion). The pipe wall thickness monitoring may include steps, by the probe assemblies, for analyzing the original wall thicknesses, wall thickness loss over time, calibration error, and measurement location repeatability error. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium.
- The present disclosure is illustrated by way of example and not limited in the accompanying figures in which like reference numerals indicate similar elements and in which:
-
FIG. 1 is an illustration of the system for corrosion/erosion monitoring; -
FIG. 2 is an illustration of a thickness monitoring controller and a piezo assembly of the system ofFIG. 1 ; -
FIG. 3 is an illustration of the thickness monitoring controller ofFIG. 2 ; -
FIG. 4 is an illustration of a switch assembly forming part of the piezo assembly ofFIG. 2 ; -
FIG. 5 is an illustration of the piezo assembly ofFIG. 2 ; -
FIG. 6 ,FIG. 7 , andFIG. 8 are illustrations of the method for corrosion/erosion monitoring; -
FIG. 9 ,FIG. 10 ,FIG. 11 , andFIG. 12 are illustrations to display the signal modulation; -
FIG. 13A andFIG. 13B (collectively referred to as “FIG. 13 ”) are drawings of one illustrative piping with installed MUT sensors in accordance with one or more aspects of the features disclosed herein; -
FIG. 14 is an illustrative network architecture of an industrial facility in accordance with various aspects of the disclosure; -
FIG. 15 is an illustrative diagram of probe assembly groupings in one embodiment of the disclosure; -
FIG. 16A ,FIG. 16B , andFIG. 16C (collectively referred to as “FIG. 16 ”) illustrate plots on a graph.FIG. 16A is a graph illustrating probability plot of measurement values for validating general corrosion in contrast to localized corrosion.FIG. 16B is a graph charting level of risk against TMLs in accordance with various aspect disclosed herein.FIG. 16C illustrates a shift in the curve depicting the level of risk against TMLs after down-selection in accordance with various aspect disclosed herein; -
FIG. 17 is a graph plot of illustrating cumulative thickness distribution for tubes with naphthenic acid corrosion; -
FIG. 18A is a corrosion sensor analytics graph illustrating TML measurements by date in one embodiment of the disclosure; -
FIG. 18B is another corrosion sensor analytics graph illustrating TML measurements by date as inFIG. 8A , but with a higher grouping sensitivity setting; -
FIG. 18C is yet another corrosion sensor analytics graph illustrating TML measurements by date as inFIG. 8A , but with an even higher grouping sensitivity setting; -
FIG. 19A andFIG. 19B are graphs in accordance with one or more aspects of the disclosure; -
FIG. 20A andFIG. 20B are also graphs in accordance with one or more aspects of the disclosure; -
FIG. 21 shows an illustrative artificial neural network configured to operate in collaboration with systems, methods, and algorithms disclosed herein; and -
FIG. 22 is a flowchart showing illustrative steps of a method performed in accordance with some embodiments disclosed herein; -
FIG. 23 is an illustration of a simplified pipe and instrumentation diagram (PID) corresponding to an illustrative corrosion/erosion monitoring system, as illustrated inFIG. 1 , in accordance with some embodiments disclosed herein. - In the following description of various illustrative embodiments, reference is made to the accompanying drawings, which form a part hereof, and in which is shown, by way of illustration, various embodiments in which aspects of the disclosure may be practiced. It is to be understood that other embodiments may be utilized and structural and functional modifications may be made, without departing from the scope of the present disclosure. It is noted that various connections between elements are discussed in the following description. It is noted that these connections are general and, unless specified otherwise, may be direct or indirect, wired or wireless, and that the specification is not intended to be limiting in this respect.
- While the disclosure may be susceptible to embodiment in different forms, there is shown in the drawings, and herein will be described in detail, specific embodiments with the understanding that the present disclosure is to be considered an exemplification of the principles of the disclosure, and is not intended to limit the disclosure to that as illustrated and described herein. Therefore, unless otherwise noted, features disclosed herein may be combined to form additional combinations that were not otherwise shown for purposes of brevity. It will be further appreciated that in some embodiments, one or more elements illustrated by way of example in a drawing(s) may be eliminated and/or substituted with alternative elements within the scope of the disclosure.
- Aspects of the disclosure relates to the monitoring and detection of corrosion and/or erosion of pipes, vessels, and other components in an industrial facility. The monitoring system may comprise a software platform for remote monitoring and analytics of historical measurements collected by a plurality of sensors affixed to the pipes and components. The monitoring system may include analytics tools for monitoring, diagnostics, and/or prediction of localized corrosion and/or general corrosion. By using the analytics systems disclosed herein, the thickness monitoring locations (TML) may be optimized to, among other things, reduce the number of measurement locations without compromising risk—i.e., down-selecting. Through down-selecting, by strategically reducing the number of probe assemblies that need to be sampled during an inspection, the amount of time/cost of an inspection is reduced while simultaneously maintaining (or even reducing) the risk profile of the industrial facility, as explained in this disclosure.
- A
system 100 for monitoring corrosion and erosion of pipes/vessels is illustrated inFIG. 1 andFIG. 2 . Thesystem 100 includes a data analytics andvisualization platform 110, anoptional gateway 120, athickness monitoring controller 130, a thickness monitoringultrasonic transducer 140 that is used for standardization purposes, and at least oneprobe assembly 150. Eachprobe assembly 150 includes aswitch assembly 160, at least one thickness monitoringultrasonic transducer 170, and at least one area monitoringultrasonic transducer 180. - The data analytics and
visualization platform 110 includes adata analytics portion 112 and avisualization portion 114. 1 Thedata analytics portion 112 is typically a cloud-based powered software that is configured to receive signals, typically wirelessly, from one or both of thegateway 120 or thethickness monitoring controller 130. These signals are analyzed by thedata analytics portion 112 to translate them into visuals for display on thevisualization portion 114. Thevisualization portion 114 may be any suitable device, e.g., a computer monitor, a tablet, a phone, etc., that are of a type that will aid an individual monitoring theplatform 110 in understanding the information regarding corrosion/erosion identified by thesystem 100. The individual may also be able to change the images/information on thevisualization portion 114 by providing further inputs to the software. - The
gateway 120 may be provided to receive signals, typically wirelessly, from thethickness monitoring controller 130, and to send such signals, typically wirelessly, to theplatform 110. For instance, it may be more economical to use thegateway 120 to establish cellular connection instead of having eachthickness monitoring controller 130 at a facility having its own data plan. In such a case, thethickness monitoring controllers 130 would use, e.g., the XBee protocol, to communicate with thegateway 120. In another example, if there is no good cellular connection at the location of thethickness monitoring controller 130, thegateway 120 could be installed at a higher location to establish cellular connection and thethickness monitoring controller 130 would submit data to thegateway 120 using, for example, the XBee protocol. - As best illustrated in
FIG. 3 , thethickness monitoring controller 130 includes amodem 131, amicroprocessor 132, apulser 133, an analog-to-digital converter (ADC) 134, anadjustable gain amplifier 135, a transmitchannel 136, and a receivechannel 137. Themodem 131 is configured to communicate with one or both of theplatform 110 and thegateway 120. Themodem 131 may use any appropriate communication option, including, but not limited to XBee 915 MHz and LTE-M/NB. Themodem 131 is configured to communicate with themicroprocessor 132. Themicroprocessor 132 may be any type of microprocessor which will provide the desired functions. Onesuch microprocessor 132 is the LPC4370 that is manufactured and sold by NXP Semiconductors. Themicroprocessor 132 is configured to communicate with both thepulser 133 and theADC 134. Thepulser 133 is preferably a high voltage pulser capacitor. TheADC 134 is preferably a 16-bit, 2 msps (million samples per second), but other ADC types may also be provided as appropriate. TheADC 134 is configured to communicate with the adjustable gain amplifier 135 (sometimes also commonly known as a variable gain amplifier). Theadjustable gain amplifier 135 preferably has a decibel range of 26-54 dB and a frequency range of 10 kHZ to 300 kHz, but other ranges may also be provided as appropriate. Thepulser 133 is configured to communicate with the transmitchannel 136 to transmit signals to the transmitchannel 136. Theadjustable gain amplifier 135 is configured to communicate with the receivechannel 137 to receive signals from the receivechannel 137. Thethickness monitoring controller 130 is preferably configured to accommodate a desired number of amplitude scans (“A-scans”) (or waveform displays). In the embodiments illustrated, thecontroller 130 is configured to accommodate sixteen A-scans (one from the thickness measurementultrasonic transducer 140 and five each from the three different probe assemblies 150). Of course, it is to be understood that as the number ofprobe assemblies 150 change and/or the number ofultrasonic transducers 170/180 are included in each probe assembly 150 (as will be discussed in further detail below), thecontroller 130 can be configured to accommodate more or less than sixteen A-scans as appropriate. - The thickness monitoring
ultrasonic transducer 140 is configured to receive signals from the transmitchannel 136 of thethickness monitoring controller 130 and is further configured to transmit signals to the receivechannel 137 of thethickness monitoring controller 130. As noted, the thickness monitoringultrasonic transducer 140 is used for standardization purposes and, thus, functions to calibrate the measurement system when a group of ultrasonic transducers are utilized (in this instance, the at least one thickness monitoringultrasonic transducer 170, and the at least one area monitoring ultrasonic transducer 180). The standardization thickness monitoringultrasonic transducer 140 works to ensure that thesystem 100 always performs the same way and functions properly, which is required by industrial standards. In the illustrated embodiment, the standardization thickness monitoringultrasonic transducer 140 is configured to perform a single A-scan. In practice, the thickness monitoringultrasonic transducer 140 is typically placed on a standardization block or a thickness calibrated metal piece to serve as a standardization transducer. - As illustrated in
FIG. 1 , thesystem 100 includes three different/distinct probe assemblies system 100, the number ofprobe assemblies 150 provided in thesystem 100 can be less than three (e.g., one or two) or can be more than three (e.g., four, five, etc.), as appropriate. Depending on the number ofprobe assemblies 150 provided in thesystem 100, minor variations/modifications may need to be made to thesystem 100 as would be understood by one of ordinary skill in the art. - As discussed above, each
probe assembly 150 includes aswitch assembly 160. As best illustrated inFIG. 4 , theswitch assembly 160 includes apower supply 161, a transmitswitch 162, amicrocontroller 163, amemory 164, a receiveswitch 165, anamplifier 166, and an optional resistance temperature detector (RTD)interface 167. Thepower supply 161 is in communication with the transmitchannel 136 of thethickness monitoring controller 130. The transmitswitch 162 is in communication with the transmitchannel 136 of thethickness monitoring controller 130. The transmitswitch 162 preferably has five “switch”channels microcontroller 163 is in communication with the transmitchannel 136 of thethickness monitoring controller 130, the transmitswitch 162, thememory 164, and the receiveswitch 165. Themicrocontroller 163 may be any type of microcontroller which will provide the desired functions. Onesuch microcontroller 163 is the PIC18 that is manufactured and sold by Microchip Technology. Thememory 164 is preferably a non-volatile memory. The receiveswitch 165 preferably has four “switch”channels amplifier 166 is in communication with the receivechannel 137 of thethickness monitoring controller 130 and the receiveswitch 165. Theamplifier 166 preferably has an amplification of 26 to 48 dB and a frequency range of 10 kHz to 300 kHz, but other levels/ranges may also be provided as appropriate. Theamplifier 166 is also preferably a two-stage amplifier, where 26 dB amplification is provided for a single stage option and 48 dB amplification is provided for a two-stage option, which can be selectable by populating or depopulating components on an amplification board. Theoptional RTD interface 167 is provided if the at least one thickness monitoringultrasonic transducer 170 incorporates an RTD 171 (as discussed below). In the illustrated embodiment, eachswitch assembly 160 is instructed bycontroller 130 to collect five A-scans (one from the thickness monitoringultrasonic transducer 170 and one from each of the four area monitoring ultrasonic transducers 180). - As discussed above, each
probe assembly 150 includes at least one thickness monitoringultrasonic transducer 170. As illustrated inFIG. 1 , eachprobe assembly 150 includes one thickness monitoringultrasonic transducer 170. Depending on thesystem 100 and theprobe assembly 150, the number of thickness monitoringultrasonic transducers 170 provided in eachprobe assembly 150 can be more than one (e.g., two, three, four, etc.), as appropriate. Depending on the number of thickness monitoringultrasonic transducers 170 provided in eachprobe assembly 150, minor variations/modifications may need to be made to theprobe assembly 150 and/orsystem 100 as would be understood by one of ordinary skill in the art. Each thickness monitoringultrasonic transducer 170 may optionally have anRTD 171 associated therewith to measure the temperature of the pipe/vessel at or near where the thickness measurement is occurring. Each thickness monitoringultrasonic transducer 170 is in communication with the fifth “switch”channel 162 e of the transmitswitch 162 and, if the thickness monitoringultrasonic transducer 170 includes theRTD 171, is also in communication with theRTD interface 167. - The thickness monitoring ultrasonic transducer 170 (as well as the 140) operates by generating high frequency ultrasonic waves (e.g., 5 MHz). These ultrasonic waves are commonly referred to as longitudinal waves (LW) and, as such, the thickness monitoring
ultrasonic transducers 170 may also be referred to as LW transducers. In the illustrated embodiment, each thickness monitoringultrasonic transducer 170 is configured to perform a single A-scan. Unlike the thickness monitoringultrasonic transducer 140, the thickness monitoringultrasonic transducer 170 is not placed on a standardization block or a thickness calibrated metal piece, but rather is placed on the pipe/vessel to measure the thickness of the pipe/vessel at the location where it is installed. - As discussed above, each
probe assembly 150 includes at least one area monitoringultrasonic transducer 180. As illustrated inFIG. 1 ,FIG. 2 ,FIG. 3 ,FIG. 4 , andFIG. 5 , eachprobe assembly 150 includes four area monitoringultrasonic transducers 180A, 180B, 180C, 180D (each also referred to as area monitoring ultrasonic transducer 180). Depending on thesystem 100 and theprobe assembly 150, the number of area monitoringultrasonic transducers 180 provided in eachprobe assembly 150 can be less than four (e.g., one, two or three) or more than four (e.g., five, six, etc.), as appropriate. Depending on the number of area monitoringultrasonic transducers 180 provided in eachprobe assembly 150, minor variations/modifications may need to be made to theprobe assembly 150 and/orsystem 100 as would be understood by one of ordinary skill in the art. The first area monitoringultrasonic transducer 180A is in communication with the first “switch”channel 162 a of the transmitswitch 162 and the first “switch”channel 165 a of the receiveswitch 165. The second area monitoring ultrasonic transducer 180B is in communication with the second “switch”channel 162 b of the transmitswitch 162 and the second “switch”channel 165 b of the receiveswitch 165. The third area monitoring ultrasonic transducer 180C is in communication with the third “switch” channel 162 c of the transmitswitch 162 and the third “switch” channel 165 c of the receiveswitch 165. The fourth area monitoring ultrasonic transducer 180D is in communication with the fourth “switch” channel 162 d of the transmitswitch 162 and the fourth “switch” channel 165 d of the receiveswitch 165. - In an embodiment, the
probe assembly 150 may comprise athickness transducer 170 and a set ofarea transducers 180 individually wired to switch/preamp assembly 160. In a different embodiment, thickness andarea transducers preamp assembly 160. In another embodiment, it also can be a set of larger probes (thickness+2 area, area+area etc.) - The area monitoring
ultrasonic transducers 180 operate by generating low frequency ultrasonic waves (e.g., 50 kHz to 500 kHz). These ultrasonic waves are commonly referred to as guided waves (GW) and, as such, the area monitoringultrasonic transducers 180 may also be referred to as GW transducers. One such type of guided wave, namely shear horizontal zero waves (called SH0 in plates or T(0,1) in piping), from GW transducers are of interest due to their non-dispersive behavior. In the illustrated embodiment, each area monitoringultrasonic transducer 180 is configured to perform a single A-scan. - The
GW transducers 180 are preferably in the form of piezo patch transducers, but may alternatively be in other forms, such as, for instance, face-shear piezo elements. In a preferred embodiment, as best illustrated inFIG. 1 andFIG. 5 , theGW transducers 180A, 180B, 180C, 180D are positioned in a rectangular configuration around theLW transducer 170, whereGW transducer 180A is positioned above and to the left ofLW transducer 170, GW transducer 180B is positioned below and to the left ofLW transducer 170, GW transducer 180C is positioned below and to the right ofLW transducer 170, and GW transducer 180D is positioned above and to the right ofLW transducer 170. When applied to a pipe/vessel, a straight line fromGW transducer 180A to GW transducer 180B is parallel to a straight line from GW transducer 180C to GW transducer 180D, and a straight line fromGW transducer 180A to GW transducer 180D is parallel to a straight line from GW transducer 180B to GW transducer 180C. Further, when applied to a pipe/vessel, a straight line fromGW transducer 180A to GW transducer 180C intersectsLW transducer 170, and a straight line from GW transducer 180B to GW transducer 180D intersectsLW transducer 170, such that an “X-shape” configuration is provided. - The
system 100, when associated with a pipe/vessel, may be utilized to measure the corrosion/erosion of the pipe/vessel. In an embodiment, onemethod 200 of measuring the corrosion/erosion of the pipe/vessel is described below and illustrated inFIG. 6 ,FIG. 7 , andFIG. 8 . - The
method 200 includes thestep 205 of manually measuring the actual longitudinal velocity and the temperature of the pipe/vessel to be inspected. - The
method 200 includes thestep 210 of manually measuring the actual guided wave velocity and the temperature of the pipe/vessel to be inspected. - The
method 200 includes thestep 215 of performing a thickness standardization measurement with the standardization thickness monitoringultrasonic transducer 140 and the RTD 171 (it is to be understood that, like the thickness monitoringultrasonic transducer 170, the standardization thickness monitoringultrasonic transducer 140 could also optionally incorporate the RTD 171). - The
method 200 includes thestep 220 of performing measurements using theprobe assembly 150A. Step 220 includes the sub-step 220 a of performing a thickness measurement with the thickness monitoringultrasonic transducer 170 and theRTD 171. Step 220 includes the sub-step 220 b of performing an area thickness monitoring with the area monitoringultrasonic transducers 180A, 180B, 180C, 180D at a first frequency. Sub-step 220 b includes the sub-step 220b 1 of performing axial scanning whereby area monitoringultrasonic transducer 180A is excited and data is recorded with area monitoring ultrasonic transducer 180B. The measurement taken insub-step 220b 1 is repeated as often as specified in configuration setting and average A-scans. Sub-step 220 b includes the sub-step 220b 2 of performing axial scanning whereby area monitoring ultrasonic transducer 180C is excited and data is recorded with area monitoring ultrasonic transducer 180D. The measurement taken insub-step 220b 2 is repeated as often as specified in configuration setting and average A-scans. Sub-step 220 b includes the sub-step 220b 3 of performing circumferential scanning whereby area monitoringultrasonic transducer 180A is excited and data is recorded with area monitoring ultrasonic transducer 180D. The measurement taken insub-step 220b 3 is repeated as often as specified in configuration setting and average A-scans. Sub-step 220 b includes the sub-step 220b 4 of performing circumferential scanning whereby area monitoring ultrasonic transducer 180C is excited and data is recorded with area monitoring ultrasonic transducer 180C. The measurement taken insub-step 220b 4 is repeated as often as specified in configuration setting and average A-scans. Thus,channels 162 a, 162 c (which are associated withGW transducers 180A, 180C) act as guided wave transmit channels whilechannels 162 b, 162 d (which are associated with GW transducers 180B, 180D) act as guided wave receive channels. The receive path further goes via theamplifier 166 to the receivechannel 137 of thethickness monitoring controller 130. - Step 220 includes the sub-step 220 c of repeating sub-step 220 b at a second frequency, which second frequency is different from the first frequency.
- Step 220 includes the sub-step 220 d of repeating sub-step 220 b at a third frequency, which third frequency is different from both the first frequency and the second frequency.
- The
method 200 includes thestep 225, which comprises repeatingstep 220 to perform measurements using theprobe assembly 150B. - The
method 200 includes thestep 230, which comprises repeatingstep 220 to perform measurements using the probe assembly 150C. - Thus, the
method 200 combines ultrasonic thickness monitoring using longitudinal waves with ultrasonic area monitoring using guided waves and, in a preferred embodiment, just one special non-dispersive shear wave mode (SH0 or T(0,1)). Themethod 200 takes representative thickness measurements, rather than trying to develop a thickness map, which will be complemented by an area monitoring feature to detect localized corrosion/erosion in-between representative thickness measurement locations. Thesystem 100 utilizes new electronics which use a single circuitry to deliver two distinctive, different excitation signals, e.g., high frequency ultrasonic waves for thickness monitoring (5 MHz) and low frequency ultrasonic waves for area monitoring (50-500 kHz), from two different types of ultrasonic transducers, e.g.,LW transducer 170 andGW transducers 180. Each excitation signal needs to be generated and processed differently. More specifically,pulser 133 of thecontroller 130 is a digital switch capable of delivering only predetermined fixed voltage levels: high voltage, low voltage and zero voltage. High and low voltage levels are normally adjustable in a range of 5V to 90V and −5V to −90V but different voltage levels are permissible as well.Microprocessor 132 signals pulser 133 to output to transmitchannel 136 one of the fixed voltage levels: ex. high voltage for a specified period of time. Example of a pulse used to excite LW transducer 170:processor 130 instructs pulser 133 to output 0V, then high voltage for a period of 100 ns, then low voltage for a period of 100 ns, then 0V. Described sequence would generate bipolar square wave of 5 MHz frequency suitable to exciteLW transducer 170. ForGW transducers 180, different frequencies and signals amplitudes are required. - As best illustrated in
FIG. 9 ,FIG. 10 ,FIG. 11 , andFIG. 12 , waveforms needed to exciteGW transducers 180 can have rather complex shapes like, ex: 5 cycle sinusoid wave superimposed on Hanning window signal (ex. half cycle cosine) shown as 330 inFIG. 12 that would allow for a smoother transition from no-signal to signal condition. To generateGW transducer 180 suitable waveforms combination of apulser 133 digital output shown aswaveform 300 inFIG. 9 ,FIG. 10 ,FIG. 11 , andFIG. 12 , in-series resistance of the transmitchannel 136 and impedance of theGW transducer 180 are used.GW transducer 180 impedance in a frequency range used to generate GW waves (50-500 kHz) is usually in majority composed of capacitance. This capacitance and mentioned in-series resistance of the transmitchannel 136 form a low pass filter.Pulser 133 under instructions from themicroprocessor 132 generates a high frequency (usually in range of tens of MHz)digital waveform 300 that when passed thru the transmitchannel 136 andGW transducer 180 capacitance results in adifferent waveform 310 than originally outputted from the pulser 133 (as illustrated inFIG. 10 ). Varying high frequency digital waveforms from thepulser 133 can generate, once passed thru the transmitchannel 136 in-series resistance andtransducer 180 capacitance, a range of analog waveforms, ex: sinusoids without Hanning windows, shown as 320 inFIG. 11 or sinusoids with Hanning windows, shown as 330 inFIG. 12 , chirp (frequency changes during duration of the pulse), ramp-up, seesaw and other. Of course, other waveforms than those as described and illustrated could also be generated. - In an embodiment, a chirp signal can be used to excite multiple frequencies at the same time from a single channel Proper software filtering can decode the individual frequency response from a single A-scan.
- By using the
system 100 andmethod 200, the time of flight and the amplitude of the echo reflected at a defect on the pipe/vessel can be evaluated. More specifically, by sending excitation signals from GW transducer 180C and receiving by GW transducers 180B, 180D, the reflection echo will be earlier in time trace in the GW transducer 180B, 180D that is closer to the damage, e.g., GW transducer 180B if the damage is to the left of both GW transducers 180B, 180D, or GW transducer 180D if the damage is to the right of both GW transducers 180B, 180D, where GW transducers 180B, 180D are positioned as illustrated inFIG. 1 andFIG. 5 . Defects as pittings or corrosion/erosion patches usually increase in size over time. Therefore, the amplitude of the echoes reflected at the defects will increase over time. Permanently installed systems therefore allow one to monitor the change of amplitude next to the time-of-flight. Monitoring changes in A-Scans after for example baseline subtraction and digital filtering reduces the complexity of the analysis and increases confidence in the inspection results. Next to baseline subtraction additional digital signal processing tools or machine learning algorithms can be used for feature extraction or pattern recognition which additionally increase confidence levels and help to detect changes earlier in time. -
FIG. 23 illustrates a simplified pipe and instrumentation diagram (PID) corresponding to an illustrative corrosion/erosion monitoring system, as illustrated inFIG. 1 , in accordance with some embodiments disclosed herein. Thesimplified PID 2300 includes numerous probe assemblies depicted as circles numbered nineteen to eighty-four. For example, three different/distinct probe assemblies PID 2300 can be any number, as appropriate. In one example, a human operator/inspector may focus the inspection on a down-selected list of TMLs, as explained herein. These down-selected TMLs may represent more efficient candidate measuring locations to capture general corrosion behavior of the entire asset, while still being able to inspect for localized corrosion. For example, substantial amount of time/energy and cost may be saved by down-selecting the number of TMLs so that only those probe assemblies with the highest probability of detecting localized corrosion are examined by the human operator/inspector. Rather than checking all of probe assemblies nineteen to eighty-four, or even randomly checking less than all of probe assemblies nineteen to eighty-four, the down-selected TMLs are a more optimal identification of which TMLs to measure. In some examples, the inspector may use a handheld or other manual device to measure wall thickness at the numbered locations on thesimplified PID 2300. In other examples, a rig or harness of sorts may be pre-installed at the numbered location on thesimplified PID 2300 to allow the inspector to measure wall thickness at each thickness measuring location. In yet another example, the inspector may be an automated machine that takes measurements at the down-selected TMLs at particular time intervals. Even in an automated measuring system, down-selecting TMLs is advantageous because it reduces the amount of processing power and network bandwidth consumed by measurement data generated by a measuring device at each numbered location on thesimplified PID 2300. For example, some large industrial facilities may have thousands upon thousands of probe assemblies that could result in a prohibitive amount of generated data. In addition, once any localized corrosion has been confirmed and repaired, a human operator may indicate as much so that any model can be updated to reflect the new wall thickness values. In addition, in some examples, if a localized corrosion is erroneously identified, then supervised human input into a machine learning or neural network, which is executing in a digital analytics platform, may refine its alerts and model accordingly. -
FIG. 13A illustrates an illustrative piping withsensors 1301 installed on the pipe in accordance with one or more aspects of the features disclosed herein. The pipe may have a flow of liquid in the direction depicted by the arrows. During an inspection, one approach may be to inspect and take measurements from each and everysensor 1 to 6 depicted inFIG. 13A . In another example, a random selection of sensors may be inspected and measured. In accordance with several of the systems and methods disclosed herein, in another example, the plurality of thickness monitoring locations (TMLs) shown at eachsensor 1 to 6 may be intelligently considered and a smaller/narrower set of TMLs may be down-selected for inspection. Moreover, in accordance with several of the systems and methods disclosed herein, the TMLs may be grouped based on one or more criteria in the process of down-selecting the TMLs. The down-selecting criteria may, in one simplified example, identify and exclude those sensors (e.g.,sensors 1 and 3) that historically measured only general corrosion in its area. Thus, by down-selecting thesystem 100 avoids using clustering, but instead uses grouping to down-select some sensors as being superfluous to the assessment of the health of the mechanical component. Thus, saving time and resources. In contrast, some prior systems attempted to reduce risk by adding more TMLs and inspections of those TMLs. However, the risk-based inspection (RBI) approach described in various aspects of this disclosure provides a superior process and system. An RBI approach may also use a model that takes into consideration other criteria such as the type of fluid being transported in the piping system, the temperature inside and outside of the pipes/components, elbow/configuration of the piping components, and other criteria. For example, the measurements at an elbow may be weighted to be more likely to be selected as part of down-selecting in a group because historically, the locations near an elbow in piping is a place that will have more turbulence and friction, thus a possibility of higher corrosion and acidity. - Referring to
FIG. 13B , probe assemblies 1302 may comprise a tethered device that captures accurate spot measurements of thickness of components. In another embodiment, probe assemblies may comprise a tethered device that captures accurate spot measurements and area monitoring. For example, the device inFIG. 13B or comparable devices may be used to capture area monitoring of the thickness of a pipe component. In yet another embodiment, the probe assembly may comprise a wireless device that captures accurate spot measurements without necessarily being in direct contact with a piping component that requires thickness monitoring. The probe assemblies may comprise one or more of thickness monitoring ultrasonic transducers, area monitoring ultrasonic transducers, and/or a combination thereof that are configured to validate general corrosion (e.g., confirm no detection of localized corrosion) in the piping system. -
FIG. 13B is a drawing of an illustrative piping with installed sensors. The sensors 1302 may be any of various types of sensors configured to measure a thickness of the piping at or near the vicinity of the point of its installation on the pipe. The sensors 1302 are typically installed in a permanent location and remains affixed to the pipe for an extended period of time (e.g., for the lifespan of that circuit of the piping, for over five years, for over three years, or other period of time). Although the sensors 1302 displayed inFIG. 13B are installed to the outside of the piping and tethered with wires, in some examples in accordance with one or more aspects of the disclosure, the sensors may be untethered and wirelessly communicate data to one or more wireless receiver/transceiver devices. In addition, although the sensors displayed inFIG. 13B are illustrated in a straight linear pattern along the longitude of the pipe, the disclosure contemplates sensors installed in any of several different patterns. For example, the density of installed sensors may be based on the direction of gravity and the type of substance being transported in the piping. For example, assuming in one example that the piping inFIG. 13B is transporting a liquid along the length of pipe from the left to the right when the bottom of the pipe is the portion of the pipe on which sensor 1302 is installed. In such an example, the sensors installed on the piping may be distributed around the circumference of the piping taking into consideration that climate conditions (e.g., rain, hail, sun) may expose portions of the pipe to greater possibility of deterioration while internal conditions in the piping (e.g., more liquid contacts the bottom of the pipe than the top of the pipe) may expose inner portions of the pipe to greater possibility of deterioration. -
FIG. 14 is an illustrative network architecture of an industrial facility with sensors, communication components, and other components in accordance with various aspects of the disclosure. Thedata analytics platform 112 may be communicatively coupled over a network, such as alocal area network 1408, to one or more networked components. For example, thedata analytics platform 112 may output to avisualization platform 114 for generation of one or more of the illustrative graphs included herein. A monitoring system may comprise thesoftware platform 112 to remotely monitor and analyze historical measurements collected by a plurality of sensors affixed to the pipes and components. The monitoring system may include analytics tools for monitoring, diagnostics, and/or prediction of areas that are candidates for localized corrosion (e.g., because the system was unable to confirm general corrosion to the area). By using the analytics systems disclosed herein, the TML may be optimized to, among other things, reduce the number of measurement locations without compromising risk—i.e., down-selecting. - In another example, the
data analytics platform 112 may trigger an alert to be generated at aremote alert device 1410. Theremote alert device 1410 may result in an immediate inspection of one or more components, or may result in particular piping components being prioritized for a subsequent inspection of the facility. - As measurements and other data are collected by the
system 1400, the data may be stored in adata store 1406 that is communicatively coupled and accessible to thedata analytics platform 112. In some examples, the data may be stored incomputer memory 1404, however, the amount of computer memory required may be high. Instead, in some examples, amodel 1412, such as a machine learning artificial neural network, may be stored at thecomputer memory 1404 for execution by aprocessor 1402, while historical data and other data may be stored at adata store 1406. In some examples, the data store may be moved into theplatform 112 although it is shown for illustrative purposes as communicating over thelocal area network 1408 with theplatform 112. -
FIG. 15 is an illustrative diagram of a plurality of sensor (e.g., probe assembly) groupings in one embodiment of the disclosure. Each probe assembly may be assigned a unique TML identifier (TML ID) as illustrated inFIG. 15 . The TML ID may be any unique letter, character, or other identifier that uniquely identifies each TML (i.e., probe assembly). InFIG. 15 , the thick-lined rectangular box around select TML ID numbers shows probe assembly groupings. In 1502, on Mar. 7, 2007, the system has groupedprobe assemblies computer memory 1404 shows that thesystem 1400 has adjusted the grouping to include/exclude one or more TMLs. In 1504, the model may recommend that the probe assembly corresponding to TML ID number four should no longer be a part of the groupID corresponding to the thick-lined rectangular box in 1504. As a result, the one or more probe assembly down-selected for that groupID may also change. Finally, in 1506, on Mar. 7, 2009, the graphical depiction shows that thesystem 1400 has further adjusted the grouping to nowgroup probe assemblies probe assemblies FIG. 16 , will change for theoverall system 100. - In one example, the grouping of TMLs into a groupID may be done in one of several different methods. For example, the initial grouping for each circuit of components at a facility may be based on the measurement data level. For every date on which measurements were taken by the probe assemblies, a new group may be triggered if a probe assembly satisfies any of the following conditions: (i) if the probe assembly is the first TML of the circuit; (ii) if the (absolute) difference between the measurement value and the preceding TML's measurement value is greater than about 0.5 to about 3.0 standard deviation of all measurements for that date, then the value of this parameter may be reduced for more conservative grouping, or increased for more aggressive grouping; (iii) if the TML's nominal wall thickness measurement is different as compared to the preceding TML's nominal wall thickness measurement; or if the TML has only one measurement historically (across all dates). In another example, the grouping of TMLs may be done in a multi-step process. In a first step, all measurements taken in a group of connected components (e.g., a circuit) on a particular date (or any other predefined time period—e.g., within a one-hour window of time, within the same week, or other) may be compared to determine how many pairs (or tuples) were measured on the particular date. In one example, any TML pairs that have less than a predetermined percentage (e.g., 70%, 80%, 60%, 75%, or other percent) of the total measurements within that survey year (or other time period) are deleted. Next, the minimum measurement value of all the TMLs may be identified and all TMLs that were paired in an earlier (e.g., first) step with that TML are assigned to the same groupID. Other examples of rules for grouping the TMLs would be apparent to a person having skill in the art after review of the entirety disclosed herein.
- Additional other illustrative rules for grouping the TMLs are contemplated in this disclosure. For example, in some rules the grouping may be reassigned based on TML pairing percentage. For a circuit of components that has at least two measurement dates, TML pairs that are grouped together at least a predetermined threshold percentage of times may be retained in the same group, but TMLs that do NOT meet this threshold may be individually assigned to separate groups using one or more rules. In yet another example, measurement dates that do not have sufficient TMLs may be dropped. For every circuit of components, the
system 1400 may consider, in some examples, only those measurement dates which have at least a predetermined threshold percentage of the maximum number of TMLs for any date. TMLs that appear in dates that do not meet this threshold may be individually assigned to separate groups. - In some examples, the
system 1400 may discard (e.g., drop) seemingly invalid measurements based on a lack of historical data, and proceed to re-group TMLs based one or more of the rules described herein. The thresholds used are hyperparameters that can be adjusted based on data set diversity and quality. This adjustment may occur at the end of the process upon data confirmation and validation. In one example, threshold percentage may be set to 75%, but with some TMLs the prior measurement might not have occurred in the past, many years. In some embodiments, a hyper-grid may be generated and used to adjust the parameters and/or hyperparameters of thesystem 1400. In some examples, the threshold setting may be strongly correlated to how many TML measurements asystem 1400 has collected for each TML ID. Thus, the threshold may be adjusted up or down based on how much data is made available to thesystem 1400. -
FIG. 16 andFIG. 17 show graph plots of various data collected and/or analyzed by thesystem 1400.FIG. 16A shows a probability plot graph of measurements values (normal is 95%) where percentage is on the Y-axis and measurement value is on the X-axis. Thesystem 1400 defaults to assuming that general corrosion has been detected, except when the plot shows that the tail is not running vertical, such as shown near the top of the graph inFIG. 16A . Thedata analytics platform 112 may validate that the pipe wall thickness measurement of the probe assembly is general corrosion and not localized corrosion by performing one or more steps. For example, in some embodiments, the validating may be performed by generating a probability plot of all pipe wall thickness measurements associated with the piping system, then grouping the plotted pipe wall thickness measurements by nominal thickness, and identifying a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm that the corrosion is likely not general corrosion. Meanwhile, where the graph shows a linear relationship, then the TMLs corresponding to those data points in the graph are exhibiting general corrosion. This approach is an advancement over systems that may have used standard deviation to build normal probability plots. Moreover, the validating step adds further assurance that thesystem 1400 is accurately detecting general corrosion and acting accordingly to down-select the appropriate probe assemblies installed on the components in the facility. Thesystem 1400 should not generate an alert (e.g., from device 1410) for general corrosion because general corrosion is pervasive and is typically not of primary interest during inspections. Rather, general corrosion is accounted for in the scheduling and planning for bulk replacement of components in a facility. - Referring to
FIG. 16B andFIG. 16C , those graphs illustrate the relationship between a risk of mis-identifying general corrosion and the quantity of thickness measurement locations (TMLs). Although the amount of risk is asymptotic to a threshold minimum amount ofrisk 1602 regardless of the number of measurement locations,FIG. 16B shows that the level of risk charted against the quantity of probe assemblies (i.e., TMLs) decreases as more TMLs are added. Meanwhile, the effects of the system and method disclosed herein are shown inFIG. 16C , which illustrates a shift in the curve depicting the level of risk charted against the quantity of TMLs after down-selection.FIG. 16B andFIG. 16C are described in more detail below in conjunction with the method steps illustrated in the flowchart ofFIG. 22 . Meanwhile,FIG. 17 is a graph illustrating cumulative thickness distribution for tubes with naphthenic acid corrosion in existing systems known in the art. -
FIG. 18A is a corrosion sensor analytics graph illustrating TML measurements by date for a specific circuit ID (or asset ID). The X-axis corresponds to TML identifiers. For practical purposes, the probe assemblies installed on a piping system may be assigned identifiers in a sequential or otherwise ordered sequence along the circuit formed by the piping system. Each TML might have an ID that shows its position upstream or downstream on the pipe. Other data cleaning and/or scrubbing of the TMLs based on positional data may be performed to harmonize/standardize the measured data for analysis. Each TML might be assigned a nominal thickness from when the pipe was first installed. One or more publicly available databases (e.g., Meridian database) may provide data, including nominal thickness measurements and specifications. Meanwhile, as the legend on the right-hand side ofFIG. 18A shows, measurements may be taken over a period of time so that historical data spanning at least a few years (i.e., an extended period of time) may be stored and analyzed. In this example, almost twenty-five years of wall thickness measurement data is stored, analyzed, and plotted inFIG. 18A .Graph plot 1802 inFIG. 18A corresponds to measurements taken on 2015 Aug. 3. Meanwhile, the other plots in the graph correspond to thickness measurements taken for each TML on the corresponding date spanning back almost twenty-five years (e.g., an extended period of time). - The
data analytics platform 112 may set one or more hyperparameter for themodel 1412 corresponding to the graph plotted inFIG. 18A . A hyperparameter is typically set before the training/learning process begins on a model; in contrast, the values of other parameters are derived through training of the model. InFIG. 18A , a graphical user interface for adjusting the grouping_sensitivity hyperparameter is displayed at the top. Thevisual platform 114 may include a graphical tool/slider through which the hyperparameter may be adjusted. InFIG. 18A , the grouping_sensitivity hyperparameter is shown set to a “standard” setting. Meanwhile, inFIG. 18B , which shows another illustration of themodel 1412, the grouping_sensitivity hyperparameter is shown set to a “medium” setting. As a result, the number of groups is only sixty-one inFIG. 18B instead of ninety-seven groups inFIG. 18A . In addition, the graph plotted 1812 inFIG. 18B is slightly different than thegraph 1802 inFIG. 18A due to the change in hyperparameter settings and TML selection methods. Furthermore, with the grouping_sensitivity hyperparameter set to “high” inFIG. 18C , the graph plotted 1822 inFIG. 18C is even more different fromFIG. 18A andFIG. 18B . The number of groups is about seventy-five while the total number of TMLs remains constant at one hundred fifty-five. - The grouping_sensitivity hyperparameter refers to the sensitivity or aggressiveness of TML grouping, and may be applied at the initial grouping stage. In some examples, a TML may be assigned to a new group when the (absolute) difference between the measurement value and the preceding TML's measurement value is greater than 1 standard deviation (SD) of all measurements for that date. This threshold can be adjusted for more conservative or aggressive grouping. A threshold less than 1 SD will result in the grouping being more sensitive to changes in measurements and will lead to a more conservative grouping. On the other hand, a threshold greater than 1 SD will cause the grouping being less sensitive to changes in measurements and will lead to more aggressive grouping (e.g., higher grouping ratios). In one example, five different grouping sensitivities may be implemented, as shown in
FIG. 18C , in decreasing sensitivity—from most conservative to most aggressive as follows: High (0.5 SD), Standard (1 SD), Medium (1.5 SD), Low (2 SD), and Very Low (3 SD). In another example, more or less than the aforementioned five groupings may be used to provide more granular or coarse sensitivity. As the grouping_sensitivity hyperparameter is applied at the initial grouping stage, as is the case hyperparameters, all subsequent grouping steps may be re-run based on the initial grouping results—i.e., the entire grouping cycle is repeated five times, once for each of the five grouping sensitivity levels. - Notably,
FIG. 18A ,FIG. 18B , andFIG. 18C (collectively referred to as “FIG. 18 ”) list a plurality of TML selection methods that may be applied to the measurement to optimize the grouping and plotting of the data points. AlthoughFIG. 18 lists three optimization functions—namely median_TML_within_groupID, minimum_average_TML_within_groupID, and minimum_variation_from_mean—other optimization functions may be used in accordance with one or more aspects of the disclosure. For example, a TML_position optimization function may be used where if one TML is to be selected, the TML at the center of the group is chosen. If two TMLs are to be selected, the group is split into two subgroups and the TMLs at the center of each subgroup are chosen, and so on. Other examples of TML selection methods are contemplated herein. For example, the optimization function may be a minimum_average_TML_within_groupID optimization function. In minimum_average_TML_within_groupID method for deciding which TML(s) to pick from each group, the method selects the TML(s) having the lowest average measurement within each group (across dates). For example, in one illustrative system using the minimum_average_TML_within_groupID optimization function, the system may calculate average measurement of each TML (across dates), rank TMLs in each group by (e.g., ascending) average measurement, and based on number of TMLs to be picked (n) from each group, pick first n TMLs. Likewise, the median_TML_within_groupID optimization function is similar to the minimum_average_TML_within_groupID optimization function, but based on the median instead of the minimum average. - In another example the optimization function may be a minimum_variation_from_mean optimization function. In a minimum_variation_from_mean method for deciding which TML(s) to pick from each group, the method selects the TML(s) having the lowest average variation from the mean measurement of the group. For example, in one illustrative system using the minimum_variation_from_mean optimization function, the system may calculate the mean group measurement for each date. Then, for each TML, for each date calculate the absolute difference from mean, and for each TML, calculate the average variation (e.g., absolute difference from mean). Next, the minimum_variation_from_mean optimization function ranks TMLs in each group by (e.g., ascending) average variation, and based on number of TMLs to be picked (n) from each group, pick first n TMLs.
- After finalizing groupings, the
system 1400 determines a TML candidate selection method and a desired number of probe assemblies per group. The number of candidates per group may be another hyper-parameter. By default, thesystem 1400 may use the greater of 1% or one for each TML group. If more than one TML candidate is to be selected, then the TML group may be divided into equally big sub-groups while preserving the TML ordering. Then, thesystem 1400 may apply a TML candidate selection method based on the one or more scenarios described herein. -
FIG. 19A is agraph 1902 showing the measured TC thickness in millimeters of a component over time. Alternatively, the graph may show the values of temperature calibration (in Celsius), temperature coefficient (e.g., 1%), corrosion rate ST (in millimeters per year), corrosion rate LT (in millimeters per year), remaining life of the component (in years), remaining half life (also in years), and the actual thickness (in millimeters). - In addition,
FIG. 19B is a rectified graph showing FSH (in percentage) values over thickness values (in millimeters or other units). The graph also illustrates the thickness range ofGate A 1904 andGate B 1906. Alternatively, the graph may chart the mV value as a substitute for FSH. Moreover, in some examples, the graph may be displayed as HF instead of rectified. -
FIG. 20A is agraph 2002 of an acid battery facility showing the measured TC thickness in millimeters of a component over time. Alternatively, the graph may show the values of temperature calibration (in Celsius), temperature coefficient (e.g., 1%), corrosion rate ST (in millimeters per year), corrosion rate LT (in millimeters per year), remaining life of the component (in years), remaining half life (also in years), and the actual thickness (in millimeters). In addition,FIG. 20B is a rectified graph of an acid battery facility showing FSH (in percentage) values over thickness values (in millimeters or other units). The graph also illustrates the thickness range ofGate A 2004 andGate B 2006. Alternatively, the graph may chart the mV value as a substitute for FSH. Moreover, in some examples, the graph may be displayed as HF instead of rectified. -
FIG. 21 illustrates a simplified example of an artificialneural network 2100 on which a machine learning algorithm may be executed.FIG. 21 is merely an example of nonlinear processing using an artificial neural network; other forms of nonlinear processing may be used to implement a machine learning algorithm in accordance with features described herein. - In
FIG. 21 , each of input nodes is connected to a first set of processing nodes. Theexternal source 2102 which is fed into input nodes may be the metrics from the results through the steps of the methods disclosed herein. Each of the first set of processing nodes is connected to each of a second set of processing nodes. Each of the second set of processing nodes is connected to each of output nodes. Though only two sets of processing nodes are shown, any number of processing nodes may be implemented. Similarly, though only four input nodes, five processing nodes, and two output nodes per set are shown inFIG. 21 , any number of nodes may be implemented per set. Data flows inFIG. 21 are depicted from left to right: data may be input into an input node, may flow through one or more processing nodes, and may be output by an output node. Input into the input nodes may originate from anexternal source 2102.Output 2104 may be sent to afeedback system 2106 and/or to data store. Thefeedback system 2106 may send output to the input nodes for successive processing iterations with the same or different input data. - In one illustrative method using
feedback system 2106, the system may use machine learning to determine an output. The output may include a leak area boundary, a multi-sensor detection event, confidence values, and/or classification output. The system may use an appropriate machine learning model including xg-boosted decision trees, auto-encoders, perceptron, decision trees, support vector machines, regression, and/or a neural network. The neural network may be an appropriate type of neural network including a feed forward network, radial basis network, recurrent neural network, long/short term memory, gated recurrent unit, auto encoder, variational autoencoder, convolutional network, residual network, Kohonen network, and/or other type. In one example, the output data in the machine learning system may be represented as multi-dimensional arrays, an extension of two-dimensional tables (such as matrices) to data with higher dimensionality. - The neural network may include an input layer, a number of intermediate layers, and an output layer. Each layer may have its own weights. The input layer may be configured to receive as input one or more feature vectors described herein. The intermediate layers may be convolutional layers, pooling layers, dense (fully connected) layers, and/or other types. The input layer may pass inputs to the intermediate layers. In one example, each intermediate layer may process the output from the previous layer and then pass output to the next intermediate layer. The output layer may be configured to output a classification or a real value. In one example, the layers in the neural network may use an activation function such as a sigmoid function, a Tanh function, a ReLu function, and/or other functions. Moreover, the neural network may include a loss function. A loss function may, in some examples, measure a number of missed positives; alternatively, it may also measure a number of false positives. The loss function may be used to determine error when comparing an output value and a target value. For example, when training the neural network, the output of the output layer may be used as a prediction and may be compared with a target value of a training instance to determine an error. The error may be used to update weights in each layer of the neural network.
- In one example, the neural network may include a technique for updating the weights in one or more of the layers based on the error. The neural network may use gradient descent to update weights. Alternatively, the neural network may use an optimizer to update weights in each layer. For example, the optimizer may use various techniques, or combination of techniques, to update weights in each layer. When appropriate, the neural network may include a mechanism to prevent overfitting—regularization (such as L1 or L2), dropout, and/or other techniques. The neural network may also increase the amount of training data used to prevent overfitting.
- In one example,
FIG. 21 depicts nodes that may perform various types of processing, such as discrete computations, computer programs, and/or mathematical functions implemented by a computing device. For example, the input nodes may comprise logical inputs of different data sources, such as one or more data servers. The processing nodes may comprise parallel processes executing on multiple servers in a data center. And, the output nodes may be the logical outputs that ultimately are stored in results data stores, such as the same or different data servers as for the input nodes. Notably, the nodes need not be distinct. For example, two nodes in any two sets may perform the exact same processing. The same node may be repeated for the same or different sets. - Each of the nodes may be connected to one or more other nodes. The connections may connect the output of a node to the input of another node. A connection may be correlated with a weighting value. For example, one connection may be weighted as more important or significant than another, thereby influencing the degree of further processing as input traverses across the artificial neural network. Such connections may be modified such that the artificial
neural network 2100 may learn and/or be dynamically reconfigured. Though nodes are depicted as having connections only to successive nodes inFIG. 21 , connections may be formed between any nodes. For example, one processing node may be configured to send output to a previous processing node. - Input received in the input nodes may be processed through processing nodes, such as the first set of processing nodes and the second set of processing nodes. The processing may result in output in output nodes. As depicted by the connections from the first set of processing nodes and the second set of processing nodes, processing may comprise multiple steps or sequences. For example, the first set of processing nodes may be a rough data filter, whereas the second set of processing nodes may be a more detailed data filter.
- The artificial
neural network 2100 may be configured to effectuate decision-making. As a simplified example for the purposes of explanation, the artificialneural network 2100 may be configured to detect faces in photographs. The input nodes may be provided with a digital copy of a photograph. The first set of processing nodes may be each configured to perform specific steps to remove non-facial content, such as large contiguous sections of the color red. The second set of processing nodes may be each configured to look for rough approximations of faces, such as facial shapes and skin tones. Multiple subsequent sets may further refine this processing, each looking for further more specific tasks, with each node performing some form of processing which need not necessarily operate in the furtherance of that task. The artificialneural network 2100 may then predict the location on the face. The prediction may be correct or incorrect. - The
feedback system 2106 may be configured to determine whether or not the artificialneural network 2100 made a correct decision. Feedback may comprise an indication of a correct answer and/or an indication of an incorrect answer and/or a degree of correctness (e.g., a percentage). For example, in the facial recognition example provided above, thefeedback system 2106 may be configured to determine if the face was correctly identified and, if so, what percentage of the face was correctly identified. The feedback system may already know a correct answer, such that the feedback system may train the artificialneural network 2100 by indicating whether it made a correct decision. The feedback system may comprise human input, such as an administrator telling the artificialneural network 2100 whether it made a correct decision. The feedback system may provide feedback (e.g., an indication of whether the previous output was correct or incorrect) to the artificialneural network 2100 via input nodes or may transmit such information to one or more nodes. The feedback system may additionally or alternatively be coupled to the storage such that output is stored. The feedback system may not have correct answers at all, but instead base feedback on further processing: for example, the feedback system may comprise a system programmed to identify faces, such that the feedback allows the artificialneural network 2100 to compare its results to that of a manually programmed system. - The artificial
neural network 2100 may be dynamically modified to learn and provide better input. Based on, for example, previous input and output and feedback from thefeedback system 2106, the artificialneural network 2100 may modify itself. For example, processing in nodes may change and/or connections may be weighted differently. Following on the example provided previously, the facial prediction may have been incorrect because the photos provided to the algorithm were tinted in a manner which made all faces look red. As such, the node which excluded sections of photos containing large contiguous sections of the color red could be considered unreliable, and the connections to that node may be weighted significantly less. Additionally, or alternatively, the node may be reconfigured to process photos differently. The modifications may be predictions and/or guesses by the artificialneural network 2100, such that the artificialneural network 2100 may vary its nodes and connections to test hypotheses. - The artificial
neural network 2100 need not have a set number of processing nodes or number of sets of processing nodes, but may increase or decrease its complexity. For example, the artificialneural network 2100 may determine that one or more processing nodes are unnecessary or should be repurposed, and either discard or reconfigure the processing nodes on that basis. As another example, the artificialneural network 2100 may determine that further processing of all or part of the input is required and add additional processing nodes and/or sets of processing nodes on that basis. - The feedback provided by the
feedback system 2106 may be mere reinforcement (e.g., providing an indication that output is correct or incorrect, awarding the machine learning algorithm a number of points, or the like) or may be specific (e.g., providing the correct output). For example, the machine learning algorithm may be asked to detect faces in photographs. Based on an output, the feedback system may indicate a score (e.g., 75% accuracy, an indication that the guess was accurate, or the like) or a specific response (e.g., specifically identifying where the face was located). In one example, a human operator/inspector may focus the inspection on a down-selected list of TMLs. Once any localized corrosion has been confirmed and repaired, a human operator may indicate as much so that themodel 1412 can be updated to reflect the new wall thickness values. In addition, in some examples, a localized corrosion may be erroneously identified in thesystem 1400, and supervised human input into a machine learning or neural network executing in thedigital analytics platform 112 may refine its alerts and model accordingly. - The artificial
neural network 2100 may be supported or replaced by other forms of machine learning. For example, one or more of the nodes of artificialneural network 2100 may implement a decision tree, associational rule set, logic programming, regression model, cluster analysis mechanisms, Bayesian network, propositional formulae, generative models, and/or other algorithms or forms of decision-making. The artificialneural network 2100 may effectuate deep learning. -
FIG. 22 is a flowchart showing illustrative steps of amethod 2200 performed in accordance with some embodiments disclosed herein. Themethod 2200 may be performed by asystem 1400 when computer-executable instructions, which are stored in a non-transitory computer-readable medium, are executed by a processor. Themethod 2200 may, among other things, down-select from among probe assemblies installed on a piping system in an industrial facility. As a result, the system and method for optimized asset health monitoring is improved because representative measurement locations are identified through down selection and the remaining probe assemblies can be disregarded during routine inspections of the piping system and other components in an industrial facility. - Regarding
FIG. 22 , instep 2202, the system is storing, in acomputer memory 1406 communicatively coupled to theprocessor 1402, historical pipe wall thickness measurements collected over a period of time from theprobe assemblies 150A installed on the piping system in the industrial facility. Instep 2204, thedata analytics platform 112 may set one or more hyperparameters, such as but not limited to a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, a group_size hyperparameter, and/or combination thereof. Once the hyperparameters are set, instep 2206, thesystem 1400 may begin training the model with at least the historical pipe wall thickness measurements stored in thecomputer memory 1406 and hyperparamter values stored incomputer memory 1404. - In
step 2208, the model stored incomputer memory 1404 may group a first set of the probe assemblies from among the numerous probe assemblies installed on the piping system. As explained in this disclosure, such as with respect toFIG. 15 , several methods are provided by which the grouping may occur in the model. After grouping the probe assemblies, thedata analytics platform 112 may assign a unique group identifier (groupID) to each set of probe assemblies. The unique groupID may be any identifier that thesystem 1400 can use to uniquely refer to the group of probe assemblies. - In
step 2210, thedata analytics platform 112 selects, based on at least the trained model, an optimization function for the operation of thesystem 1400. Numerous illustrative optimization functions are described in this disclosure, including but not limited to a median_TML_within_groupID optimization function, minimum_average_TML_within_groupID optimization function, minimum_variation_from_mean optimization function, and/or TML_position optimization function. The decision to select a specific optimization function causes subsequent identification and measurement steps to be effected. For example, insteps system 1400 identifies, based on the model stored in thecomputer memory 1404 and selected optimization function, a probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system. In some examples, thesystem 1400 may identify a single probe assembly for the entire groupID to be representative of the area being measured. In other examples, multiple probe assemblies may be identified to be representative of the groupID. The number of TMLs assigned to be down-selected from a group may be based on one or more rules. This is based on the value of the maximum standard deviation of the group, in one example: if any group has a maximum standard deviation less than or equal to 0.25, then one TML is chosen from the group. for an increase in max SD by 0.25, the number of TMLs selected increases by one (i.e., if it is between 0.25-0.5, then two probe assemblies may be chosen from the group, and so on). In addition, the 0.25 step value can be modified for adjusting the sensitivity of TML selection. Decreasing the 0.25 value leads to more TMLs being selected in each group (i.e., a conservative approach), and increasing the 0.25 value leads to less TMLs being selected in each group (i.e., an aggressive approach). - In
step 2214, during the inspection, thesystem 1400 may disregard all remaining probe assemblies in each groupID except the probe assembly identified from each groupID. The system may measure the wall thickness of each of the identified probe assemblies for each groupID, but exclude the other probe assemblies in the groupID. Thus, the system down-selects from among the plurality of probe assemblies installed on a piping system. At least one benefit of down-selecting the number of probe assemblies to use during an inspection is the time savings that results. For example, a human inspector that might have previously checked each probe assembly may now check measurements at a reduced number of probe assemblies without substantially increasing the risk of missing dangerous localized corrosion. In one example, atstep 2214, thesystem 1400 may output a human-readable report listing those probe assemblies a human inspector should manually inspect for wall thickness measurements. The output may be ordered in any of various ways—e.g., based on highest risk of localized corrosion, based on geographic convenience from a known start position of the human inspector, or other order. - For example, as illustrated in
FIG. 16B andFIG. 16C , when the amount of risk is graphed against the quantity of measurement locations taken, the amount of risk is asymptotic to a threshold minimum amount ofrisk 1602 regardless of the number of measurement locations increases. Importantly, when the quantity of measurements is decreased, the delta change in risk increases at an increasing rate as shown bygraph 1604—stated another way, reducing the number of probe assemblies that are sampled can increase the risk to an unsafe amount. Thesystem 1400 andmethod 2200 disclosed herein, however, shifts the graph from aninitial risk graph 1606 to a morefavorable risk graph 1608. Therefore, down-selecting the number of probe assemblies required to have actively checked during an inspection, by identifying those that are the most statistically probable to be general corrosion/degradation to the piping wall, results in reduced inspection time/cost while simultaneously maintaining (or even reducing) the risk profile. - Finally, in
step 2216 inFIG. 22 , thethickness monitoring controller 130 may receive and send the pipe wall thickness measurement of the probe assembly from each groupID for inspection. Thethickness monitoring controller 130 may send the measurement data (and any other data) to thedata store 1406 for historical recordkeeping and analytics, as well as to thedata analytics platform 112 for analysis and generation of visualizations. For example, the wall thickness measurements may show that a particular segment of pipe in the piping system is suffering from degradation other than general corrosion such that it rises to the level of dangerous, localized corrosion and must be replaced within a particular period of time. In another example, pipe wall thickness measurements may be taken at one or more of a pipe, tank, vessel, and/or pipeline at a facility. - While particular embodiments are illustrated in and described with respect to the drawings, it is envisioned that those skilled in the art may devise various modifications without departing from the spirit and scope of the appended claims. It will therefore be appreciated that the scope of the disclosure and the appended claims is not limited to the specific embodiments illustrated in and discussed with respect to the drawings and that modifications and other embodiments are intended to be included within the scope of the disclosure and appended drawings. Moreover, although the foregoing descriptions and the associated drawings describe example embodiments in the context of certain example combinations of elements and/or functions, it should be appreciated that different combinations of elements and/or functions may be provided by alternative embodiments without departing from the scope of the disclosure and the appended claims. Further, the foregoing descriptions describe methods that recite the performance of several steps. Unless stated to the contrary, one or more steps within a method may not be required, one or more steps may be performed in a different order than as described, and one or more steps may be formed substantially contemporaneously. Various aspects are capable of other embodiments and of being practiced or being carried out in various different ways. It is to be understood that the phraseology and terminology used herein are for the purpose of description and should not be regarded as limiting. Rather, the phrases and terms used herein are to be given their broadest interpretation and meaning. The use of “including” and “comprising” and variations thereof is meant to encompass the items listed thereafter and equivalents thereof as well as additional items and equivalents thereof
Claims (20)
1. A method for down-selecting from among probe assemblies installed on a piping system, wherein the probe assemblies are configured for pipe wall thickness monitoring, the method comprising:
setting a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter for a model, before training the model;
grouping, by the model executing on a processor, a first set of the probe assemblies based at least on historical pipe wall thickness measurements collected from the probe assemblies installed on the piping system over a period of time;
assigning a unique groupID to each set of probe assemblies;
selecting, by the model after training the model, an optimization function from among a plurality of optimization functions for the model;
identifying, by the model, a single probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system; and
sending, by a thickness monitoring controller associated with the piping system, a pipe wall thickness measurement of the single probe assembly from each groupID for inspection.
2. The method of claim 1 , wherein the probe assemblies comprise at least a resistance temperature detector, a thickness monitoring ultrasonic transducer, and an area monitoring ultrasonic transducer configured to detect localized corrosion in the piping system.
3. The method of claim 2 , further comprising:
validating that the pipe wall thickness measurement of the single probe assembly is general corrosion and not localized corrosion by:
generating a probability plot of all pipe wall thickness measurements associated with the piping system;
grouping the plotted pipe wall thickness measurements by nominal thickness; and
failing to identify a non-linear relationship in the probability plot of pipe wall thickness measurements grouped by nominal thickness to confirm the general corrosion.
4. The method of claim 1 , further comprising:
during the inspection, down-selecting by disregarding all remaining probe assemblies in each groupID except the single probe assembly from each groupID to reduce a number of inspection samples measured without compromising a risk profile of the piping system.
5. The method of claim 1 , wherein the grouping of the first set of the probe assemblies is further based at least on inspection information provided to the system and historical pipe wall thickness measurements collected over a period of time from the probe assemblies installed on the piping system.
6. The method of claim 1 , wherein the plurality of optimization functions comprises median_TML_within_groupID, minimum_average_TML_within_groupID, and minimum_variation_from_mean.
7. The method of claim 6 , wherein the plurality of optimization functions comprises TML_position.
8. The method of claim 1 , wherein the piping system comprises a tank, and wherein a first probe assembly of the probe assemblies is configured to measure a wall thickness of the tank.
9. The method of claim 1 , wherein the pipe wall thickness monitoring comprises measuring the thickness of pipe wall at a specific probe assembly, wherein the pipe wall is located at one or more of a pipe, tank, vessel, and pipeline.
10. The method of claim 9 , wherein the pipe wall thickness monitoring comprises, by the probe assemblies, analyzing the original wall thicknesses, wall thickness loss over time, calibration error, and measurement location repeatability error.
11. The method of claim 1 , further comprising:
storing, in computer memory communicatively coupled to the processor, historical pipe wall thickness measurements collected over an extended period of time from the probe assemblies installed on the piping system; and
training, by the processor, the model with at least the historical pipe wall thickness measurements stored in the computer memory.
12. The method of claim 11 , wherein the model comprises an artificial neural network.
13. A system for detecting localized corrosion to a plurality of components that transport materials across a distance, the system comprising:
a plurality of probe assemblies affixed to one or more of the components, wherein each of the plurality of probe assemblies corresponds to a unique identifier;
a data store configured to store historical wall thickness measurements collected over a period of time from measurements performed by the probe assemblies;
a model trained on the historical wall thickness measurements in the data store and with hyperparameters comprising at least a grouping_sensitivity hyperparameter; and
a monitoring apparatus comprising a processor and a memory storing computer-executable instructions that, when executed by the processor, cause the system to perform steps comprising:
grouping, based on the model, a first set of the probe assemblies;
assigning a unique groupID to each set of probe assemblies;
selecting, based on the model, an optimization function from among a plurality of optimization functions;
identifying, based on the model and selected optimization function, a probe assembly for each groupID for wall thickness monitoring of the components, wherein each groupID corresponds to the unique identifier corresponding to the identified probe assembly; and
outputting a list of the unique identifiers corresponding to any groupID.
14. The system of claim 13 , wherein the plurality of probe assemblies comprise at least a thickness monitoring ultrasonic transducer and an area monitoring ultrasonic transducer configured to detect localized corrosion to the components, wherein the probe assembly identified from each groupID comprises more than one probe assembly of the plurality of probe assemblies, and wherein the memory of the monitoring apparatus stores computer-executable instructions that, when executed by the processor, cause the system to perform steps comprising:
sending, by a thickness monitoring controller associated with the components, a wall thickness measurement of the probe assembly from each groupID for inspection;
during the inspection, disregarding all remaining probe assemblies in each groupID except the more than one probe assembly from each groupID; and
validating that the wall thickness measurements of the more than one probe assembly from each groupID fails to identify general corrosion.
15. The system of claim 13 , wherein the wall thickness measurement of the probe assembly from a first groupID comprises a thickness of a wall of a pipe component at the probe assembly.
16. The system of claim 13 , wherein the wall thickness measurement of the probe assembly from a first groupID comprises a thickness of a wall of a tank component at the probe assembly.
17. The system of claim 13 , wherein the hyperparameters comprise a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter, and wherein the plurality of optimization functions comprises median_TML_within_groupID, minimum_average_TML_within_groupID, minimum_variation_from_mean, and TML_position.
18. A non-transitory computer-readable medium storing computer-executable instructions that, when executed by a processor, cause a system to down-select from among probe assemblies installed on a piping system, by performing steps comprising:
storing, in a computer memory communicatively coupled to the processor, historical pipe wall thickness measurements collected over a period of time from the probe assemblies installed on the piping system;
setting a hyperparameter for a model;
training, by the processor, the model with at least the historical pipe wall thickness measurements stored in the computer memory;
grouping, by the model executing on the processor, a first set of the probe assemblies;
assigning a unique groupID to each set of probe assemblies;
selecting, based on the model, an optimization function from among a plurality of optimization functions;
identifying, based on the model and selected optimization function, a probe assembly corresponding to each groupID for pipe wall thickness monitoring of the piping system; and
sending, by a thickness monitoring controller associated with the piping system, a pipe wall thickness measurement of the probe assembly from each groupID for inspection.
19. The non-transitory computer-readable medium of claim 18 , wherein the hyperparameters comprise at least one of a grouping_sensitivity hyperparameter, a threshold_measurements hyperparameter, and a group_size hyperparameter; and wherein the plurality of optimization functions comprises median_TML_within_groupID, minimum_average_TML_within_groupID, minimum_variation_from_mean, and TML_position.
20. The non-transitory computer-readable medium of claim 18 , further storing computer-executable instructions that, when executed by the processor, cause the system to perform steps comprising:
during the inspection, disregarding all remaining probe assemblies in each groupID except the probe assembly identified from each groupID.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US18/265,433 US20240036011A1 (en) | 2020-02-28 | 2020-12-08 | System and method for corrosion and erosion monitoring of pipes and vessels |
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US202062982751P | 2020-02-28 | 2020-02-28 | |
US18/265,433 US20240036011A1 (en) | 2020-02-28 | 2020-12-08 | System and method for corrosion and erosion monitoring of pipes and vessels |
PCT/US2020/063683 WO2022125059A1 (en) | 2020-12-08 | 2020-12-08 | System and method for corrosion and erosion monitoring of pipes and vessels |
Publications (1)
Publication Number | Publication Date |
---|---|
US20240036011A1 true US20240036011A1 (en) | 2024-02-01 |
Family
ID=77490761
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US18/265,433 Pending US20240036011A1 (en) | 2020-02-28 | 2020-12-08 | System and method for corrosion and erosion monitoring of pipes and vessels |
US17/800,545 Pending US20230064270A1 (en) | 2020-02-28 | 2021-03-01 | System and method for corrosion and erosion monitoring of fixed equipment |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/800,545 Pending US20230064270A1 (en) | 2020-02-28 | 2021-03-01 | System and method for corrosion and erosion monitoring of fixed equipment |
Country Status (5)
Country | Link |
---|---|
US (2) | US20240036011A1 (en) |
EP (1) | EP4111172A4 (en) |
CN (1) | CN115380199A (en) |
CA (1) | CA3166618A1 (en) |
WO (1) | WO2021171273A1 (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115614678B (en) | 2022-12-16 | 2023-03-10 | 成都秦川物联网科技股份有限公司 | Intelligent gas pipeline electrochemical corrosion assessment method, internet of things system and medium |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4398421A (en) * | 1981-12-23 | 1983-08-16 | Hartford Steam Boiler Inspection And Insurance Company | Ultrasonic thickness measuring apparatus and method |
JP3747921B2 (en) * | 2003-06-20 | 2006-02-22 | 株式会社日立製作所 | Nondestructive inspection apparatus and nondestructive inspection method using guide wave |
GB2437547B (en) * | 2006-04-28 | 2010-07-14 | Genesis Oil And Gas Consultant | Method and apparatus for inspecting pipes |
JP5530405B2 (en) * | 2011-07-25 | 2014-06-25 | 日立Geニュークリア・エナジー株式会社 | Nondestructive inspection method and nondestructive inspection device |
WO2014190268A1 (en) * | 2013-05-24 | 2014-11-27 | The Penn State Research Foundation | Flexural modes in non-destructive testing and inspection |
EP2995942A1 (en) * | 2014-09-10 | 2016-03-16 | Shell Internationale Research Maatschappij B.V. | A pipe fault detection system |
-
2020
- 2020-12-08 US US18/265,433 patent/US20240036011A1/en active Pending
-
2021
- 2021-03-01 CN CN202180017534.3A patent/CN115380199A/en active Pending
- 2021-03-01 CA CA3166618A patent/CA3166618A1/en active Pending
- 2021-03-01 WO PCT/IB2021/051697 patent/WO2021171273A1/en unknown
- 2021-03-01 US US17/800,545 patent/US20230064270A1/en active Pending
- 2021-03-01 EP EP21760344.8A patent/EP4111172A4/en active Pending
Also Published As
Publication number | Publication date |
---|---|
CA3166618A1 (en) | 2021-09-02 |
US20230064270A1 (en) | 2023-03-02 |
EP4111172A1 (en) | 2023-01-04 |
WO2021171273A1 (en) | 2021-09-02 |
EP4111172A4 (en) | 2024-01-24 |
CN115380199A (en) | 2022-11-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10909280B2 (en) | Methods and systems for structural health monitoring | |
US20220205967A1 (en) | Systems and methods for placing networked sensors within a facility for fugitive emissions monitoring | |
US8494810B2 (en) | Component adaptive life management | |
Hashemian | State-of-the-art predictive maintenance techniques | |
RU2533321C1 (en) | Method for adaptive forecasting of residual operating life of complex objects, and device for its implementation | |
US20240036011A1 (en) | System and method for corrosion and erosion monitoring of pipes and vessels | |
Lee et al. | An enhanced prediction model for the on-line monitoring of the sensors using the Gaussian process regression | |
CN113916673B (en) | Safety early warning method and system based on container state monitoring | |
Schoefs et al. | The αδ method for modelling expert judgement and combination of non-destructive testing tools in risk-based inspection context: application to marine structures | |
EP4260057A1 (en) | System and method for corrosion and erosion monitoring of pipes and vessels | |
CN116437243B (en) | Remote diagnosis method for natural gas metering analysis equipment | |
JP2024507920A (en) | System and method for continuously monitoring the condition of nuclear reactor internals | |
Rastegaev et al. | Estimating the probability of detecting a delamination in the wall of equipment depending on the set of used methods of nondestructive testing and ways of its improvement | |
JP5150302B2 (en) | Ultrasonic inspection data evaluation apparatus and ultrasonic inspection data evaluation method | |
Gerards-Wünsche et al. | A framework for assessing the reliability of crack luminescence: an automated fatigue crack detection system | |
RU2742282C1 (en) | Method of controlling parametric fail-safety of articles by state parameters | |
CA3100826C (en) | Condition monitoring of ultrasonic transducers and probes | |
Park et al. | Stochastic Remaining Fatigue Life Assessment Considering Crack Inspection Results | |
Araseethota Manjunatha et al. | Diagnosis of Corrosion Process in Nuclear Power Plant Secondary Piping Structures | |
KR20230167389A (en) | Intelligent Pipe Inspection Machine | |
WO2022033992A1 (en) | Autonomous penetrant testing | |
CN118190074A (en) | Multi-parameter on-line monitor for tank farm storage tank | |
Mead et al. | Investigation into maritime military platform hull defects automation of sensors and processing: Research-in-progress | |
Fields | In-service pressure vessel inspections | |
Badalyan et al. | Data-processing algorithms for automatic operation of ultrasonic systems with coherent data processing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
AS | Assignment |
Owner name: MOLEX, LLC, ILLINOIS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SCHIEKE, SASCHA;LUTOLF-CARROLL, DANIEL;SIGNING DATES FROM 20230815 TO 20230816;REEL/FRAME:065999/0612 |