14th International Conference on the Properties of Water and Steam in Kyoto

Corrosion Behavior of Steam Turbine Materials for Geothermal Power Plants

Hiroshi Takaku1 , Li-Bin Niu1 , Hiromasa Kawanishi2 , Naoki Takamura2 , Kensuke Monma 2 ,
Yoshihiro Sakai3 , Mitsuo Yamashita 4 and Kunio Shiokawa5
1
Faculty of Engineering, Shinshu University, 4-17-1 Wakasato, Nagano 380-8553 Japan

E-mail: takakuh@gipwc.shinshu-u.ac.jp
2
Student of Graduate School of Eng., Shinshu University, 4-17-1 Wakasato, Nagano 380-8553 Japan
3
Fuji Electric Systems Co.Ltd., 1-1 Tanabeshinden, Kawasaki 210-9530 Japan
4
Fuji Electric Advanced Technology Co. Ltd., 4-18-1, Tsukama, Matsumoto City 390-0821 Japan
5
Fuji Electric Advanced Technology Co. Ltd., 1-1 Tanabeshinden, Kawasaki 210-9530 Japan

To evaluate the corrosion behavior of steam turbine materials for geotherma l power
plants, corrosion tests are conducted in the simulated geothermal water with corrosive
chemicals. A newly developed high purity 9%CrMoV steel for rotor and a modified
heat-treated 16Cr-4Ni steel for blade had a good general corrosion resistance, compared
to the conventional materials. Due to the galvanic corrosion effect, rotor materials were
subjected to the more severe crevice corrosion than blade materials. The SCC initiation
and SCC propagation did not occur up to the 4000 h corrosion test in the H2 S free waters,
although the sulfide induced stress corrosion cracking occurred in water with hydrogen
sulfide. The combinations of corrosive chemicals on the electrochemical corrosion
behavior were investigated for specimens with and without oxide films formed in the
simulated geothermal water. The corrosion behaviors were discussed mainly based on the
oxide film characteristics, material properties and the water quality.

1. Introduction the systematic data and knowledge have not been

obtained sufficiently for maintenance, design and
As compared with the conventional fossil fuels, others of the steam turbine equipments in the
the geothermal energy is one of the clean and geothermal power plant.
renewable energies because of almost no carbon In this work, some kinds of corrosion tests of the
dioxide (CO2 ) emission, and it could produce the steam turbine materials for the geothermal power
vast energies most effectively by the existing plants are experimentally conducted in a simulate
technologies or by a little modification of the geothermal water, to evaluate the generic corrosion
present technologies. The geothermal fluid, water characteristics.
and steam, and their mixtures, of which quality
varies site to site contains many kinds of the 2. Experimental Procedures
corrosive chemicals such as chlorides, sulfates,
hydrogen sulfide etc. which induce the corrosion Steam turbine materials tested are shown in Table
problems. The corrosion of components of the 1. That is two rotor materials of a newly developed
geothermal plants is one of the most important high purity 9%CrMoV steel (steel B) and the
subjects to be evaluated [1-3]. On the other hands, p r e s e n t l y u s e d 1%CrMoV steel ( steel B ) ,

Table 1. Chemical composition of steels used (mass%) (*: not analyzed)

Material C Si Mu Ni Cr Mo V Cu
9%CrMoV (Steel A) 0.15 0.09 0.09 1.21 9.69 1.35 0.22 *
1%CrMoV (Steel B) 0.28 0.1 0.69 0.7 1.16 1.14 0.27 *
13%Cr (Steel C) 0.21 0.32 0.62 0.46 13.36 * * *
16Cr-4Ni (Steel D) 0.038 0.27 0.36 4.32 15.61 0.08 * 3.1

718
14th International Conference on the Properties of Water and Steam in Kyoto

27.5 8 4 24

6.25

4
P

1
R3

20
4
1
20

36

4
50
58
15
6.25

8 30

0.2
37.5
Precrack
Fig.1. Configuration and size of Blunt Notch
Compact Tension (BNCT) specimen. 2.0

Table 2. Applied maximum stress for BNCT Fig. 2. Configuration and size of Double
specimens. Cantilever Beam (DCB) specimen.
Specimen Number
1 2 3 4 5 6 maximum stress can be calculated from the
Materials
following equation.
0.8 0.2 1.0 0.2 1.2 0.2
A/9%CrMo 663 663 829 829 995 995 2K/(R)1/2
B/1%CrMo 546 546 682 682 818 818
C/13%Cr 624 624 780 780 936 936 where is the stress, K is the stress intensity
D/16Cr-4Ni 658 658 823 823 988 988 factor and R is the radius of the hole. K is
calculated according to the following equation [5].
Unit: MPa, 0.2 : 0.2% proof stress at 90
KEh{3h(a0.6h) 2 h3 }1/2
and two blade materials of the conventional 13%Cr {4(a0.6h) 3 h2 a }-1
steel (steel C) and the 16%Cr-4%Ni steel (steel D)
subjected to the modified heat treatment.
where E is the Youngs modulus, is the crack
The corrosion test was conducted at 90 in a opening displacement, h is the half value of the
simulated geothermal water which contains 2000 specimen height and a is the crack length.
ppm Cl- as NaCl, 50 ppm SO4 2- as Na2 SO4 and As shown in Tables 2 and 3, the levels of applied
approximately 3000 ppm CO2 by a continuous gas stress and stress intensity are variable for BNCT and
bubbling method as the main corrosive chemicals. DCB specimens, respectively.
The effect of corrosive chemicals on the general Anodic polarization curves were measured as one
corrosion and also the crevice corrosion were of the electrochemical corrosion tests, to evaluate
investigated. The general corrosion test using the the existence of the passivation state and the
plate type specimen with thickness 2 mm and the stability of the corrosion films during the stopping
crevice corrosion test using the coupled specimen period of power plants. The measurement according
with the rotor and blade materials were conducted. to JIS G0579 is carried out using the automatic
The stress corrosion cracking sensitivity (SCC) equipment, in the various simulated geothermal
was evaluated using the two kinds of specimens waters with the variable corrosive chemicals. The
shown in Figures 1 and 2. Namely, the blunt notch system is composed of the three electrodes which
compact tension (BNCT) specimen is used for SCC are the sample electrode, the Pt couterpart electrode
initiation and the double cantilever beam compact and the Hg/HgCl reference one. The scanning speed
tension (DCB) specimen is for SCC propagation of the potential is 20 mV/min in this work.
[4].
The stress is applied to the zone near the circular 3. Results and Discussion
tip by the inserted wedge in BNCT specimens. 3.1 General Corrosion and Crevice Corrosion
Thereby, the maximum stress is formed at the Figure 3 shows the general corrosion rate by mdd
inside of the hole -top on BNCT specimens. The [mg/(dm2 day)]. It was clarified that the general

719
14th International Conference on the Properties of Water and Steam in Kyoto

Table 3. Applied stress intensity factor for DCB specimens.

Specimen Number
Materials
1 2 3 4 5 6
K 1901 1897 2196 2184 2494 2512
A/9%CrMoV
a 26.05 26.09 25.99 26.08 26.09 25.97
K 1867 1945 2229 2268 2597 2592
B/1%CrMoV
a 25.99 25.77 26.07 25.78 25.75 25.78
K 1868 1884 2239 2237 2600 290
C/13%Cr
a 26.22 26.08 26.07 26.09 26.04 26.10
K 1874 1921 2192 2231 2583 2556
D/16Cr-4Ni
a 26.23 25.81 26.07 25.77 25.77 25.95
K: Stress intensity factor/MPamm1/2 , a: Crack length/mm

300 300
Steel A Steel B

200 200
Corrosion rate (mdd)

100 100

0 0
3 3
Steel C Steel D

2 2

1 1

0 0
500 1000 2000 4000 500 1000 2000 4000
Corrosion time (h)
Fig. 3. General corrosion rate, mdd [mg/(dm2 day)].

corrosion rate of the blade materials (steel C and D) of Cr2 O3 by the X-ray photoelectron spectroscopy
shows smaller approximately 1/100 than that of (XPS). Although these tight films may suppress the
rotor materials (Steel A and B), and also that the general corrosion, the localized break of these films
new rotor material of steel A and the modified due to Cl- ion may be easy to induce the corrosion
heat-treated blade of steel D have the very high pits.
general corrosion resistance. The aspect of the Figure 5 shows the typical aspect of corroded
as-scaled and de-scaled surfaces after the 4000 h. surfaces in the crevice using coupled specimens of
corrosion test is shown in Fig. 4. For steel B, the blade materials (steels C and D) and rotor steel B. It
corroded surface is very porous and also no is clarified that the steel B was subjected to the
passivation film is formed. On the other hand, for severe corrosion as compared with the both blade
other three materials with the high Cr content, the materials. It is considered that this phenomenon is
thin film is formed on surfaces of under the due to the galvanic corrosion effect in addition to
as-scaled condition, and also the corrosion pits are the electric cell effect due to the difference in the
observed in these materials. It is found that the oxygen concentration between the crevice and its
corrosion films of these materials having the high outside water.
general corrosion resistance are mainly composed 3.2 SCC Sensitivity As shown in Table 2, the SCC

720
14th International Conference on the Properties of Water and Steam in Kyoto

As-scaled Descaled As-scaled Descaled

A A B B

A A B B

C C D D

C C D D
250m

Fig. 4. Aspect of as-scaled and descaled surfaces.

initiation test was conducted under the three stress 3.3 Electrochemical Corrosion Behavior By the
levels condition of the maximum applied stress 0.8 anodic polarization method, the corrosion tendency
0.2 , 1.0 0.2 and 1.20.2 at 90 strengths in each could easily evaluate from the existence of the
material, up to the test duration 4000 h. However, it passivation region, its length, the current density in
is clarified that no SCC initiated for all materials. passivation region, and so on. Figures 7 and 8 show
Are shown in Fig. 6 the fractographs after the SCC effects of the Cl- ion concentration on the corrosion
propagation test conducted for 4000 h at the applied for the materials polished by Emery paper. From
maximum stress intensity. No cracks elongated the figures, it is clarified that the Cl- ion accelerates
from the precrack as SCC for all materials in this the corrosion of all materials, and that the corrosion
work. The hydrogen sulfide induced SCC (SSC), tendency is almost the same one tested in the
especially for 13%Cr steel, occurred in our other simulated geothermal water at 90 . Figure 9
work conducted in the water containing the H2 S [3]. shows the result tested in the water sampled from
It is considered that the environmental- induced the corrosion loop water in which many kinds of
crackings such as SCC and corrosion fatigue may metallic ions were dissolved. In this case, were
be rather suppressed when the general corrosion observed the activated dissolution peaks for all
will be dominant in the relatively severe corrosive materials except steel B, and this water may be very
environment. severe for the materials corrosion due to the

721
14th International Conference on the Properties of Water and Steam in Kyoto

250m

Fig. 5. Aspect of descaled surfaces after 2000 h crevice corrosion test.

Mechanical Mechanical
fracture fracture

Precrack Precrack
A B

Mechanical Mechanical
fracture fracture

2mm
Precrack Precrack
C D
Fig. 6. Fracture surfaces of DCB specimens after SCC test for 4000 h (K=2200 MPa/mm1/2 ).

722
 14th International Conference on the Properties of Water and Steam in Kyoto
Current Density (A/cm2 )

Current Density (A/cm2 )

0.8 0.8
0.7 : Steel A 0.7 : Steel A
0.6 : Steel B 0.6
: Steel B
0.5 : Steel C 0.5
0.4 0.4 : Steel C
: Steel D
0.3 0.3 : Steel D
0.2 0.2
0.1 0.1
0 0
-0.1 -0.1
-0.2 -0.2
-1.5 -1.2 -0.9 -0.6 -0.3 0 0.3 -1.5 -1.2 -0.9 -0.6 -0.3 0 0.3

Potential (V) Potential (V)

Fig. 7. Anodic polarization curve in water with 50 Fig. 8. Anodic polarization curve in water with 50
ppm SO4 2- + 1800 ppm Cl- +CO2 . ppm SO4 2- + 10000 ppm Cl- +CO2 .
Current Density (A/cm2 )

0.8

Current Density (A/cm2 )

: Steel A 0.8
0.7 : Steel A
: Steel B 0.7
0.6 : Steel B
: Steel C 0.6
0.5 0.5 : Steel C
0.4 : Steel D : Steel D
0.4
0.3
0.2

0.3
0.2
0.1 0.1
0 0
-0.1 -0.1
-0.2 -0.2
-1.5 -1.2 -0.9 -0.6 -0.3 0 0.3 -1.5 -1.2 -0.9 -0.6 -0.3 0 0.3

Potential (V) Potential (V)

Fig. 9. Anodic polarization curve in water with 50 Fig. 10. Anodic polarization curve of materials
ppm SO4 2- + 1800 ppm Cl- +CO2 + metallic ions. with scale by 500 h corrosion in water with 50
ppm SO4 2- + 10000 ppm Cl- + saturated air at
room temperature.
dissolved metallic ions. Figure 10 shows the result
of specimens with the scale (film) formed by the
500 h test in the simulated geothermal waters (3) The corrosion by the anodic polarization
mentioned above. Compared with the other curves, measurement corresponds with that in a simulated
it is clear that the corrosion films of all materials 90 water. The films formed by corrosion in a
were extremely stable and had the good corrosion simulated 90 water were stable even in the
resistance in the severe corrosive water. severe corrosive waters.
4. Conclusions References
Essential results obtained are as follows: [1] Y. Sakai, The Thermal and Nuclear Power, 50, 778
(1) 9%CrMoV steel for new-developed rotor and (1999).
16Cr-4Ni steel for modified heat-treated blade have [2] Geothermal Education Office, Homepage (2003).
a good general corrosion and crevice corrosion [3] Y. Sakai, M.Yamashita, K. Shiokawa, L.B.Niu, and H.
resistance. The Cr content in materials controls the Takaku, Proc. ICOPE-03, 3, 297 (2003).
corrosion. [4] H. Takaku and H. Kimura, Proc. Inter. Symp. on
(2) All materials showed no SCC sensitivity in a Plant Aging and Life Prediction of Corrodible
Structures, 793 (1995).
simulated geothermal water without H2 S, although
[5] Damage Tolerant Handbook Part 2, Appendix 11.0,
the H2 S in a water accelerates the SCC. Battelle Columbus Lab. (1975).

723