L.

de Kerret

Material and process improvements in condenser tubing

L. de Kerret1, P. Gerard1
1)
Valtimet, 21150 Les-Laumes, FRANCE

Abstract
The reliability of the surface condenser is a key factor in plant performance level and maintenance
cost optimization. This is especially the case for thermal nuclear plants where condenser raw wa-
ter ingress can introduce contamination into the chemically-controlled, steam/water loop potentially
causing damage to sensitive equipment.
Two important parameters must be taken into account when attempting to optimize the quality and
the reliability of condenser tubing. They include selecting the appropriate material according to the
cooling water corrosion level present. A second and equally important parameter is the manufac-
turing of the tubing product itself.

This paper will identify methods to optimize manufacturing processes and improve tubing quality,
according to VALTIMET’s 30 years of condenser welded tubing production experience.

Those methods complete the core manufacturing process (forming and welding), through im-
provement of the metallurgical homogeneity (Roll Sink Anneal), Tube integrity (Laser welding),
Tube reliability (NDT), and improvement of the corrosion resistance (bright annealing and passiva-
tion).

The paper will also present an overview of the principal materials used for condenser tubing em-
phasizing the importance of the material choice according to operational environment and specific
corrosion risks, using VALTIMET’s returns of experience accumulated over the past decades.

1
L. de Kerret

1 Alloy choice

1.1 Materials history overview

The introduction of the first stainless steel welded tubing began in the U.S. with “ordinary” grades
such as UNS S30403 or UNS S31603, which provided good service records in fresh water.
S31603 was used for a short period in seawater applications but proved to be susceptible to loca-
lized pitting and crevice corrosion in concentrated chloride environments.

The titanium development in Europe and Japan in 1970 offered a timely solution. Demonstrating
outstanding resistance to general and localized attack in high chloride environments, it has pro-
vided nearly forty years of trouble-free seawater service for the power generation and process
industries. In the late 70’s, due to the titanium market crisis, highly alloyed stainless steels, also
called super stainless steels, were developed to offer a lower-cost alternative solution. These new
alloys included UNS S31254 and UNS N08367 super austenitic alloys and UNS S44735 and UNS
S44660 super-ferritic alloys. However, the use of these alloys has been quite limited mainly to the
use of titanium retubing programs in Europe and USA.

Since the year 2000, material prices for Nickel and Molybdenum have been unstable and fluctuat-
ing. As a result, the duplex and lean duplex stainless steels including UNS S31803, UNS S32003,
UNS S32304 and UNS S32101 have been developed as a cost-effective alternative to traditional
standard austenitic stainless steel alloys for use in mild cooling water service. Recently, super-
duplex alloys have also been developed in strip form offering additional material options for brack-
ish and seawater service.

Today, titanium and stainless steel welded tubing represent the vast majority of the condenser
tubing market. The pallet of corrosion-resistant alloys covers a large spectrum of applications in
terms of corrosion resistance and corrosion constraints. The material selection criteria for con-
denser tubing is mainly based on the choice of the most suitable alloy according to the cooling
water quality.

1.2 Condenser Tubing Material Categories

Figure 1 provides a ranking of “typical” stainless steel alloys according to their Critical Crevice
Temperature (CCT) per ASTM G48. This test characterizes their crevice corrosion resistance per-
formance in a very aggressive ferric chloride solution. A relation between the CCT of a given alloy
and its performance in chloride-containing waters can therefore be established based on actual
testing and direct experience in the form of chloride limits expressed as ppm and presented on the
right-hand axis.

2
L. de Kerret

CHLORIDE LIMIT HEAT

60 °C
EXCHANGER AMBIENT T°
S44735 ppm
50

40
S44660
N08367
15000
30
10000
CCT(°C) per ASTM G48

20 S31803 5000

10 N08904
2000
0
10 15 20 25 30
S31703
35 40 45 50 1000
-10
500
-20
S31603
200
-30
S30403
100
-40 PRE NUMBER (Cr+3.3Mo+16N)

Figure 1: Critical Crevice Temperature per ASTM G48 and Maximum Chloride Levels According to
the PREN of Certain Stainless Steel Alloys.

This chart enables the reader to distinguish different categories of stainless steel alloys according
to their corrosion resistance behavior:

“ordinary” stainless steels with “basic” corrosion resistance, including austenitic alloys
S30403 and S31603, or ferritic alloy S43005. The corrosion resistance of these alloys is
typically oriented towards mild water service with very low chloride content. This category
also includes the newly developed lean duplex family including alloys S32101, S32304 and
S32003 alloys in particular.

super stainless steels with high corrosion resistance, including supper-ferritic alloys S44735
and S44660, super-austenitic alloys N08367, S31254 and S34565, and super-duplex alloy
S32750. Super stainless steels are highly alloyed stainless steels with high Mo, Cr and N
contents, and may be used for brackish or sea water service.

Intermediate corrosion resistant alloys with improved corrosion resistance over “ordinary”
grades include the duplex alloy S31803 or the austenitic alloys N08904, S31703 or
S31726.

Note: The main aggressive agent in cooling water to be considered, is the chloride content. To
avoid risks of bio-fouling, it is a common practice to perform chlorination. However, this adjunction
must be controlled, and the condenser tubing material must be chosen accordingly.

3
L. de Kerret

The chloride content must be measured at the condenser inlet, and not in the incoming water. The
reason is that cooling towers contribute to a chloride accumulation, which can be harmfull if not
taken into account.

The PREN or Pitting Resistance Equivalent Number index method may be the easiest and most
common method to rank and compare the different alloys according to their pitting corrosion resis-
tance, even though this index is purely empirical. PREN is calculated using the following formula.

PREN = %Cr + 3.3 (%Mo) + 16 (%N)

Table 1 provides the typical PREN, along with the typical tensile requirements for alloy tubing
commonly used for condenser service. Ranking of the alloys according to the PREN index leads to
the same categories of condenser tubing alloys:

“ordinary” stainless steels with PREN between 18 and 30

“improved corrosion resistance” alloys with PREN in the range of 30-40

super stainless steels with PREN beyond 40

YS0.2 TS
ASTM UNS El% PREN
MPa MPa
S30403
A 249 170 485 35 18
TP 304L
S31603
A 249 170 485 35 23.6
TP 316L
A 789 2
S32101 530 700 30 26
A 790
A 789 3
S32003 485 690 25 30.2
A 790
A 789 2
S32304 400 600 25 25.6
A 790
A 789
S31803 450 620 25 34.6
A 790
S31703
A 249 205 515 35 35.1
TP 317L
S31726
A 249 240 550 35 35.3
TP 317LMN
4
A 249 N08904 215 490 35 34.2
4
A 249 S31254 310 675 35 43.3
4
A 249 S34565 415 795 35 46
5
A 249 N08367 310 690 30 44.5
1
A 268 S44660 450 585 20 39.2
6
A 268 S44735 415 515 18 42.2
A 789 7
S32750 550 800 15 42.5
A 790
R50400
B 338 275 345 20 N/A
Ti GR. 2
Table 1: Typical PREN and Typical Tubing Tensile Requirements (min.) According to ASTM for
Commonly Used Condenser Tubing Alloys

4
L. de Kerret

For seawater usage, titanium remains the best technical solution, with an uncontestable superiori-
ty, especially for industries like power generation or desalination, with a perfect return of experi-
ment after decades of utilization and an installation track record approaching 40 years of operation
without one reported corrosion event.

This first part presents a pallet of corrosion resistant stainless and titanium alloys for condenser
and heat exchanger tubing service. After a careful and extensive analysis of the operating condi-
tions, the alloy selection can be made among this vast choice of grades which correspond to dif-
ferent needs in terms of erosion and corrosion resistance levels, mainly dictated by, but not only
limited to, the cooling water chemistry.

2 Manufacturing Process Improvements

Heat exchanger tubes can be of two types: seam-welded or seamless. Seamless are often consi-
dered as more reliable, being used in the primary loop of nuclear power-plants for instance. Their
metallurgical homogeneity is, indeed, not to be questioned. However, seamless tube manufactur-
ing process is long and complex, includes several cold works and heat treatments. This leads to a
higher manufacturing cost, felt in the product price, especially for condenser thin-wall tubes. On
the other hand, welded tube represents a valuable alternative in many applications, not only due to
its lower cost, but considering its proven reliability (countless references of welded heat exchanger
tubes), and the outstanding characteristics featured: mechanical, geometrical, corrosion behavior.

To achieve welded tubing expected performance, the manufacturing process stands as a major
parameter. With more than 30 years of experience in welded tube manufacture, Valtimet has de-
veloped some techniques to improve this process. After a brief process overview, we will focus on
Valtimet particular techniques, insuring homogeneity of tube section: the WGS, the RSA (for aus-
tenitic welded tubes) and also Bright Annealing. Some tests and experiments will be presented,
demonstrating the influence of specific manufacturing processes on tube reliability.

2.1 Manufacturing process in brief

Figure 2: Valtimet Titanium and stainless steel welded tubing manufacturing process

Figure 2 describes Valtimet titanium and stainless steel welded tubing manufacturing process.
Tube mill strip is uncoiled and fed through a multiple stand forming mill which shapes the strip into

5
L. de Kerret

a tube. The tube is autogenous welded in an inert atmosphere with a non-consumable tungsten
electrode (T.I.G. welding) or with the laser technology. Dimensional properties are achieved when
the as-welded tube passes through a final, multi-stage sizing mill. Mechanical and thermal
stresses are relieved by in-line induction heating under inert atmosphere (bright annealing). Fol-
lowing that, the tube is straightened, 100% Eddy Current (EC) tested, line marked, then automati-
cally cut. Finished tube ends are deburred. The product is subsequently dimensionally inspected,
tested through NDT controls (pneumatic testing, EC inspection and Ultrasonic Testing) and pack-
aged to customer specification.

2.2 Weld Gas Smoother

Considering a seam-welded tube cross section, the weld is obviously never 100% invisible. There
are some inherent characteristics to the welding process, which lead to a geometrical difference
from the base material (laminated strip).

This is specially the case with TIG welding: Tungsten Inert Gas process is an autogenous welding,
resulting in a wide weld bead (up to several times the wall thickness). During solidification, the
weld bead is pulled down by gravity, even more considering the pressure applied on the tube
sides, by the welding rolls. If no care is taken, the weld bead will collapse down.

Such a phenomenon will result in a shape irregularity, harmful to the mechanical resistance (espe-
cially fatigue), to the corrosion resistance (crevice corrosion) if it forms a confinement zone, and
harmful to the tube controllability (especially for Ultrasonic control).

To avoid this phenomenon, VALTIMET has developed a process, named Weld Gas Smoother. As
indicated by its name, a gas flow under pressure is injected into the tube, during welding, to hold
up the weld bead. This gas is, of course, inert, to avoid any oxidation of the ID surface; Argon is
used for that purpose. In addition, a special care is taken, to use the right parameters: weld rolls
pressure and positioning, sufficient gas flow... to insure the control of weld dimensions and geo-
metry. Figure 3 shows the principle of the Weld Gas Smoother process.

Figure 3: Weld Gas Smoother process principle

6
L. de Kerret

2.3 Roll Sink Anneal

For Austenitic Stainless Steel tubes manufacturing, ASTM A249 and A688 require a cold work
operation on the welded tubes before annealing; this cold work operation can be carried out on the
entire section of the tube (weld +parent metal), or only on the weld . Today, several technologies
exist to satisfy this requirement (bead rolling, bead hammering, roll sinking, redrawing…).

Valtimet has chosen to use another technology: the

Roll Sink Anneal process. This technique provides
a cold work operation during sizing (Roll Sink), fol-
lowed by a bright annealing.

The sizing is operated by 4 horizontal and 4 vertical

roll sets, as shown on figure 4. Passing through
those rolls, the tube diameter is reduced, so as to
reach the final tube diameter. Diameter reduction
goes up to 12% reduction, depending on the ma-
Figure 4: Sizing OD reduction principle
terial grade and on the wall thickness.

The weld bead material being softer than the base

material, most of the cold work is supported by the
weld. This is easily visible on figure 5, representing
a TIG welded S31603 tube: with a 10% OD reduc-
tion sizing, the dimensional reduction of the weld
zone is up to 16%.
Figure 5: Weld bead profile before and
after sizing, TIG welded S31603 tube
Figure 6 shows the evolution of mi-
cro-hardness in Weld and Base met-
al, throughout the manufacturing
process, on a TIG welded S30403
tube, which supported an OD reduc-
tion of 10%. It clearly demonstrates
that sizing takes the weld bead hard-
ness at the same level than the base
material.
Figure 6: Micro-hardness evolution TIG welded S31603
tube

7
L. de Kerret

The second step is an in-line annealing treatment, which main purpose is to relax the residual
strains from forming and sizing. The most visible result is a homogenization of mechanical proper-
ties, as shown on figure 6 (Hardness values after bright annealing).

For austenitic stainless steel tubes, annealing leads to a re-crystallization of the weld material, and
dissolution of ferrite, as shown on Figure 7. Here, annealing has lowered the ferrite rate down to
0.02%.

Figure 7: S30403 TIG welded tube cross section after nitric acid etching

VALTIMET has developed a Bright Annealing process, which ensures no oxidation, neither on
external tube surface, nor on internal surface, thanks to an inert atmosphere protecting the tube
during Heat treatment (induction with an encircling coil) and cooling (carbon sleeves cooled by
water flow).The utility of a bright annealing compared to annealing + pickling will be detailed in a
next chapter.

The main proof of RSA efficiency and reliability would be its references. As a matter of facts, RSA
process has been already qualified by more than 30 companies in the world. Today, more than
200.000 km of tubes have been manufactured by Valtimet with RSA process; more than 400 pow-
er stations in the world, fossil or nuclear, are equipped with condenser and feed water heater
tubes cold worked by RSA.

2.4 Influence of cold work technology on weld corrosion resistance

The strengths and drawbacks of each cold work technology is very well known now, exepted may
be the influence on the corrosion resistance of the weld. The present work8 will compare a tube
cold worked with RSA process and a tube cold worked with bead rolling, with ASTM G48 method
A test in order to evaluate the pitting corrosion susceptibility of both welds.

The investigated tubes were manufactured at CST hyderabad plant by CST VALINOX, on the
same line, and with the same operating conditions. Their final dimensions were 31.75 mm OD,

8
L. de Kerret

0.80 mm wall thickness. The tubes were manufactured in 304L and TIG welded. In both cases,
after cold work, an in line annealing at 1050°C was performed.

Table 2: Nominal chemical composition of tested AISI 304L TIG welded tubes (weight fractions)

C Cr Ni Mn Si P N S Fe
304L 0.030 18.2 10.1 2.00 0.75 0.045 0.10 0.030 Bal.
max max max max max max

Figure 8 shows micrographs of both Bead Rolled tube weld and Roll Sinked tube weld. The micro-
graphic examinations show no evident differences between the two tube samples .The microstruc-
tures are very similar as well as the ferrite content (1.25% for the bead rolled weld and 1.30% for
the non bead rolled weld).

Figure 8: Micrographs and dendrite/austenite interface of the

(a) bead rolled and (b) roll-sinked welds.

Based on the chemical composition, the critical pitting temperature of UNS 30403 stainless steel is
estimated to be 10°C (Table 2). However, welding is known to induce microstructural modifications
and is thus expected to modify the corrosion susceptibility of the welded product. For that reason,
the critical pitting temperature of the tubes was investigated starting to temperatures below 10°C.

According to the ASTM G48 Method A standard (Ferric Chloride Pitting Test), samples were im-
mersed into a ferric chloride solution (PH ≈ 1) during 24 hours. The temperature of the ferric chlo-
ride test solution was controlled via a thermostated water bath at 5°C (heater chiller). This tests
allows the determination of the Critical Pitting Temperature (CPT), which is defined as the temper-
ature at which pits of at least 25 μm depth are developed on the metal (ASTM G48 standard).

9
L. de Kerret

Figure 9: (a) Roll-sinked and (b) bead rolled tube after 24 hrs at 5°C in the ferric chloride solution

After 24 hours immersion in the ferric test solution at 5°C, both tubes exhibited strong pitting cor-
rosion features. Figure 9 shows that :

Pitting corrosion is localized along the weld with extent of the attack leading to holes.

The rolled tube was significantly more attacked than the non-rolled one.

This observation is consistent with the corrosion rates obtained from the weight loss of each tested
sample (Table 3).

Non-rolled tube (Roll Sink) Bead-rolled tube

1.40 mm/y 3.10 mm/y
55 mpy 122 mpy
Table 3 : Corrosion rates after 24 hours at 5°C in the ferric chloride test solution

This experiment clearly demonstrates that Bead Rolling technology is detrimental to the pitting
corrosion resistance of the weld zone. Roll Sinking process appears then as an interesting alterna-
tive. Regarding the corrosion resistance, it is very well known that the surface finish is an impor-
tant parameter. Therefore, for instance, a rolling process after welding might induce blending of
pollutants at the weld surface. In such a case, surface decontamination thanks to a nitric acid pick-
ling/passivating process would result in a surface free from of any contamination such as oxides,
rusts, ferrous particles…that promote pitting corrosion by acting as preferential initiation sites.

2.5 Influence of Bright Annealing on corrosion resistance

Investigations9 were carried out on UNS S44735 in order to assess the impact on the corrosion
resistance of the different heat treatment processes which can be used during the welded tubing
manufacturing. A conventional ASTM G48 test has been performed on three different states ac-
cording to heat treatment processes:

“As received”: UNS S44735 strip without any additional heat treatment than the one
performed during the strip production

10
L. de Kerret

“900-Air”: strip which has been heat treated at 900°C under air (open air annealed) and
pickled to remove the residual oxidation due to the oxidizing environment during the heat
treatment process (representative of tubes which are open air annealed and pickled)

“900-H2”: welded tube which has been heat treated at 900°C under hydrogen protective
atmosphere (representative of tubes which are bright annealed with Valtimet process)

Table 4 shows the corrosion rate (C.R.) measured thanks to the weight loss of the coupons.

Thickness Area Time Weight loss Corrosion Crevice spots

Materials T (°C)
(mm) (cm2) (h) (g) Rate (µmpy) (arbitrary units)

"900-H2" 0.66 19.84 50 24 0.0005 12 No

"900-Air" 0.67 12.07 50 24 0.1990 7842 10/20

"As received" 0.68 9.19 50 24 0.0001 5 No

Table 4: Weight loss corrosion of samples under investigation (ASTM G48 Method B)

This allows ranking of the three materials when tested according crevice susceptibility in ASTM
G48 Method B environment as follows:

UNS S44735 "As received” ~ "900-H2" >> "900-Air"

Figure 10, Figure 11 and Figure 12 show the crevice specimens after testing of the three samples.
While “As-received” and “900-H2” materials did not present any sign of crevice corrosion (except
one spot on one specimen of “900-H2”), “900-Air” material exhibited significant crevice corrosion
under the TFE-fluorocarbon blocks associated with a very high corrosion rate of the coupons.

Figure 10 : ASTM G48 Method B – “900-Air” specimens after 24 h. in ferric chloride solution

Figure 11 : ASTM G48 Method B – “900-H2” specimens after 24 h. in ferric chloride solution

11
L. de Kerret

Figure 12 : ASTM G48 Method B – “As received” specimens after 24 h. in ferric chloride solution

Tests conclusion: The ASTM G48 test provides more severe conditions than corrosion tests in
seawater (i.e. the fitness-for-purpose environment according to the medium in contact with
materials in heat exchangers) but has the drawback of being an artificial medium not
representative of service in seawater. Nevertheless, This test show a consequent improvement
coming from the heat treatment process at 900°C under Hydrogen protective atmosphere in
comparison with the other heat treatment under open air. Furthermore, those crevice corrosion
tests showed a very high susceptibility of UNS S44735 material heat treated under oxidizing
atmosphere, i.e. open air annealed followed by an acidic pickling to remove the residual oxidation.

Those results can be generalized to all stainless steels and lead to the conclusion that welded
tubes which are bright-annealed during the production process show a better corrosion resistance
than the ones which are open air-annealed then pickled.

2.6 Influence of weld technology on corrosion resistance

2.6.1 TIG and Laser weld description

For welded condenser tubing manufacture,

two welding technologies are available: La-
ser and TIG. Both are autogenous welds
(without added filler metal), but, as the two
micrographics on figure 14, the result is
completely different. The biggest advantage
of Laser welding is actually the welding
speed, especially for heavy wall stainless
steel tubes. Indeed, as shown on figure 13,
Figure 13: TIG and Laser welding speed comparison
the higher the wall thickness, the faster is
Laser welding compared to TIG welding.

12
 L. de Kerret

(a) (c)

(c)
(a)

(b)

Figure 14 : weld micrographics of a) 0.5 mm Thickness Ti Gr.2 TIG welded tube,

b) 0.5 mm Thickness S31803 TIG welded tube, c) 0.9 mm Thickness S30403 Laser welded tube

Moreover, the energy repartition is different with TIG and Laser weld: Laser uses the keyhole ef-
fect, and provides a very high temperature gradiant and a minor Heat Affected Zone (HAZ). The
next paragraph will detail what could be a consequence of this metallurgical difference.

2.6.2 Compared pitting corrosion susceptibility for super austenitic welded tubes

10
This study aims at evaluating the resistance of a super austenitic stainless steel welded tube to
pitting corrosion, and more specifically to compare the influence of welding process on the pitting
susceptibility. The alloy studied is UNS N08367, and its chemical composition is presented in Ta-
ble 5. Since it is highly alloyed, UNS N08367 has a PREN of 44.3.

Chemical composition of tested UNS N08367 tubes, in %

C Mn P S Si Cr Ni Mo N Cu

0.022 0.71 0.023 0.0003 0.38 20.5 24 6.2 0.21 0.44

Table 5 : Chemical composition of AL-6XN Laser and TIG welded tubes.

The evaluation of tube Critical Pitting Temperature is conducted using tests following ASTM G48 E
(24 hours exposure). During this experiment, if minimum pitting attack is observed, the next test is
performed at a temperature 5°C lower than the previous one. If less than minimum pitting attack is
observed, the next test is conducted 10°C higher. The procedure is repeated using a new tube
and a fresh solution. For each steel at each temperature, test is repeated twice.

The first test campaign investigating pitting corrosion of UNS N08367 Laser and TIG welded, gave
respectively 80°C and 50°C as CPT values. Tested at their respective CPT, UNS N08367 Laser
tubes showed pits on both the base metal and the weld while UNS N08367 TIG tubes developed
holes and pits on the weld only. The second test campaign gave similar results. Figure 15 shows

13
L. de Kerret

the aspect of Laser and TIG welded tubes at their CPT and 5°C below. At 80°C, Laser welded
presents holes and pits on both the base metal and the weld. At 50°C, TIG welded shows holes
and pits on the weld; the base metal is not affected because the temperature is not high enough.

The sequence of increasing resistance to pitting corrosion is the following:

UNS N08367 TIG (50°C - weld) < UNS N08367 Laser (80°C - base metal)

(a) UNS N08367 Laser 75°C (b) UNS N08367 Laser 80°C

W
W

(c) UNS N08367 TIG 45°C (d) UNS N08367 TIG 50°C
Figure 15: Pictures of UNS N08367 Laser welded tubes tested at 75°C (a) and at 80°C (b) and
pictures of UNS N08367 TIG welded tubes tested at 45°C (c) and at 50°C (d).

For UNS N08367 tubes, laser welding process provides better pitting corrosion resistance than
TIG welding process. Indeed, the laser welded tubes resist until the base metal limit (80°C). Con-
trary to the TIG welded tubes, the laser weld is not the weak point of the tube.
A similar experiment10 has been performed on 1.4565 TIG and Laser welded tube, showing the
same tendency. For super-austenitic tubes, Laser welding is then more appropriate than TIG weld-
ing.

2.7 Influence of passivation

Valtimet recommends including a passivation step in the manufacturing route, for austenitic stain-
less steel condenser tubes. The present work11 was conducted to evaluate the interest of nitric
passivation, to improve corrosion resistance of a UNS 31603 (316L) TIG welded condenser tube.

14
L. de Kerret

The 316L welded tubes were passivated in a 25% HNO3 solution at 25°C, during 15 minutes.

Then, pitting potential measurements were performed by recording polarization curves, according
to ASTM G61, in a 3.5% NaCl solution, at a temperature of 25°C. The solution was not de-
aerated, to reproduce more accurately the condenser tubes operating conditions. The electro-
chemical cell was a conventional three
electrodes cell, involving the metal surface
to be analyzed, a platinum counter elec-
trode, and a Saturated Calomel Electrode
(SCE) as a reference.

The polarization curves on figure 16 dis-

play the measurement of the current de-
veloped on the sample as a function of the

Figure 16: Polarization curves of passivated 316L time and the applied potential.
tubes, in non-deaerated NaCl solution

To analyze these curves in terms of pitting suscep-

tibility, the pitting potential is determined in each
curve. This pitting potential Ep correspond the po-
tential at which the current density increases conti-
nuously. It corresponds to the potential at which pits
are developed on the stainless steel. The higher Ep
the more resistant to pitting the stainless steel is.
Figure 17: Values of the pitting potential Ep
The values of the Ep are presented on figure 17. of 316L stainless steel, in 3.5% NaCl solution

Figure 17 shows that in the modified conditions of ASTM G61, when the solution is not deaerated
(O2 remaining), Ep is higher for samples passivated with HNO3 than for samples not passivated
with HNO3. As a higher Ep means a lower pitting susceptibility, the internal and external surfaces
of the tube are highly susceptible to pitting when they are not passivated with HNO3.

CONCLUSION: The promoting effect of passivating 316L stainless steel with HNO3 is clearly ob-
served in 3.5% NaCl solution containing O2, regarding that non-passivated sample are more sus-
ceptible to pitting corrosion than passivated samples.

15
L. de Kerret

3 Non Destructive Testing Improvements

The term Non-Destructive Testing (NDT) regroups all the controls performed on the product, which
do not cause damage or modify the product itself. This is a highly valuable solution, to insure that
the very specimen delivered to the customer has been controlled and is compliant. In opposition, a
destructive testing method will consist in testing a sample that will not be usable afterwards, with
the assumption that the delivered product is similar to the sample tested.

Valtimet experience acquired through the welding of more than 190,000 km of titanium and more
than 230,000 km of stainless steel welded condenser tubing, has resulted in the identification of
specific risks linked to the product. To ensure that none of those defects are delivered to the cus-
tomer, a tailor-made control process including Air Pressure Testing, Eddy Current (EC) testing and
Ultrasonic Testing (UT) has been developed, far beyond usual standards requirements.

3.1 NDTs calibration principle

Parameters to be taken into account for the calibration of the NDT equipment to establish the min-
imum sensitivity level for rejection are not only the shape and the size of the natural defect but
also their possible development during further manufacturing process or service life under severe
conditions. Extensive defects analysis on titanium and stainless steel welded tubes showed clearly
two groups in which typical defects can be divided:

Existing defects from the strip material (laps, blisters, slivers, …)

Defects caused by the welding process itself (porosities, lack of penetration, lack of fusion …),
or by forming (edges mismatch, …)

To adjust test equipment and set the rejection level, a standard defect is used. Its shape and di-
mensions is specified in the different usual standards. For instance, for Eddy Current (EC) testing
of Titanium welded tubing, ASTM B338 standard defect is one drilled hole of 0.78 mm diameter in
the parent metal. For Ultrasonic Test (UT), these calibration defects are usually several notches,
on ID and OD surface, in longitudinal or transverse direction.

The first step is then to adjust the test equipment, to optimize the detection of the standard defect.
It will consist, for UT test, in adjusting the probes angles for instance. When the output signal am-
plitude of standard defect is satisfactory, the rejection level can be set. To characterize this rejec-
tion level, we will use the quality criterion K:

K = amplitude of minimum natural defect rejected / amplitude of standard defect.

16
L. de Kerret

Most of the manufacturing specifications for welded tubes (ASTM B338, ASTM A450, R113L…)
specify a quality criterion K=1. This means that all natural defects with a signal amplitude lower
than the standard defect one will not be rejected. On the other hand, Valtimet has chosen to re-
duce this ratio, in order to perform a more severe control. This rejection level is determined
through a defects analysis and takes into account the standard deviation of the signal results after
minimum 10 passages of the standard. This generally leads use to use a quality criterion (quality
criterion K = ratio between natural defect amplitude and reference defect amplitude) which is more
stringent than ASTM standards requirements.

3.2 Air Pressure Test

Two kinds of testing are commonly used on welded condenser tubing: pressure differential testing
and air-under-water testing. For U-bent tubes, this test is usually replaced by a hydrostatic testing.

The air-under-water (AUW) testing is the traditional testing method which was first developed on
the market. It is performed by placing tubes in a well-lit tank of water while an operator walks the
length of the tank looking for bubbles. This test is very sensitive, and allows locating and identify-
ing a leakage defect…However, the detection of leaks is operator dependent as it is a visual me-
thod and thus subject to operator fatigue (8-hour shifts) ; it is not a reliable test at the industrial
scale. Valtimet used to perform AUW testing before to experiment the pressure differential test for
a better reliability of leak detection.

The pressure differential detective method is based on the measurement of the pressure drop over
a constant period of time. This pressure drop is measured by differential method, comparing the
tested tube pressure to a reference volume pressure. This method offers a better reliability and
repeatability of the test compared to air under water test and does not require subjective judgment.
Moreover, the recordings can be kept and thus the test results can be checked later.

The pneumatic testing of welded tubes, by either the pneumatic test or the air underwater method,
is a pressure leak test enabling to detect small thru-wall leaks due to defects like weld stoppages
but does not apply for all defects which do not provoke immediate tube leaks during the test pe-
riod. Leak tests have never been able to detect geometrical defects nor welding defects such as
lack of penetration or edges mismatch.

3.3 Eddy Current Test

Valtimet performs an EC test using the encircling differential coils method working generally at a
frequency of 10 to 30 kHz. EC test allows observing a conductivity variation in the tube; this im-
plies that either a volumetric defect resulting in a local cross-section variation, or a metallurgical
anomaly, will be seen.

17
L. de Kerret

Figures 18 and 19 show examples of defects detected by EC test, either on Stainless Steel tubes
or on Titanium tubes. There is one weld defect, made during the welding process, and one strip
defect, coming from the strip supplier.

OD view X20 A-A cross section X40

Figure 18: Defect on the external surface of the tube (sliver in the strip) detected by EC test
(K=0.3 - measured depth=0.22 mm – measured length: 13 mm)

1.57 mm length

Transverse view X30 Outside Diameter view X35

0.04 mm depth

Transverse view X200

Figure 19 : weld shrinkage on a stainless steel tube, detected by Eddy Current with K=0.58
micrographics and US & EC signals.

Those two defects have been detected by Eddy Current test, with a quality criterion set to K=0.5.
However, respecting the standards of ASTM, for example, a K=1 would be used , and those de-
fects would not be detected. This clearly demonstrates the utility of using a quality criterion lower
than 1, depending on the product. This choice is usually made by Valtimet for a new order, after a
trial campaign and a defect catalog study. The use of more severe rejection threshold permits a
reliable detection of short and punctual defects like rolling defects.

For equipment calibration, VALTIMET’s calibration notches are minimum three drilled holes not
larger than 0.787 mm in diameter, spaced out by around 500 mm and forming an angle of 120°

18
L. de Kerret

between each other, with one of them located in the weld. This is another improvement of usual
standards, which specify only one hole in the parent metal: the use of 3 holes allows to ensure a
correct centering of the tube versus EC coils, if the 3 signals are leveled. Moreover, one hole in
the weld allows focusing the control on weld defects detection.

The differential method compares actually one encircling coil response with the second coil re-
sponse. If a punctual defect is going through the first coil, and not yet through the second coil, it
will be detected. However, a long defect, like for example an edge mismatch, over several decime-
ters, will not be detected easily by this method. For that reason, it is necessary to add another
NDT in complement to EC test, which is Ultrasonic Test.

3.4 Ultrasonic Test

Ultrasonic test is the most global NDT used for welded tube manufacturing, although it is comple-
mentary with EC and air pressure test, therefore it cannot be used as unique replacement test.

Longitudinal control

Transverse control

Figure 20 : Ultrasonic Test principle

As explained on figure 20, Ultrasonic Test on welded tubing can be performed in two directions:
longitudinal and transverse. Those two methods will detect respectively longitudinal-shaped de-
fects and transverse-shaped defects. The equipment calibration, for those two methods, require
respectively longitudinal and transverse standard notches (on both ID and OD), usually of 10%
tube thickness depth.

For Titanium welded tubes, ASTM (B338), as most usual standards, require to perform a longitu-
dinal control only. For Stainless steel welded tubes, most standards do not even require any Ultra-

19
L. de Kerret

sonic Test. The coming examples will intend to show that those requirements may not be suffi-
cient, to ensure welded tubing integrity and reliability.

3.4.1 Longitudinal Ultrasonic Test

The Longitudinal Ultrasonic method has proved to be the only way to discriminate the most detri-
mental types of weld defects such as edges mismatch, lack of fusion, cracks…. Indeed, those
defects do not provoque any leakage during air pressure test, and are often regular over a conse-
quent length, therefore they cannot be stopped by EC test; However, they present a geometrical
irregularity which will act as a mirror, reflecting Ultrasonic waves.

Figures 21 to 23 provide some defects examples, in both Stainless Steel and Titanium, all de-
tected with Longitudinal Ultrasonic Test. For each defect, the sensivity level is indicated by the K
ratio, for both shot directions (L1 & L2).

Cross section X30

Measured crack depth: 0.12 mm
Sensivity levels: KL1 = 0.4 ; KL2 = 0.6

Cross section X200

Figure 21: ID crack and lack of fusion on a Stainless Steel tube, detected by Longitudinal UT

ID surface view, X50

Sensivity level: K = 0.7 ID surface view, X40

Figure 22: Hot crack in the weld, on a Ti Grade2 welded tube

20
L. de Kerret

Cross section X40 Cross section X200

Figure 23: Edges mismatch, on a Ti Grade2 welded tube (K=0.6 – measured depth=0.15 mm)

It is important to keep in mind that all those defects have been detected only thanks to longitudinal
UT control. Many of such defects in stainless steel welded tubes will not be detected by EC test,
therefore Valtimet emphasizes the need for Ultrasonic control on Stainless steel welded tubes.
Moreover, for each of those example, the sensivity level is lower than 1. This means that they
would not have been detected with a quality criterion of 1, respecting usual specifications.

3.4.2 Transverse Ultrasonic Test for Titanium welded tubes

UT inspection based on longitudinal calibration notches in the weld enables a reliable detection of
weld defects. Past investigations12,13 have however proven this test is not severe enough to detect
other kinds of defects which could be particularly harmful on titanium welded tubing. Indeed, Tita-
nium is a very reactive material, and porosities are susceptible to be trapped in the weld during
solidification, as in figure 25. Moreover, some shorts defects on the ID surface would not be de-
tected by UT. This kind of volumetric defects being too short for a reliable longitudinal Ultrasonic
detection, Transverse ultrasonic control is needful.

Figure 24 : Strip transversal defect on the external surface of a 22 x 0.5 mm Ti Grade2 tube
(KT1=0.4 - KT2=0.54 – measured length=2.1 mm)

21
L. de Kerret

Figure 25 : Weld defect (gas occlusions) of a 17 x 0.5 mm Ti Gr2 tube (K=0.5)

Figures 24 and 25 give examples of strip and weld defects which can only be detected by a UT
control based on transversal calibration notches in the weld. Those kinds of defects proved to be
leaking with a helium test.

This kind of inspection is therefore complementary to EC test and longitudinal UT control. A com-
bination of those three different tests is key to ensure full reliability of welded titanium tubes.

Figure 26: most frequent defects in welded tubing, and the detection method associated.

22
L. de Kerret

Conclusion:
When designing a condenser, the tubing material choice is to be made according to the future
operating conditions. For seawater usage, Titanium is the best technical solution, even though
super stainless steels can represent an interesting alternative, provided the alloy chosen suits to
the cooling water chemistry.

The manufacturing process of welded tubes stands also as a major parameter, especially for aus-
tenitic stainless steel tubes, to insure a corrosion resistance as high as the material itself. For in-
stance, attention must be paid to the weld cold-work: Roll Sink technology is proven to reduce the
potentiality of weld to become a preferential pitting site, unlike bead rolling technology. Also, Bright
annealing and laser welding technologies appears as improvements of respectively classical air-
annealing-pickling and TIG welding, reducing the localized corrosion susceptibility. Moreover,
VALTIMET strongly recommends the use of passivation for austenitic stainless steel condenser
welded tubes. This process increases indeed the pitting corrosion resistance of inside and outside
tube surface, in aerated conditions.

The best manufacturing process, transforming the most fit-for-purpose material, cannot do without
a strong Non-Destructive-Test policy. This is the keystone of the condenser tubing know-how,
which allows securing the tubes sanity, avoiding strip or weld defects of being delivered to con-
denser manufacturer. Unfortunately, the usual standards requests are often not severe enough,
and it is up to the welded tube manufacturer to strengthen the production standards. For instance,
the use of both ultrasonic control and eddy current inspection should be compulsory for stainless
steels, these two methods being complementary. Ultrasonic Transverse method also appears ne-
cessary, especially for titanium welded tubing, considering the typical titanium weld defect shape.
VALTIMET also recommends lowering the rejection level, to detect natural defects smaller than
standard calibration defect.

23
L. de Kerret

References
1. “Super-ferritic stainless steels – the cost-effective answer for heat transfer tubing”, by Daniel Janikowski
and Edward Blessman, NACE CORROSION 2008 conference & expo

2. LDX 2101 Duplex Stainless Steel technical sheet, Outokumpu

3. “AL 2003 LDSS As a Substitute for Type 316L”, John Dunn and David Bergstrom, Stainless Steel World
2006 Conference

4. “Market and Trends – Stainless Steel” Powerpoint presentation, by Anders Olsson, Norsk Staldag 2006

5. AL6XN Sourcebook, Allegheny Ludlum, 2002

6. “AL 29-4C alloy, A Superferritic Stainless Steel for Power Plant and Desalination Condenser Tubing”
brochure, by ATI Allegheny Ludlum, 2007

7. Duplex Stainless Steel brochure, Outokumpu, Feb 2007

8. Valtimet Internal Technical report, “Influence of a bead rolling operation on the corrosion resistance”,
P.Gerard, September 2007

9. Vallourec Research Aulnoye internal Technical report 2006-COR-06044, H. Marchebois, October 2006

10. Valourec Research Aulnoye internal technical report 2009-SCC-0163, C. Plennevaux, April 2009

11. Valourec Research Aulnoye internal technical report 2009SCC00141, H. El Alami, April 2009

12. Valinox Welded internal technical report GD/AR/CR/4008/93, MM. Durand and Rousset, April 1993

13. Valinox Welded internal technical report PG/RG/500596, P. Gerard and P. Gili, February 1996

24