EVALUATION OF STEEL AND TSA

COATING IN A
CORROSION UNDER INSULATION
(CUI) ENVIRONMENT
by:
Russell Kane
Honeywell Process Solutions

Monica Chauviere & Keith Chustz

ExxonMobil Research & Engineering
 Organization

ƒ Introduction
I t d ti
ƒ Background
ƒ Test Method
ƒ Experimental CUI Data
ƒ Discussion
ƒ Conclusions

2
 Introduction

ƒ Energy
E conservation
ti h has expanded
d d th
the use
of thermal insulation to reduce energy loss.
ƒ Insulation coveringg the metal surface can
create conditions conducive for corrosion
that can proceed in an insidious manner
ƒ Because this damage occurs out-of-sight,
out of sight,
serious and sometimes unexpected
consequences can take place.
ƒ Notable corrosion incidents under thermal
insulation have made this topic a major
concern for many companies involved in
petroleum refining and petrochemical
operations.

3
 Background (2)

ƒ The
Th mechanism
h i off CUI iinvolves
l
three requirements:
- Insulation produces a poultice
that can hold moisture and
dissolved species on the metal
surface
- Variable and high temperatures
and heat transfer across metal
surfaces increases the rate of
corrosion reactions through
wet/dry conditions
- Moisture and corrosives can
enter from the surrounding
atmosphere and leach from the
thermal insulation.
ƒ This special
Thi i l combination
bi ti off
variables is what commonly
defines the CUI environment. Comparison of actual plant CUI corrosion rates measurements

4
 CUI Testing

ƒ The
Th mostt challenging
h ll i obstacle
b t l iis
predicting the efficacy of coating
systems which are already in
service.
ƒ A variety of polymeric and metallic
coatings are currently in use.
ƒ However, long g term performance
and efficacy must be proven.
ƒ Recent developments in the field
of CUI have provided methods
that can be used to simulate CUI
in the laboratory under typical
conditions of exposure (e.g.
isothermal, wet/dry, cyclic
t
temperature).
t )
- This activity has resulted in a new
standard – ASTM G189-07 Standard
Guide for Laboratory Simulation of
Corrosion Under Insulation
Insulation.

5
 Purpose of this Study

ƒ The
Th presentt studyt d was developed
d l d to
t utilize
tili ttechniques
h i as
described in ASTM G189 for the evaluation of CUI under
simulated service conditions.
ƒ This study involved the use of TSA coating (Grade 1350) on
carbon steel (3 mil anchor pattern) and selected insulative
materials.
ƒ To simulate longer term service conditions, tests were
conducted on coated specimens and insulation that were
aged
g p prior to testing.
g
ƒ The results from TSA coated specimens were compared to
those produced on bare carbon steel.
ƒ Six CUI material conditions were used:
- Bare carbon steel
- TSA coated steel
- Defected TSA coated steel
- Each was tested with mineral wool and calcium silicate insulations.
6
 Test Methods

ƒ ASTM CUI simulation

i l ti proceduresd were used.
d
ƒ This procedure involved the use of a 2-inch (51 mm) O.D.
pipe
p p ggeometry y with internal heating
g and external thermal
insulation.
ƒ Ring specimens were exposed to CUI conditions on the
outer surface. Specimen evaluation included:
- Visual examinations
- Electrochemical corrosion
measurements
- Chemical analyses.
ƒ The ring specimens were
isolated from each other
using TFE spacers.
ƒ Specimens & insulations
were aged d tto simulate
i l t LT
service behavior.
7
 Aging Pre-Treatment

Material Aging Treatment Temperature Duration

(C / F) (hrs)
Bare Steel Immersion in test solution 23 / 75 48
TSA Coated Immersion in test solution 23 / 75 48
Steel
Defected TSA Immersion in test solution 23 / 75 48
Coated Steel
Mineral Wool Immersion in test solution 82 / 180 48
Insulation
Calcium Immersion in test solution 82 / 180 48
Silicate
I l i
Insulation
* Test Solutions – distilled water containing reagent sulfuric acid to pH 5.0 and 100 ppm NaCl

8
 CUI Test Set-Up

Exposure Cell

Test Solution: DI Water + H2SO4 to pH 5; 100 ppm NaCl

Cyclic Wet-Dry conditions: 20 hrs/82 C and 4 hrs/120 C
Mineral Wool and Calcium Silicate Insulations tested

Electrochemical
g
Monitoring

SmartCET Online, Real-Time Corrosion Transmitter

9
 Data Analysis
ƒ Electrochemical
El t h i l E Evaluation
l ti
- Corrosion rates were plotted
versus time using a
combination of linear
polarization (LPR) and
harmonic distortion analysis
(HDA).
- The B value correction was
based on the HDA
measurements made during
the actual exposure test
test.
- Pitting Factor (PF) was
assessed using the standard
deviation of the corrosion
current divided by the average
corrosion current during
measurement cycle (7 min.)
Š PF > 0
0.1
1 (Localized Corrosion)
Š PF < 0.01 (General Corrosion)
ƒ Corrosion
C i A Analyses
l
- The ring specimens were
visually examined for corrosion
or coating damage
damage.
- The working electrode was
used for a destructive
examination including:
Š Metallographic sectioning
Š Examination of the corrosion
- The TSA coating thickness was
assessed by metallographic
cross-sectioning and
microscopical measurements.
- Scrapings from the surface
deposits on the ring specimens
and pH measurements of the
insulation were taken after
t ti
testing.
10
 Steel in Mineral Wool & Calcium Silicate

ƒMineral
Test 1 - LPRCorrosion Rates for Steel - Cyclic Temperature Wet Dry

Mi lW
Wooll 100.000

LPR Corrosion Rate

- Initial Corrosion rates 10.000

were high

Corrosion Rate (mpy))

1.000

- Distinct wet/dry periods 0.100

Š Each wet/dry cycle produced

a corrosion rate spike 0.010

ƒCalcium Silicate 0.001

8-Jan-07 9-Jan-07 10-Jan-07 11-Jan-07 12-Jan-07 13-Jan-07 14-Jan-07 15-Jan-07 16-Jan-07 17-Jan-07 18-Jan-07 19-Jan-07
Date/Time

- Maximum
a u co
corrosion
os o rates
a es Test 2 - LPRCorrosion Rates for Steel - Cyclic Temperature Wet Dry

more consistent with time 1.000

LPR Corrosion Rate

- No evidence of wet/dry
periods Corrosion Rate (mpy)) 0.100

- Corrosion rate is low but

highly variable 0.010

0.001
8-Feb-07 9-Feb-07 10-Feb-07 11-Feb-07 12-Feb-07 13-Feb-07 14-Feb-07 15-Feb-07 16-Feb-07 17-Feb-07 18-Feb-07 19-Feb-07
Date/Time

11
 Corrosion of Carbon Steel in Mineral Wool

LPR Corrosion Rate (w/B value = 25 mV from HDA)

Pitting Factor (Test 1: C. Steel - Cyclic Temperature Wet Dry)
100.000 1.000

LPR Corr Rate (default B)

Pitting Factor

10.000 0.100
Rate (mpy)

Factor
Pitting F
Corrosion R

1 000
1.000 0 010
0.010

0.100 0.001

Isothermal drying
period
0.010 0.000
21-Dec-06 10-Jan-07 30-Jan-07 19-Feb-07 11-Mar-07 31-Mar-07 20-Apr-07 10-May-07
Time (Date / Time)

ƒ Corrosion rate decreased with time of exposure.

ƒ Corrosion rate initially up to 40 mpy (1 mm/yr) decreasing to <0.1 mpy (0.03 mm/yr)
ƒ Pitting tendency is initially high but also reduced with time
time.
ƒ Corrosion decreases to baseline rate during drying period at end of the test. No
evidence of dry out.
12
 Corrosion of Carbon Steel in Calcium Silicate

LPR Corr Rate (default B) LPR Corrosion Rate (w/B value = 26 mV from HDA)
Pitting Factor Pitting Factor (Test 2: C. Steel - Cyclic Wet Dry)
10.000 1.000

0.100

1.000
Rate (mpy)

Isothermal drying

Factor
period

Pitting F
Corrosion R

0 010
0.010

0.100

0.001

0.010 0.000
16-Jan-07 26-Jan-07 5-Feb-07 15-Feb-07 25-Feb-07 7-Mar-07 17-Mar-07 27-Mar-07 6-Apr-07 16-Apr-07 26-Apr-07 6-May-07
Time (Date / Time)

ƒ Corrosion rates generally between 0.1 to 1.0 mpy (0.003 to 0.03 mm/yr)
ƒ Some spikes during rehydration to higher rates
ƒ Corrosion rate becomes variable during final drying period
period.
ƒ High PF caused by highly variable corrosion rates and not pitting.
13
 TSA Coated Steel in Mineral Wool

LPR Corrosion Rate (w/B value = 15 mV from HDA) & Pitting Factor
(Test 1: TSA Coated Steel - Cyclic Temperature Wet Dry)
10.000 1.000

LPR Corr Rate (default B)

Pitting Factor

1.000 0.100
n Rate (mpy)

g Factor
0.100 0.010

Pitting
Corrosion

Isothermal drying
period

0.010 0.001

0.001 0.000
21-Dec-06 10-Jan-07 30-Jan-07 19-Feb-07 11-Mar-07 31-Mar-07 20-Apr-07 10-May-07
Time (Date / Time)

ƒ Initial corrosion rates around 0.01 mpy (0.003 mm/yr)

ƒ Corrosion rates decrease to detection limit – near 0.001 mpy (0.0003 mm/yr)
ƒ Corrosion became intermittent likely due to passivation of the aluminum
aluminum.

14
 TSA Coated Steel in Calcium Silicate

LPR Corrosion Rate (w/B value = 15 mV from HDA) Pitting Factor

(Test 2: TSA Coated Steel - Cyclic Temperature Wet Dry)
10.000 1.000

LPR Corr Rate (default B)

Pitting Factor

Isothermal drying
period 0.100

1.000
on Rate (mpy)

Pitting Factor
0.010
Corrosio

0.100

0.001

0.010 0.000
16-Jan-07 5-Feb-07 25-Feb-07 17-Mar-07 6-Apr-07 26-Apr-07 16-May-07
Time (Date / Time)

ƒ Corrosion rates were higher than in mineral wool and were in the range
0.1 - 1 mpy (0.003 – 0.03 mpy)
ƒ Highly variable corrosion rates result in high PF not due to localized corrosion
ƒ During final drying period corrosion rate reduce to < 0.010 mpy (0.0003 mm/yr)

15
 Corrosion of Defected TSA

ƒ Corrosion
C i off d defected
f t d TSA was lik likely
l d
driven
i b
by corrosion
i
of exposed steel areas however, it could have been an
acceleration of the TSA corrosion rate.
ƒ Corrosion rates were generally higher tested with calcium
silicate versus mineral wool:
- Initiallyy 1 – 10 mpy
py ((0.03 – 0.3 mm/yr)
y ) in both types
yp of insulation
- Corrosion rates decreased with time to 0.01 – 0.1 mpy (0.0003 –
0.003 mm/yr) in mineral wool
- Corrosion rates stayed
y higher
g in calcium silicate in the range
g of
1 – 10 mpy (0.03 – 0.3 mm/yr).
ƒ Corrosion did not result in measureable sacrificial loss of
TSA.
ƒ Good adhesion was observed for the TSA around the
defected areas.

16
 Other Observations

ƒ Forced
F d aeration
ti off the
th test
t t solution
l ti did nott change
h th
the
corrosion rate of carbon steel, TSA or defected TSA
specimens.
ƒ CUI corrosion rates varied with the level of water availability
- When the water injection was stopped corrosion rates decreased very
low levels
- But, once hydrated, the mineral wool insulation could sustain
corrosion of steel at 82 C without water injection for a prolonged
period. Not so in calcium silicate.
ƒ Post-test
P pHH measurements
- The pH of the mineral wool after testing averaged 5.1.
- The pH of the calcium silicate after testingg was onlyy slightly
g y higher
g
with an average value of 5.5 versus initial readings of near 10.
ƒ Post-test surface deposit analyses using SEM/EDS showed
evidence of elements contained in the corrosion p products,
test solution and the thermal insulation with no evidence of
unexpected species.
17
 Conclusions

ƒ The
Th highest
hi h t corrosioni rates
t observed
b d during
d i CUI exposures
occurred for bare steel during the initial month under mineral
wool insulation. Corrosion rates were in the range of 1 to 40
mpy (0.03
(0 03 tto 1
1.0
0 mm/yr).
/ )
ƒ TSA coated specimens showed the lowest corrosion rates
for all cases with corrosion rate of < 0.01 mpy (0.0003
mm/y) under mineral wool insulation and between 0.1 and 1
mpy (0.003 to 0.03 mm/y) for calcium silicate insulation.
ƒ Defected TSA coated specimens
p showed higher
g corrosion
rates than non-defected TSA. This was primarily due to the
increased activity of the carbon steel in the defected region
or increased activity of the TSA.
ƒ No measurable sacrificial reduction in the TSA coating was
noted or could be confirmed by post-test evaluation.

18
 Conclusions (2)

ƒ Post-test
P t t t analyses
l off surface
f deposits
d it showed
h d only
l
evidence of elements contained in the corrosion products,
test solution and the thermal insulation.
ƒ Air purging of the test solution used in the CUI tests, did not
increase the corrosion rates in the test over those obtained
in tests using the test solution exposed to ambient air.
ƒ Mineral wool and calcium silicate responded differently to
the wet and iso-thermal drying periods.
- Corrosion rates were more variable with time in the calcium silicate insulation.
- The corrosion rate of the bare steel in the test with mineral wool was
maintained at the low but finite rate during drying indicating that drying of the
insulation was not achieved at 82 C.
- B comparison,
By comparison the corrosion rate of the bare steel in the calci
calcium
m silicate test
dropped to a very low rate and became intermittent suggesting drying of the
insulation.

19