Reformer Tube Inspection

and
Remaining Life Assessment as per
API 579-1/ASME FFS-1

by: Ketan Upadhyaya

TCR Advanced
Vadodara
Basics Quick view on reformer heater tube operation
Reformer tubes are used for cracking hydrocarbon (natural gas) to
produce hydrogen and carbon dioxide either in presence of steam
(called steam reformer) or in presence of oxygen (auto-thermal
reforming).

Steam reforming reaction:

CH4 + H2O ⇌ CO + 3 H2 (700 – 1000°C)

Auto-thermal reforming reaction:

2CH4 + ½O2 + CO2 → 4H2 + 3CO + H2O (900 – 1000°C)

Hydrogen generation is an endothermic reaction, takes place in

presence of catalysts (nickel monoxide).
RLA Reformer tube inspection
• The reformer tube life is influenced by operating temperature

• It is experienced that 10 to 20°C overshoot can lower the tube

life substantially.

• Reformer tubes are therefore one of the most focused

inspection activities during the turnarounds.

• The increased demand for production requires higher flow

rates. Shorter reaction time of cracking necessitates increased
furnace temperature.
Primary reformer materials

• Reformer tubes (HPMA, G4852, KHR35CT, HP39W)

• Inlet header (P22, 321H, 347H)
• Inlet tube/pigtail/hairpin (P22, 321H, 347H)
• Outlet tube/pigtail/hairpin (800H)
• Hot outlet header (800H)
• Cold outlet header (Low alloy steel, P22, internal
refractory lined)

4
Reformer tube failures

• Failures due to creep rupture

• Short term overheating and bowing
• Thermal shocks
• Internal carburization and metal dusting
• Metallurgical degradations
• Design aspect and support system

5
 Creep (Elevated service temperature components)

Creep failures occur in hot-section components as a

result of continuous exposure of high temperature
and stress during operation.

Unacceptable dimensional changes, creep rupture

and local failure by creep-crack growth constitute
failures.

6
SEM creep damage

7
Bowing of reformer tube

• As the reformer furnace operation is taken at high temperature and

pressure due to thermal expansion and contraction at elevated
temperature and due to prolonged exposure to high temperature the
microstructure of the material is subjected to degraded and bending or
bowing of the tubes occur.

8
Bowing of reformer tube

Schematic drawing of bowing of reformer tube

Acceptable up to 1.5XOD as long as not wedging (OISD) 9

 Baldness in reformer tubes (Oxidation)

• Corrosion and oxidation resistance of most materials conventionally used in

steam cracking industry relies on formation of a protective chromia (Cr2O3)
layer.

• It is noticed that the scale formed on 25Cr-35Ni has a rapid initial growth rate
, but at a certain point, intrinsic stresses causes the scale to spall off. New
oxide is formed on the spalled areas, and chromium from the alloy is rapidly
consumed.

10
Baldness in reformer tubes Fuel / air ratio

Bald tube

Comparison between two reformer tubes for showing baldness

11
 Baldness: oxidation damage

• The carbides grow at the grain boundary with ageing.

• At the outer surface reacts with the oxygen in the furnace
environment, when the Chromium depletes because of the grain
boundary carbide precipitation, the oxides form on the outer
surface of tube.
• Thus the oxidation at the outer surface led the Cr depletion region
near outer surface.
• The variation of Cr concentration in the matrix from outer to the
inner edge of the service expose tube is generally observed to be
more at the outer surface.

12
 Internal carburization

• Internal carburization can be caused by process upset

• Severe carburization can lead to brittle failure

13
 Internal carburization

• Internal carburization can be caused by process upset

– Degradation of catalyst
– Imbalance of Feed and steam
– Upstream process upset
• Carburization extent: it is observed up to 50% of wall thickness (from ID
side)
• Carburized layer is not sound wall, and hence does not bear the internal
pressure
• Severe carburization can lead to brittle failure
– Brittle fracture can occur during idling time essentially during pressure test, DP
test for catalyst check
• Relative magnetic permeability measurements are helpful
• MOCs: Kubota developed AFTALLOY (Al and Si added proprietary grade)

14
Thermal shock

• Sudden change in temperature causes thermal shock that may

result in abrupt, catastrophic failure of tubes.

15
 Localized overheating
 Direct flame impingement:

• Mis-alignment of burners and / or their clogging creating disturbance in flue gas path
can lead to direct flame impingement to the tubes.
• As a result, tubes tend to damage locally by overheating.

16
 Deteriorated catalyst
 Overheating by choking of catalyst:

• The nickel monoxide catalyst used

in cracking of hydrocarbons is
generally in pellet form.
• Due to ageing or increased flow
rate, the catalysts undergo rubbing
and relative movement of particles
making them more friable.
• The fine particles settle down and
choke the gas path such that a
localized channel would get form
and passage of gas gets completely
resisted.

17
Metallurgical degradation
• The carbide coarsening mechanism is temperature and time dependent
phenomena. Higher the temperature faster is the coarsening of carbides.

• For temperature exposure between 700-800°C, the primary carbides tend to

get transformed from their eutectic morphology to compact blocks.

• Between 800-900°C the secondary carbides tend to coalesce and they get
reduced in number due to carbon diffusion to the primary carbides.

• When the material is exposed between 900-1000°C the secondary carbides

also tend to disappear over a Prolonged exposure and primary carbide appear
blocky, continuous in nature as shown in Figure 1. This results in reduction in
creep rupture strength

18
Metallurgical degradation

Primary carbides
Secondary carbides

Formation of compact blocky carbides With aging at 800°-900°C the secondary carbides
tend to coalesce and they get reduced in number
with diffusion to primary carbides.

19
Metallurgical degradation

Scanning electron micrograph:

With aging at 800°-900°C the secondary carbides tend to coalesce and they get
reduced in number with diffusion to primary carbides.
20
Remaining Life Assessment
RLA What is Remaining Life Assessment
The heater tubes are built with certain assumption on nominal
design and reasonable life of operation, generally about one or
two decades.

At end of this period tubes are either replaced or life is extended.

The tube replacement is generally planned when failure rate

increases.

What is important is: DAMAGE MECHANISM causing failure.

RLA Approaches
A fairly accurate remaining (residual) life assessment can be
performed for CREEP rupture failures through destructive tests.
This include Accelerated Stress Rupture test.

Similarly, Non-destructive Test approach for RLA is also adopted

based on metallurgical condition assessment of tube and
correlating the CREEP strength to applied stress.
RLA How is it derived
RLA is an output of quantitative figure derived from:

• Estimate of accumulated tube damage (the life fraction used

up) based on operating pressure, the tube metal temperature,
and the corrosion rate and the exposed service life.

• Available knowledge of the actual rupture strength of a given

tube, this is generally provided by the alloy manufacturer.

• In absence of parametric creep rupture data, accelerated stress

rupture tests may be adopted.
RLA Destructive test approach

Part I
Laboratory analyses
RLA Laboratory analyses

• Dimensional measurement
• Macrostructural observation
• Microstructural observation
• Accelerated stress rupture test
RLA Laboratory analyses

Dimensional measurement

Thickness measurement
– outer surface oxidation rate
- Inner surface metal dusting (carburization)

Increase in outer diameter

Creep Curve: Basic Concepts

Creep properties are generally determined by

means of a test in which a constant uniaxial
load or stress is applied to the specimen and
the resulting strain is recorded as a function of
time.

28
 RLA Laboratory analyses
Accelerated stress rupture test

• Similar to tensile test preparation, specimens are prepared and

subjected to tensile loading (constant load) at pre-selected
elevated temperature.

• Generally 2 or 3 tests are performed at different temperatures

and/or stress
Creep rupture test

• Usually a tensile bar

• Dead load applied

• Strain is plotted with time

• Test usually ends with rupture (failure)

30
Creep curves
Start Rupture
Stress

Time
Min Creep Rate
Strain

Time

Strain
Strain API579 method
Rate 1
𝐿=
ε𝑜 𝑥 𝜔

31
Mechanisms of Creep

32
RLA Laboratory analyses: ASR
As stress decreases, time to
rupture increases

As temperature decreases,
time to rupture increases

Stress
A fair degree of accuracy
can be gained with lower
stress values and higher test
temperature regimes

Time to failure OR LMP

Plotted when test
temperature is Plotted when test
constant temperature is varied
 RLA Laboratory analyses
Use of accelerated stress rupture test results
• Having multiple test results handy, prepare the stress v/s LMP
curve.
• Interpolate the actual (operating) stress and determine LMP
• Determine the residual time by averaging the trend on tube
metal temperature using LMP equation
LMP = (T + 273) x (C + Log10 t)
where, LMP is determined from stress rupture curve, T is
temperature in °C, t is the residual life in hours and ‘C’ is
material constant – generally between 20 to 30.
RLA NDT Approach

Part II
Shutdown inspection on reformer heater tubes and remaining life
assessment approach
Present scenario of inspection
During turnarounds generally following activities are targeted:

• Maintenance of burners and their alignments

• Maintenance of refractory, hot gas manifold chambers
• Visual inspection of reformer tubes
• Scanning for detection of internal fissure in reformer tubes
• Outer diameter measurement of tubes
• Random dye penetration tests of inlet / outlet pigtail joints
Present scenario of inspection
General criteria for tube replacement:

• Tube bowing condition

• Stresses increase at intrados

• Visual appearance / marking of localized heating

• Ultrasonic attenuation, e.g. tubes exhibiting highest dB value or
exceeding 64 to 70 dB
• Calculated creep strain based on outer diameter
measurements, tubes exceeding 3% are generally prioritized
Present scenario of inspection
Still missing the outcome of inspection

• What is remaining life ?

The answer lies to in-depth understanding of damage causing

mechanisms, responsible to tube failure
 Role of NDT to identify damages
Damage NDT method for detection Information available
mechanism
Creep damage Diametric measurements Often increased diameter by 3 to 6%.
Ultrasonic scanning for attenuation measurements Increased attenuation indicating presence of creep voids and / or
creep fissures
Eddy current testing Detection of creep fissures at later stage.
Metallography Creep fissures and void at primary carbide chains, detected
generally after 2nd stage damage, when surfaced out to OD

Radiography Direct film imaging can detect possible advanced stage of creep
fissures
Localized Pyrometers / Thermovision in operation Increased skin temperature either by flame / flue gas impingement
overheating or damage of catalysts leading to disturbed path of internal gases

Metallography Overheated structures around 1000°C for short duration would

dissolve the secondary carbide precipitates.

Degradation of Pyrometers / Thermovision in operation Increased skin temperature

catalysts Pressure gauges and on-line readings Change in pressure between inlet and outlet gases would reflect
disturbance of flow and damage of catalysts

External tube Visual inspection By touch and feel, smoothening of surface texture reflects external
oxidation oxidation
Thickness measurements Reduced wall thickness.
Thermal fatigue Visual inspection and ARTIS Typical bowing of tubes reflect disturbance in thermo-mechanical
stress distribution
Spring tension / counter weight balance readings Indicates disturbance in thermo-mechanical stress distribution
at penthouse area
Creep detection by UT
Detection of creep fissures: typical attenuation comparison
Creep detection by UT: Effect of carbides

TR Mode ultrasound wave attenuates by:

 Scattering at dendrite boundaries
 Secondary precipitates
 From creep damage, seen as formation of voids
 From interlinked voids (fissures) at mid-wall
Creep detection by UT: Effect of macro-structure

TR Mode ultrasound wave attenuation also differs with

 Fully columnar grain to Equiaxed and Columnar macrostructures
 Creep detection by UT: Effect of carbides

Alloy:KHR35CT, 100X, service exposed material (125000h),

Observation: creep voids are along primary carbide. Isolated creep

cavities are at core along with tendency for carbide coarsening.

Sound attenuation measured is 60 to 64 dB

 Creep detection by UT: Effect of carbides

Alloy HK40, Service exposed 29 years,

Primary carbides at dendritic grain boundaries. Outer and inner

edges show oxidation cracking. Aligned cavities are observed
along primary carbide at core. The secondary carbides are nearly
absent within the grains.

Sound attenuation measured is 56 to 62 dB

Microstructural degradation v/s Ultrasound attenuation

The microstructure of unused IN519

tubes is shown in Figure left, having
dendritic columnar austenite grains
oriented perpendicularly to tube walls.
Attenuation level: 35 – 55 dB

Upon aging, the structure degrades. The

original morphology of eutectic carbides
gets modified after exposure at
elevated temperatures.
Attenuation level: 55 – 62 dB

At life fraction near to 80 - 95%, the

microstructure is coarsened secondary
carbides with presence of inter
dendritic micro fissures.
Attenuation level: 62 – 72 dB
Severity Classification (Statistics based on general
Reformer Tube grade material)

Class - 1 Class - 2 Class - 3 Class - 4 Class - 5

Sound Aligned voids & Voids & creep Macro cracking

Creep Voids
Metal fissures Fissures with > 75%
Micro Fissures
(in as cast between mid-wall to OD, through wall
Inner Wall
condition) ID & mid-wall greater thickness
Critical
Slight Moderate Severe
No Damage Replacement
Damage Damage Damage
needed
- After 40% Life. After 55% Life After 75% Life After 100% Life

Class - 1 Class - 2 Class - 3 Class - 4 Class - 5

46
Progressive creep cracking (Reformer tubes)

Creep damage initiates from mid-wall

Balance between temperature gradient and internal working pressure

47
ARTiS model for integrated approach
Index (higher is worse)
Test
0 (least aged) 1 2 3 (most aged)
Apparent change in
Good condition with Apparent localized Presence of crack,
General visual weld or apparent
no significant bulging or shiny bulged with craze
examination offset or abnormal
abnormality. surface pattern
coloration
Smooth surface
Good surface
Visual baldness texture on touch and - -
roughness
feel
< 0.1X of tube Up to 0.5X of tube Up to 0.8X of tube Up to 1X tube
Bowing of tubes
diameter diameter diameter diameter
Ultrasonic
Up to 50 db Up to 58 db Up to 70 db > 70 db
attenuation*
Creep strain < 2% 2 to 3 % 3 to 5 % > 5%
Dilation of secondary Presence of
Presence of isolated
Microstructure carbides towards interconnected,
or oriented creep
Microstructural without any grain boundary with parallel micro
voids preferably
condition significant grain coarsening of primary cracks normal to
normal to principle
coarsening grain boundary principle stress
stress direction
carbides direction
Overall index Sum of individual indexing
Automated Reformer Tube Inspection System (ARTiS)
ARTiS is abbreviated to Automated Reformer Tube Inspection System.

This is a robotic crawler to aid ultrasonic testing of reformer tubes in a more

systematic manner and provide tabular and interactive digital output.

The method follows same principle of manual scanning.

While crawling it measures:
• Ultrasonic creep fissure detection
• Outer diameter of tube
• Tube bowing assessment

The outcome of inspection is more systematic and traceable throughout the

tube height.
Automated Reformer Tube Inspection System (ARTiS)
Comparison Manual method ARTiS
Requirement of scaffolding Needed Not needed
Cost of scaffolding At actual Nil
Time required for erection and 1 to 2 days Nil
removal of scaffolding
Requirement of D.M. water Continuous Limited
Spillage and wetting of Uncontrolled Nil
surrounding during test
Time of test (3 persons team) 3 to 5 days for 100 tubes 2 to 3 days for 100 tubes
Resolution of test result 1.5 - 2.0 meters 0.1 meter
Outer diameter measurement One or two locations manually Every 0.1 meter
Tube bowing Qualitative judgment or by Quantitative measure
plumb measurement, during UT scan
additional time
Safety consideration Higher risk Lower risk
(elevated work area) (platform area)
Reporting Manual data entry Software based with statistical
data analysis
Crawler components

Water tank
with pump

Drive

OD
measurement

Adjustable arm

UT Probes
Automated Reformer Tube Inspection System (ARTiS)
Automated Reformer Tube Inspection System (ARTiS)
Automated Reformer Tube Inspection System (ARTiS)
What if Level 1 criteria are not satisfied?

Level 3 assessments require determination of remaining life based on the inspection results used for the
Level 1 assessment. The assessment approach is based on the calculation of accumulated creep strain till
the date of operation. For the evaluation, operating parameters are to be obtained from time to time since
the commissioning till the date of assessment.

These data include original dimensions of the tubes, the start of run and end of run temperatures for each
time interval. The start and end of run conditions for each of the time intervals long with measured tube
metal temperatures during operation and pressure variations becomes the input parameters for life
assessment calculations. The changes / variation in pressure and tube metal temperature are generally
associated with catalyst changeovers. All these data are difficult to obtain and often not traceable.

On certain reformer designs, the tube metal temperatures for all of the tubes are not feasible to measure
which is the major influencing factor for creep life. It is therefore, necessary to calculate that effective tube
metal temperature for individual tube is to be used for future life calculations. The effective tube metal
temperature is defined as the temperature that caused the present level of creep damage to the tubes in
terms of creep strain (i.e. increase in diameter) accommodating all occasional changes in tube skin
temperatures as well as thermal cycling.

The creep strain used for calculations is necessarily a steady state creep rate of increase (i.e. stage II as
shown in the creep curve >> Next slide.)
Automated Reformer Tube Inspection System (ARTiS)

The creep strain to

be used in
calculations is
necessarily a steady
state creep rate.
Automated Reformer Tube Inspection System (ARTiS)

The creep strain is calculated

from maximum increase in
diameter along length of
tube.

Dividing the total creep

strain to number of hours
the tube is operated.

The creep strain rate along

with the initial stress for
each interval of loading
provides input parameters
for calculation of effective
tube skin temperature.

The curves for creep strain

rate v/s stress are available
with service providers
through series of actual
(T+273) = P / [20 – Log (creep rate)] creep rupture tests.
Automated Reformer Tube Inspection System (ARTiS)

From calculated effective

tube metal temperature,
new working life (DL) can be
calculated.

Data are made available by

manufacturer of alloy.

Generally provided with

procurement of new tubes.

DL = 10 {P / (T-EFF + 273)}
Automated Reformer Tube Inspection System (ARTiS)

From calculated effective

tube metal temperature,
new working life (DL) can be
calculated.

Data are made available by

manufacturer of alloy.

Generally provided with

procurement of new tubes.

DL = 10 {P / (T-EFF + 273)}
Automated Reformer Tube Inspection System (ARTiS)
Automated Reformer Tube Inspection System (ARTiS)

Typical output of remaining life with API 579 ASME FFS1 approach
Automated Reformer Tube Inspection System (ARTiS)
Conclusions
The non-destructive testing of reformer heater tubes has undue
emphasis predominantly on ultrasonic attenuation
measurements.

Change in microstructural condition such as carbide coarsening,

secondary carbide precipitation and depletion / dissolution of
carbides from the grains largely affect the ultrasound
attenuation mechanism and significantly affect the creep
strength of the material.

Microstructural changes are necessarily to be categorized with

respect to internal fissures and diameter correlation.
Conclusions
Significant or complete loss of ultrasound energy may be
indicative of presence of mid-wall fissures which requires
confirmation by radiography and metallography techniques.

The tube life assessment based only on NDT approach has so far
remained in isolation, where only a few techniques like
diameter measurements and loss of ultrasound energy are
considered for judgment.
TCR always believe to work as
a team

Thank you