Generalized Corrosion of the CANDU Steam Generator ... · To investigate the influence of pH water chemistry on corrosion of carbon steel SA 508 cl.2 material, corrosion experiments

1013.1

Generalized Corrosion of the

CANDU Steam Generator Tubesheet Material

Dumitra Lucan, Manuela Fulger Institute for Nuclear Research

P.O. Box 78, RO-115400, Campului Street no. 1, Mioveni, Romania [email protected], [email protected]

ABSTRACT

The secondary side environment is known to be affected significantly by several factors dependent on the balance of plant conditions (condenser leaks, condensate polishing, and coolant system materials) and operational conditions, particularly through their thermal-hydraulic effects. The presence of tube-tubesheet crevices and restricted flow areas within sludge or surface deposits provide for local concentration sites for various impurities, including acidic. The generalized corrosion can appear and affects the steam generator performances. It is very important to understand the generalized corrosion mechanism in the purpose to evaluate the quantities of corrosion products which exist in the steam generator after a determined period of operation.

The purpose of this work consists in the assessment of corrosion behaviour of the tubesheet material (carbon steel SA508 cl.2) at normal secondary circuit parameters (temperature-2600C, pressure-5.1MPa). The testing environment was the demineralised water without impurities, at different pH values regulated with morpholine and cyclohexylamine (all volatile treatment – AVT).

The results are presented like micrographies and graphics representing loss of metal by corrosion, corrosion rate, the total corrosion products, the adherent corrosion product, the released corrosion products and the release of the metal.

1 INTRODUCTION

By operation in aqueous environment at high temperature and pressure, the structural materials from secondary circuit cover with protective oxide films, which maintain the corrosion rate in admissible limits. A lot of potential factors exist, which conduct to degradation of the protective films and consequently to intensification of the corrosion processes. The existing experience of different nuclear power plants shows that the water chemistry has an important role in maintaining integrity of the protective oxide films. The presence of the crevices can involve certain conditions, when the pH-value can rise to 7 and 10.

To investigate the influence of pH water chemistry on corrosion of carbon steel SA 508 cl.2 material, corrosion experiments by static autoclaving and electrochemical methods were performed.

The experimental results allowed us to establish the contribution of the pH water chemistry in evolution of the steam generator tubesheet corrosion processes.

mailto:[email protected]

mailto:[email protected]

1013.2

2. EXPERIMENTAL

The principal goals of the experimental program referring of the most important steam generator material (carbon steel SA 508 cl.2) consisted in the assessment of the corrosion kinetics for the samples tested in aqueous environments having different pH values.

Corrosion experiments included in the present paper have been carried out on the SA 508 cl.2 samples by autoclaving at parameters specific for the secondary circuit of the CANDU steam generator: temperature 260ºC, pressure 5.1MPa. The testing environments utilized were demineralised water with pH = 7.5, 8.5 and 9.5 (AVT – all volatile treatment).

The testing periods were 1896 hours, 1848 hours and and 1896 hours respectively. The specimens used were from steam generator tubesheet material (carbon steel SA

508 cl.2) and had a parallelepipedic shape having the dimensions 15x10x2 mm. Experimental work included: gravimetric analyses, optical microscopic analyses and

electrochemical measurements (potentiodynamic polarization and impedance). 3. RESULTS AND DISCUSSION

The goal of the work consists in the assessment of the kinetics corrosion for the carbon steel SA 508 cl.2 -material of the tubesheet- tested in demineralised water with different pH values and the experimental results processing in the purpose to including in a future database steam generator.

To investigate the water chemistry effects on characteristics of corrosive films formed on SA 508 cl.2 material, a lot of corrosion experiments by electrochemical methods and static autoclaving were performed. The electrochemical determinations were performed by potentiostatic method in aqueous solutions with different pH, at room temperature.

Some examples of experimental results for the testing of the SA 508 cl.2 samples for different times in demineralised water environments with pH=7.5, 8.5 and 9.5 (AVT) at secondary circuit steam generator specifica parameters (2600C and 5.1MPa) are presented in the Fig. 1 ÷ Fig. 3.

a)

b)

Fig. 1 Surface morphology (x200) (a) and aspect of the superficial layer (x1000) (b) for SA 508 cl.2 exposed 1896 h in demineralised water, pH=7.5 AVT, t=260 C and p=5.1MPa0

Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, Sept. 8-11, 2008

1013.3

a)

b)

Fig. 2 Surface morphology (x200) (a) and aspect of the superficial layer (x1000) (b) for SA508 cl.2 exposed 1848 h in demineralised water, pH=8,5 AVT, t=2600C and p=5.1MPa

a)

b)

Fig. 3 Surface morphology (x200) (a) and aspect of the superficial layer (x1000) (b) for SA 508 cl.2 exposed 1896 h in demineralised water, pH=9.5 AVT, t=2600C and p=5.1MPa

The exposure times for the metallographic analysis of samples tested at a pH = 7.5

were: 312 h, 624 h, 864 h, 960 h, 1272h, 1584 h and 1896 h. The SA 508 cl.2 samples, tested for 312 hours in demineralised water at a pH = 7.5 a uniform, continuous, adherent oxide layer is observed, whose thickness is smaller than or equal to 1μm.

When the testing time was 624 hours in the same conditions the presence of the oxide is noticed on sample surfaces, bright black in color. The sample surfaces are entirely covered by oxide; there are no uncovered spots while the visual aspect is almost identical for the entire surface. Thickness of the oxide layer is about 1.2μm. The aspect of the sample in demineralised water at a pH = 7.5 is presented in Fig.1a. The oxide layer on the samples tested for 1272 hours is uniform, continuous, adherent, while its thickness ranges between 1.0÷1.5μm. The results of the 1584 hours exposure was the occurrence of an oxide layer with a uniform thickness of 1.5μm÷1.7μm. The thickness of the oxide layer existing on the samples tested for 1896 hours is about 2μm, Fig.1b.

The exposure times for the metallographic analysis of the samples tested at a pH = 8.5 were: 288 hours, 576 hours, 960 hours, 1272 hours, 1560 hours and 1848 hours. On the SA 508 cl.2 samples, tested for 288 hours in demineralised water with a pH=8.5 adjusted with morpholine and cyclohexylamine, at 2600C and a pressure of 5.1MPa a uniform, continuous and adherent oxide layer is noticed, which thickness is smaller than or equal to 1μm. The surface morphology for some samples exposed for 288 hours in demineralised water at a pH=8.5 (AVT) at parameters specific to the steam generator secondary circuit shows the presence of the oxide, its color being bright black. The sample surfaces are completely covered by oxide, there are no uncovered spots, while the visual aspect is almost identical for


1013.4

the entire surface. It is to be noticed that the oxide on the samples tested 288 hours in demineralised water with a pH=8.5 is, in this case, uniform, continuous, adherent, and its thickness ranges between 0.8μm and 1μm. The oxide is uniform, with brown-red shadows and formed in continuous film on the samples surface. For the SA 508 cl.2 samples tested for 576 hours in demineralised water with a pH=8.5 the result of the exposure was the formation of an oxide layer with a uniform thickness of 1.2μm. In this case the oxide uniformity is noticed. The aspect of the oxide layer existing on the samples tested 576 hours is shown that the film thickness on these samples is about 1.5μm. The surface morphology for the samples exposed for 576 hours has an uniform aspect.

The aspect of the oxide layer existing on the surface of samples tested for 960 hours in demineralised water with pH=8.5. The film is uniform, continuous and adherent. The uniformity and continuity of the oxide film is observed and the surface morphology for the samples exposed for 960 hours in demineralised water with pH=8.5.

The oxide film is uniform, continuous, adherent and has a thickness smaller than 1μm for a sample exposed for 1272 hours in demineralised water with a pH=8.5 in conditions specific to the operation of the secondary circuit.

The aspect of the oxide layer on the surface of samples tested for 1560 hours in demineralised water with a pH=8.5 is uniform, continuous and adherent. The uniformity and continuity of the oxide layer can be noticed by the surface morphology of samples exposed for hours in demineralised water with a pH=8.5. The oxide layer is uniform, continuous, adherent and has a thickness smaller than 1.8μm for 1560 hours of exposure and 2.2μm for 1848 hours.

The aspect of the oxide layer and the surface morphology, respectively, for a sample exposed for 1848 hours in demineralised water at a pH=8.5 and pH=9.5 under operating conditions specific to the secondary circuit are presented in Fig.2 and Fig.3. 3.1 Comparison of outputs of tests performed at pH=7.5, pH=8.5 and pH=9.5

Fig. 4÷Fig. 10 comparatively present the corrosion kinetics for: metal loss by corrosion; corrosion rate; totally formed corrosion products; adherent corrosion products; released corrosion products; corrosion products release, and the release rate of metal at a pH = 7.5, 8.5 and 9.5, respectively. For the weight loss due to corrosion and corrosion rate it is noticed that, in the case of a pH = 9.5 these have the smallest values (Fig. 4 and Fig. 5). Also, in the case of totally formed corrosion products the smallest values have been obtained in exposure in solution at a pH=9.5 (Fig.6). In the case of adherent corrosion products, the smallest values have been reached for the solution with a pH = 7.5 (Fig.7), but the values for the tests at pH = 9.5 solution range in their immediate neighbourhood, without significant differences between them. In the case of released corrosion products, the release rate for products and metal the highest values are obtained for the samples tested in the solution with a pH =7.5 (Fig. 8 ÷ Fig. 10). This can be explained by the fact that magnetite solubility has higher values for solutions with a smaller pH. An example in this sense is the fact that, at 200ºC the magnetite solubility is 2 μg/kg for a solution with pH=9.5 while for a pH=8.5 the magnetite solubility increases, reaching 60μg/kg. It can be stated that corrosion kinetics for: metal loss by corrosion; corrosion rate; totally formed corrosion products; adherent corrosion products; released corrosion products and the metal release rate at a pH=7.5, 8.5 and 9.5, respectively, evolve following power-type or logarithmic laws: the smallest corrosion rates are obtained in the case of exposure in a pH=9.5 solution, [1] ÷ [9].


1013.5

y = 2,2801x0,3308

R2 = 0,917

y = 2,8722x0,25

R2 = 0,9141

y = 1,6701x0,3975

R2 = 0,9795

0,000

2,000

4,000

6,000

8,000

10,000

12,000

0 10 20 30 40 50 60 70 80 90Time (days)

Loss

wei

ght (

g/m

2)

pH=8,5pH=7,5pH=9,5Power (pH=8,5)Power (pH=7,5)Power (pH=9,5)

y = 2,2801x-0,6692

R2 = 0,9784

y = 2,5309x-0,717

R2 = 0,9599

y = 1,6701x-0,6025

R2 = 0,991

0,000

0,050

0,100

0,150

0,200

0,250

0,300

0,350

0,400

0,450

0,500

0 10 20 30 40 50 60 70 80 90Time (days)

Cor

rosi

on ra

te (g

/m2.

day)

pH=8,5pH=7,5pH=9,5Power (pH=7,5)Power (pH=8,5)Power (pH=7,5)Power (pH=9,5)

Fig. 4 Loss of metal by corrosion vs. time: ■ pH=7.5;♦ pH=8.5,▲ pH =9.5

Fig. 5 Corrosion rate vs. time: ■ pH=7.5;♦ pH=8.5,▲ pH =9.5

y = 3,8951x0,2563

R2 = 0,9234

y = 3,1238x0,3308

R2 = 0,917

y = 2,3047x0,3975

R2 = 0,9795

0,000

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

0 10 20 30 40 50 60 70 80 90Time(days)

Tota

l cor

rosi

on p

rodu

cts

(g/m

2)


Fig. 6 Total corrosion products vs. time: ■ pH=7.5;♦ pH=8.5,▲ pH =9.5

y = 2,2775Ln(x) - 0,8647R2 = 0,9508

y = 2,5249Ln(x) - 0,7886R2 = 0,9132

y = 2,7205Ln(x) - 1,8979R2 = 0,9534

0,000

2,000

4,000

6,000

8,000

10,000

12,000

0 10 20 30 40 50 60 70 80 90Time(days)

Adh

eren

t cor

rosi

on p

rodu

ct (g

/m2)

pH=8,5pH=7,5pH=9,5Log. (pH=7,5)Log. (pH=8,5)Log. (pH=9,5)

y = 0,4485Ln(x) + 0,4022R2 = 0,987

y = 0,5335Ln(x) + 0,8059R2 = 0,9606

y = 0,8542Ln(x) - 0,9551R2 = 0,9447

0,000

0,500

1,000

1,500

2,000

2,500

3,000

3,500

0 10 20 30 40 50 60 70 80 90Timp (zile)

Rel

ease

d co

rros

ion

prod

ucts

(g/m

2)

pH=8,5pH=7,59,5Log. (pH=8,5)Log. (pH=7,5)Log. (9,5)

Fig. 7 Adherent corrosion products vs. time

■ pH=7.5;♦ pH=8.5,▲ pH =9.5 Fig. 8 Released corrosion products vs. time

■ pH=7.5;♦ pH=8.5,▲ pH =9.5

y = 0,891x-0,765

R2 = 0,9348

y = 1,6706x-0,8968

R2 = 0,9531

y = 0,3688x-0,5291

R2 = 0,9626

0,000

0,020

0,040

0,060

0,080

0,100

0,120

0,140

0,160

0,180

0,200

0 10 20 30 40 50 60 70 80 90Time (days)

Rel

ease

rate

of c

orro

sion

pro

duct

s (g

/m2.

day)


y = 0,6341x-0,7562

R2 = 0,9376

y = 0,972x-0,8306

R2 = 0,9224

y = 0,2673x-0,5291

R2 = 0,9626

0,000

0,020

0,040

0,060

0,080

0,100

0,120

0,140

0,160

0 10 20 30 40 50 60 70 80 90Time (zile)

Rel

ease

rate

of m

etal

(g/

m2.

day)


Fig. 9 Release rate of corrosion products vs. time

■ pH=7.5;♦ pH=8.5,▲ pH =9.5 Fig. 10 Release rate of metal vs. time ■ pH=7.5;♦ pH=8.5,▲ pH =9.5


1013.6

In the Table 1 were presented the fitting and the equations for the corrosion kinetic in

Fig.1 ÷ Fig.10. Table 1 The equations for the kinetic corrosion specific parameters

No. Parameter pH=7.5 pH=8.5 pH=9.5

1. Loss of metal by corrosion

y = 2.8722x0.25 y = 2.2801 x0.3308 y = 1.6701 x0.3975

2. Corrosion rate y = 2.5309 x-0.717 y = 2.2801 x-0.6692 y = 1.6701.x-0.6025

3. Total corrosion products

y = 2.4755ln(x)+1.0074 y = 3.109 ln(x)-0.6686 y = 3.8951 x0.2563

4. Adherent corrosion products

y = 2.2775ln(x) -0.8647 y = 2.5249 ln(x)-0.7886 y = 2.7205ln(x)-1.8979

5. Released corrosion y = 0.5335ln(x)+0.8059 y = 0.4485ln(x)+0.4022 y = 0.8542ln(x)–0.9551 6. Release rate of

corrosion products y = 1.6706x–0..8968 y = 0.891 x –0.765 y = 0.3688 x –0.5291

7. Release rate of metal y = 0.972x–0.8306 y = 0.6341 x –0.7562 y = 0.2673 x –0.5291

Fig. 11 and Fig.12 show the results of electrochemical measurements performed by the potentiodynamic method. By assessing the corrosion rates measured in demineralised water at a pH=9.5 for samples tested by autoclaving at a pH=9.5: PD 7 – as received, PD 8 – tested 240 hours and PD 9- tested 3600 hours and pH=9.5 for samples tested by autoclaving at a pH=8.5: PD 1 - tested 288 h, PD 3 – tested 960 h, PD 4 –tested 1272 h, PD 6 – tested 1848 h, it resulted that the oxide films formed on the samples surface at higher exposure times provides them a relatively high corrosion resistance.

T

T

T

E(m

V)

log(I)(log(A))

-200

-400

-600

-800

0

200

400

600

800

1000

1200

1400

-3-4-5-6-7-8-9-10 -2

Fig. 11 Potentiodynamic curves for SA508 cl.2 tested at pH=9.5: PD 7- as received, PD 8 – tested 240 h, PD

9 –tested 3600h

E(m

V)

log(I)(log(A))

-200

-400

0

200

400

600

800

1000

1200

1400

-8-9-10-11-12 -7

PD6 PD8

PD9

PD7 PD4 PD1

PD3

Fig. 12 Potentiodynamic curves for SA508 cl.2 tested at pH=8.5: PD 1- tested 288h, PD 3 – tested 960 h, PD 4 –tested 1272h, PD 6 – tested 1848h

CONCLUSIONS

Corrosion and mechanical damages within the steam generator are the result of complex interactions among various factors:


1013.7

� strict control of water chemistry; � proper thermo-hydraulic design; � selection of construction materials for the steam generator; � utilization of materials compatible for the entire secondary circuit; � conditions of completion of equipment and facilities; � the operation manner.

Corrosion can be generalized or localized, the latter having the following forms:

denting, wastage, pitting, SCC (= Stress Corrosion Cracking).

Corrosion is in close correlation with the composition of metallic materials, with the properties of the environment where these work and with the operating parameters of the equipment.

When the chemical parameters of the coolant are maintained at the values imposed by

specifications, corrosion is not a problem but from this point of view it is very important the knowledge of the deviation from the normal operation conditions of the steam generator.

In order to attain the objective of the paper (i.e. establishing the corrosion kinetics of

the carbon steel SA508 cl.2 alloy and the systematization of experimental results for the implementation in a database related to the CANDU steam generator), corrosion tests have been conducted at parameters specific to the secondary circuit of the steam generator in demineralised water at a pH=7.5, pH=8.5 and pH=9.5.

The exposure environment was demineralised water with a pH=7.5, pH=8.5 and

pH=9.5 adjusted with morpholine and cyclohexylamine, at a temperature and pressure of 2600C and 5.1MPa, respectively.

Corrosion testing has been performed for carbon steel SA508 cl.2 samples for 1896

hours at pH=7.5 and 1848 hours at pH=8.5.

By using gravimetry and descaling of filmed samples exposed for certain intervals of time, the corrosion kinetics of the carbon steel SA508 cl.2 has been established.

The films formed on samples after autoclaving and the morphologies of sample

surfaces have been assessed by metallographic microscopy.

In the case of samples tested in solution of pH=7.5 for 312 hours the thickness of the films is smaller than or equal to 1μm, this one increasing with the increase of the testing time, reaching 2μm after 1896 hours of testing.

In the case of samples tested in solution of a pH=8.5 for 288 hours the thickness of the

films is between 0.8μm and 1μm, this one increasing with the increase of the testing time, reaching 2.2μm after 1848 hours of testing.

The corrosion kinetics has been established for the corrosion-induced loss of metal,

corrosion rate, totally formed corrosion products, adherent corrosion products, released corrosion products, release of corrosion products and metal release.


1013.8

A comparison is presented between the corrosion kinetics for the normal value of the operation pH = 9.5, and also for the values 7.5 and 8.5, which made the objective of the present paper.

For the corrosion-induced loss of weight and for the corrosion rate it is noticed that in

the case of pH=9.5 solution these have the smallest values.

In the case of released corrosion products, their release rate and metal release rate, the highest values are obtained for the samples tested in a pH=7.5 solution. This can be explained by the fact that magnetite solubility has higher values for solutions with a smaller pH.

REFERENCES [1]. D. Lucan, M., Fulger, L., Velciu, Gh., Savu, Corrosion Kinetic of Incoloy 800 in Normal

Operation of Steam Generator, Institute for Nuclear Research Pitesti, Internal Report 6749, 2003.

[2]. D. Lucan, M., Fulger, L., Velciu, Corrosion Kinetic of Carbon Steel SA 508 cl.2 at Different pH Values of the Cooling Water from CANDU Steam Generator, Institute for Nuclear Research Pitesti, Internal Report 7539, 2006.

[3]. D. Lucan, M., Fulger, Ghe., Jinescu, The Ageing of CANDU Steam Generator Due to Localized Corrosion, 5-th International Seminar on Primary and Secondary Side Water Chemistry of Nuclear Power Plants, Eger, Hungary Sept. 17-20, 2001.

[4]. D. Lucan, M., Fulger, Gh., Savu, L., Velciu, Experimental Research Concerning CANDU Steam Generator Components, International Congress on Advanced Nuclear Power Plants, Seoul, KOREA, May 15-19, 2005.

[5]. D. Lucan, M., Fulger, Influence of the Silicon Compounds on the SA 508 Corrosion in High Temperature and Pressure Water, International Congress on Advances in Nuclear Power Plants, Nice, France, May 13-18, 2007.

[6]. J.A. Gorman, A.P.L., Turner, R.W., Staehle, A.P., McIlree, Control of Corrosion on the Secondary Side of Steam generators, Proceeding of the Meeting, Improving the Understanding and Control of Corrosion on the Secondary Side of Steam Generators, Virginia, USA, October 9-13, 1995.

[7]. G.R. Rundell, Oxidation Kinetics of Heat Resistant Alloys, Avesta Corrosion Management, Sweden, no.2, 1985.

[8]. D.H. Lister, Corrosion Products in Power Generating Systems, AECL 6877, Canada, 1980.

[9]. G.F. Taylor, Corrosion Monitoring in CANDU Nuclear Generator Stations, AECL 5648, Canada, 1977.


Documents

Generalized Corrosion of the CANDU Steam Generator ... · To investigate the influence of pH water chemistry on corrosion of carbon steel SA 508 cl.2 material, corrosion experiments