· Home · Table of Contents · Industrial Plant & Structures

Detection and Measurement of Reformer Tube Creep Don Searle Technology Consultant, pndt pty.ltd. Australia. Contact

Abstract

This paper concerns the BRUCE technique: Boiler and Reformer Ultrasonic Creep Evaluation.

Creep, arising as a consequence of the high operating temperatures and stress, primarily governs the life of catalyst tubes in gas reformer service. In addition to the pressure stress, thermal stresses are introduced during plant start-up, shut-down and upsets. It is the decay of these stresses at high temperature during operation, by the creep process, that results in creep damage and reduced tube life. The more rapid the rate of temperature change during start-up, shut-down or upset conditions, the higher the magnitude of thermal stress introduced into the catalyst tube and hence greater the degree of creep damage which occurs during subsequent operation. Higher operating temperatures will also increase the rate of creep damage accumulation.

Creep is a progressive damage process, which produces a physical change in diameter of the catalyst tube with increased service time. As creep progresses, voids will nucleate at grain boundaries within the material and with further service exposure the creep process initiates micro-cracks in the tube. In catalyst tubes, the thermal stress is maximised close to the internal surface and it is at this location that the creep damage starts. Once cracks form they will subsequently propagate by creep from the internal to external surface of the tube.

Initiation and growth of creep induced cracks from the internal sub-surface of the catalyst tube clearly makes Non Destructive Evaluation (NDE) more difficult. However, it has been recognised for some time, that monitoring changes in catalyst tube diameter and hence creep strain, will provide an indication of the tube condition. Catalyst tubes have a relatively low creep ductility and depending on the material type, strain levels in excess of 3 to 4% can be indicative of exhausted creep life and therefore this level of plastic deformation is a useful indicator for tube replacement.

In the past, creep strain measurement in catalyst tubes has been undertaken using micrometer or vernier based instruments in both static and dynamic formats. However, after studying the effectiveness of these earlier techniques, a reverse order approach was taken, in that the inspection system design firstly considered the detection of large cracks, then small cracks, then micro-cracks, then plastic deformation and finally a baseline representing material and construction constants. An "Ultrasonic based technique" to measure tube diameter, external and internal tube thickness and diametric values was developed. In addition to diametric measurements, the BRUCE system also simultaneously measures the creep damage accumulation by comparing the ultrasonic wave attenuation test to test. The relationship between ultrasonic attenuation and the level of creep damage was first determined by studies undertaken on reformer plants in Europe.

Combining these design criteria and insitu measurements, provided a more accurate means for the monitoring of the catalyst tube condition, managing risk and hence planning future tube replacement schedules.

Keywords: BRUCE Creep Damage Reformer RBI

Introduction

There have been a number of approaches taken to assess tube condition, the primary variable being largely driven by the available technology of the time.
pndt advanced inspection division, designed and built, a custom radio controlled robot, to establish a base line on which to determine the current condition of reformer tubes and a process to be put in place to define the on going evaluation of tube condition, creep and safe remaining life.
BRUCE is clearly a "Risk Based Inspection" tool.

OBJECTIVES

The criteria used to underpin the technical design of a suitable system was as follows:

• The system needs to measure deformation of the tube very accurately.
• The system needs to measure both outside diameter and internal diameter.
• The system needs to measure the wall thickness of the tube.
• The system needs to measure certain metallurgical properties as indicators of creep, particularly acoustic attenuation.
• The system needs to detect growth in number and size of microvoids.
• The system needs to detect cracking.
• The system needs to be operated in "Time of Flight Diffraction" mode for the purpose of detecting and evaluating cracks, their size, orientation and to utilise the ToFD permanent record display to overlay test result onto test result over time.

The technical design needed to cover the above criteria is represented by Fig 1.

Fig 1:

Critical "Controls" needed in the system design.


The technology needed to consider resolution.
Resolution of a 1mm change in the radius needed to be represented by 25 measurable increments. ie 1/25mm per increment or 0.04mm.

SOLUTION:
This was achieved by using a Roller Search Unit (RSU) filled with a medium which allowed the system to resolve a change in the tube external diameter four times more accurately than measuring the tube material. This was achieved by using a transmission medium with an acoustic velocity of 1565m/sec as compared to 5960 m/sec.

Demineralised water applied as a couplant needed to be minimized to prevent ingress into the furnace refractory.


SOLUTION:
The purposeful use of Roller Search Units (RSU's) hydraulically guarantees the interface and simultaneously provides an ultrasonic window with minimum use of water and potential spillage. Fig 2. represents the gauge, width of surface wetting, used to regulate and monitor the water consumption rate.

The method and equipment needed to be intrinsically safe and operable within the environment of a furnace, including effects of RF interference and electronic noise.


The entire system was 12V DC based

The system required flexibility to progress and scan over butt welds and surface irregularities.


SOLUTION:
This was not originally anticipated but by changing out the driven tyre specification to a pneumatic drive tyre, from that of a solid drive wheel system, the problem was overcome.

Fig 2:

TEST RESULT OBJECTIVES

• To compile a database of the baseline diametric tube profiles, by measuring the co-ordinates of the tube outer surface within the probe array and computing the outside diameter of the tube, at intervals of 10mm from the bottom of the reformer tube to the top.
• To compile a database of tube thickness variations from nominal at intervals of 10mm along each tube, similarly from the bottom of the reformer tube to the top.
• To determine by calculation and create a database of the internal diameter by subtracting the pipe thicknesses across the diametric co-ordinates from the outside pipe diameter.
• To compile an attenuation profile from transducer to transducer using "echo dynamics technology" and to scale the results 0-100% along the tube length at intervals of 10mm.
• In the event of any creep observation being made from the diametric measurements, the system would perform at the measured location a pulse echo crack detection scan and or perform a ToFD scan of the tube.
• In order to carry out this inspection "pndt" within the terms of an intellectual property rights agreement with AEA Sonomatic, utilised a "Microplus" ultrasonic generator/receiver and data collection system. The system software was programmed by pndt technologists specifically for the BRUCE project.

DIAMETRIC MEASUREMENTS

Fig 3: Represents the format of the diametric tube measurements

REPRESENTATIONS and FORMAT

Measuring position and channel number.

The four plots are scaled against a Y axis which represents the unit variation of the quadrant on each radius.no.s 1-4
25 units represents 4% creep or 4% change in diameter and each unit is dividable by 20.

Calibration

A sample creep calibration display is attached as Fig 5.

Fig 5: Cross section of the Reformer.

The method to confirm calibration is to insert a 0.66mm shim between the RSU and the tube outer surface which in effect compresses the liquid transit time between the transducer and the interface by that amount

TUBE THICKNESS DETERMINATION See fig 4

Fig 4:

Simultaneous to the diametric measurement, the tube thickness is determined. A separate graphical representation can be made of the tube thickness by changing the resolution scale to the equivalent of 20mm of the spun cast material. This step of producing B scan images is only necessary where creep is noted from the diametric data.

A further facility is available in the software package and this is the ability to convert the scan images into text data, where each interface is posted as a physical measurement. When imported into a spreadsheet format the information can be mathematically treated for trending and comparison test to test and levels of change can be calculated and expressed in R squared format. The image data is converted by a software program called scandata.exe

The tube thickness requires that the range be extended to cover 0 - 200 on the Y axis. The section thickness definition is impaired due to the outer surface rough profile that is the result of the spun casting technique used to manufacture the tubes.

Scandata exe can also be run on the file to produce in spreadsheet format the interrupt distances between the OD and the ID interfaces.

Table No 1. displays the text format after conversion from the b scan image using the scandata .exe file sequence

SCANNING PROCEDURE

Access to the reformer tubes is generally via the side entry doors, which allow access along the tunnel roof. Scaffolding is placed on the roof to protect the refractory. This platform forms the basis of connecting BRUCE to each of the reformer tubes. A radio control system is used to direct the scanner between the top and the bottom of the reformer including the velocity at which the unit is to scan.

Fig 6: represents the scanning robot being retrieved from the scan start position by remote control

Results

The results obtained with the BRUCE system are within the expectation of the design.

The system satisfactorily produced

• A graphical and digital representation of each of the longitudinal diametric values of the tube quadrants.These results clearly display the profile of the tube diameters. The system clearly defined dynamic irregularities of the scanner by any change in one radius with a simultaneous change in the corresponding radius.
  If both diametrics change to the negative the tube is in a creep condition and will be flagged as such for echo dynamic studies and for verification of growth by the next inspection.
• A table of exceptions are obtainable where the radial extensions corresponded with corresponding wall thickness anomalies.
• An echo dynamics plot is easily obtained, to relate attenuation versus creep damage. Echo dynamics is important when the attenuation values across the tube quadrants exceed 16 dB from the nominal.
• A database of primary interface measurements allowing various mathematical models to be carried out to identify and to create an index for creep damage of the reformer tubes.

Acknowledgements:

The author wishes to acknowledge the assistance provided by the "Mechanical Engineering Team" at BHP HBI Boodarie W.A.

© AINDT , created by NDT.net

|Home|    |Top|