AECL-5849

ATOMIC ENERGY W S & L'ÉNERGIE ATOMIQUEOF CANADA UMITED f & J T DU CANADA LIMITÉE

EFFECT OF CONDENSER WATER IN-LEAKAGE ON

STEAM GENERATOR WATER CHEMISTRY

Paper no. IWC-77-12

by

P.V. BALAKRISHNAN

38th Annual AAeeting of the International Water Conference

Pittsburgh, Pennsylvania, November 1-3, 1977

Chalk River Nuclear Laboratories

Chalk River, Ontario

January 1978

EFFECT OF CONDENSER WATER IN-LEAKAGEON STEAM GENERATOR WATER CHEMISTRY

Paper IWC-77-12 Presented at 38th Annual Meeting |International Water Conference *

Pittsburgh, Pennsylvania1-3 November 1977

Atomic Energy of Canada LimitedChalk River Nuclear LaboratoriesChalk River, Ontario KOJ 1JO

January 1978

AECL-5849

Effet des infiltrations d'eau du condenseur sur lachimie de l'eau des générateurs de vapeur "̂

par

P.V. Balakrishnan

Résumé

Des environnements corrosifs peuvent être engendrés dansdes générateurs de vapeur à partir d'infiltrations d'eau derefroidissement dans le condenseur. Une évaluation théoriqueainsi qu'expérimentale de l'agressivité de tels environnementsest effectuée pour les eaux de refroidissement du condenseurutilisées dans les centrales nucléaires CANDU-PHW*. Les calculsont montré que des solutions de chlorure fortement concentrées -acides dans le cas d'infiltrations d'eau de mer et alcalinesdans les autres cas considérés - seraient produites à l'intérieurdu générateur de vapeur. Des expériences effectuées dans unmodèle de chaudière ont montré que des infiltrations d'eau demer provoquent une corrosion rapide des composants en acier aucarbone seulement lorsque l'on utilise AVT** pour le contrôlede la chimie de l'eau. L'emploi d'un réactif non volatil,^comme dans le traitement au phosphate conforme a permis d'éviterla corrosion rapide de l'acier au carbone.

Selon nos études, le traitement au phosphate conformedurant les infiltrations d'eau de mer semble souhaitable.

Document IWC-77-12 présenté à la 38 e m e Assemblée annuellede l'International Water Conference, Pittsburgh, Pennsylvanie,1 - 3 novembre 1977.

CANada Deuterium Uranium - eau lourde pressurisée

* *AVT - Ail Volatile Treatment

L'Energie Atomique du Canada, LimitéeLaboratoires Nucléaires de Chalk River

Chalk River, Ontario KO j 1J0T • -imo AECL-5849Janvier 1978

EFFECT OF CONDENSER WATER IN-LEAKAGEON STEAM GENERATOR WATER CHEMISTRYt

\

by

P.V. Balakrishnan

ABSTRACT

Corrosive environments may be generated within steamgenerators from condenser cooling water in-leakage. Theoreticalas well as experimental evaluation of the aggressiveness ofsuch environments is being carried out for the condenser-coolingwaters used at CANDU-PHW* nuclear power stations. Calculationshave shown that highly concentrated chloride solutions - acidicin the case of sea-water, m-leakage, and alkaline in the restof the cases considered .- would be produced within the steamgenerator. Experiments in a model boiler showed that sea-waterm-leakage caused rapid corrosion of carbon steel componentswhen only AVT** was used for water chemistry control. Use of anon-volatile reagent, as in the congruent phosphate treatment,avoided the rapid corrosion of carbon steel.

On the basis of our studies, congruent phosphatetreatment during sea water in-leakage appears desirable.

+ Paper IWC-77-12 presented at 38th Annual Meeting,International Water Conference, Pittsburgh, Pennsylvania,1-3 November 1977.

* CANada Deuterium Uranium - Pressurized Heavy Water

** AVT - All Volatile Treatment

Atomic Energy of Canada LimitedChalk River Nuclear LaboratoriesChalk River, Ontario KOJ 1J0

January, 1978

AECL-5849

EFFECT OF CONDENSER WATER IN-LEAKAGE

ON STEAM GENERATOR WATER CHEMISTRY

by

P.V. Balakrishnan

Corrosion from the secondary side has contributed to

many boiler tube leaks in nuclear steam generators. The world

experience in steam generator tube failures has been documented

by Stevens-Guille and Hare in a series of reports. There

has been no corrosion-induced tube failure in any steam

generator of a CANDU-nuclear power station.

Corrosion in steam generators is caused by the accumu-

lation there of chemicals which are by themselves corrosive or

which, on concentration, generate a corrosive environment. Such

chemicals may be introduced into the steam generator with the

feed water via a leaking condenser which allows the raw

condenser cooling water to mix with the condensate. Undesir-

able chemicals may enter the steam generator also with the

make-up water because of improper operation of the make-up

water treatment plant. The phosphates used for water treatment

in some PWR* steam generators have also contributed to tube

failures.

Printed by permission of Conference Organizers

* Pressurized Water Reactor

CONCENTRATION OF NON-VOLATILE SOLUTES IN THE STEAM GENERATOR

Steam generators, by nature, are concentrators of non-

volatile solutes. Thus, a continuous injection of non-volatile

matter with the feed water into the steam generators causes the

solute to accumulate. If some of the water is continuously

blown down from the steam generator, the concentration which

the solute can attain in the bulk of the water is limited. The

steady state concentration factor will be the ratio of the feed

rate to the blowdown rate.

In addition to the above bulk concentration process,

there may be a further concentration within low flow areas in

the steam generators. The crevices within the tube-to-

tubesheet joints and those between the tube support plates and

tubes, sludge accumulated on the tube sheet, and porous

deposits on the tubes, are examples of such areas. Solute

concentration within porous deposits has been studied in :i(23) •}

detail. Models ' have been suggested for the

concentration process within deposits, all based on wick

boiling.(4)

Concentration of solutes in the water should lead to

changes in the composition of the solution, as the equilibria

(hydrolytic, ionization and solubility) in solution cause the

formation of different ionic species and precipitates. The

change in the composition of the solution may very well

generate an aggressive environment and cause corrosion problems

in the steam generator.

In the following, the effect of ingress of condenser

cooling water on the chemistry in CANDU-PHWR* steam generators

is discussed.

PHWR - Pressurized Heavy Water Reactor

REACTIONS IN STEAM GENERATOR WATER BETWEEN SOLUTE SPECIES

The analyses for major species in the cooling waters

now being used or soon to be used for the condensers in

CANDU-PHWR steam supply systems are given in Table 1.

The increase in concentration of the ionic species due

to evaporation in the steam generator results in the

precipitation of compounds whose solubility products have been

exceeded. This alteis the relative composition of solutes in

the water and the water may become acidic or alkaline, depend-

ing on the nature of the compounds that have precipitated. The

compounds that may precipitate at somewhat low concentrations

are :

calcium and magnesium carbonates

calcium and magnesium hydroxides

calcium sulphate and magnesium hydroxysulphate

Within crevices, very high concentrations are possible and

salts of higher solubility (e.g. sodium chloride, sodium

sulphate, etc.) may precipitate.

The following reactions are considered in this work:

HCO 3 -> OH + CO 2t . . . (1)

H 20 = H + + 0H~; K W = [H+][OH~] ... (2)

HSO~ = H + + S0~2; K H S Q = [H+][SO~2]/[HSO~] ... (3)

CaSO4(s) = C a+ 2 + SO~ 2; K C a S Q = [Ca + 2][SO~ 2] ... (4)

Ca(OH)2(s) = C a+ 2 + 20H"; K C a ( O H ) = [Ca + 2][OH~] 2 ... (5)

Mg(OH)+ = Mg + 2 + 0H~; K M g Q H = [Mg+2][OH~]/[Mg(OH)+] ... (6)

Mg(OH)2(s) = Mg+ 2 + 2OH~; K M g ( Q H ) = [Mg+2][OH~]2 ... (7)

MgSO4(s) = Mg+ 2 +"'so~2; K M g S Q = [Mg+2][so~2] ... (8)

Na2SO4(s) = 2 Na+ + SO~2; K^ S Q = [Na+]2[SO~2] ... (9)

2 4

NaCl(s) = Na+ + Cl"; K N a C 1 = [Na+][Cl~] ... (10)

CaCl2(s) = Ca+ 2 + 2 Cl"; K C a C 1 = [Ca+2][Cl"]2 ... (11)

MgCl2(s) - Mg+ 2 + 2 Cl"f K M g C 1 = [Mg+2][cl"]2 ... (12)

The above scheme of reactions is, perhaps, too simple. It

needs to be verified, since we do not really know that the

precipitates that would be formed in a steam generator have all

been included. No double or basic salts are considered in this

scheme because of a lack of thermodynamic data. The bicarbonate

ion is assumed to dissociate completely, the resulting carbon

dioxide being carried away from the system by steam. Because of

these simplifications, the results from this work may be taken

as only approximately defining the chemistry conditions in the

steam generator. Moreover, this scheme considers leakage of

salts iiito steam generator water that is untreated. The

results, however, should apply to water treated with a volatile

amine (e.g. ammonia, cyclohexylamine or morpholine) according to

the All Volatile Treatment (AVT) because the reagent usually

distributes preferentially to steam and because it is a weak

base and almost completely un-ionized at the operating

temperature of steam generators.

The equilibrium constants for the reactions at the

temperatures considered here are listed in Table 2. The terms

within the square brackets on the right-hand side of the

equations for equilibrium constants are the activities of the

corresponding ions. The activities are obtained by multiplying

the concentration in mol/kg by the activity coefficient. Use

of concentrations instead of activities in the expression for

equilibrium constants yields the concentration quotients, Q,

which are related to the thermodynamic equilibrium constants,

K, by the equation

K = Q.r ... (13)

The activity coefficient quotients, V, are described in terms

of the individual mean ionic activity coefficients, y, by

expressions analogous to those describing the equilibrium

constants (e.g. rNa;;SOi< = Y Na+#YSO'2) a n d a r e o b t a i n e d

from the modified Debye-Huckel expression of the form

S is a constant, for a given temperature; (Az ), the difference

in the sum of squares of the ionic charges betv;een the right-

hand side and the left-hand side of the equation representing

the chemical reaction; A, a constant at a given temperature for

a given equilibrium; and y the ionic strength of the medium.

The ionic strength, \i , is related to the molal ionic

concentration, m., and ionic charges, Z., by the relation

2u = h 2 m±z± ... (15)

i

The parameters for evaluation of activity coefficient quotients

are also shown in Table 2.

CALCULATIONS

The calculations consist of two parts. In the first

part the impurities entering the steam generator at a certain

rate from condenser in-leakage are considered to concentrate

there because water is removed as steam. The change in

concentration of a non-volatile solute is derived from the mass

balance equation

d CBv = F.C - B.C ... (16)dt F B

v = volume of the steam generator

C = concentration in the steam generatorB

C = concentration in the feed water, as a result of

condenser in-leakage

F = rate at which feed water enters the steam generator

B = rate at which the steam generator water is blown down

Equation (16), on integration, gives

CB = [1 - exp( - | . t ) ] . . . (17)

for the concentration at time, t. At steady state, when the

rate of input of impurities into the steam generator is equal

to the r a t e of removal by blowdown,

CB = F.CF/B . . . (18)

Calculations are done for the period from the start of the

condenser leak to steady state.

In the second part, the bulk water in the steam

generator is considered to concentrate within crevices, by

arbitrary factors until either equilibrium with respect to all

solute species is attained or the elevation of boiling point of

the solution because of increasing solute concentration

prevents the solution from boiling at the pressure in the

secondary system at the highest temperature attainable in the

steam, generator - the primary coolant temperature. The

elevation of boiling point, AT, is calculated using the

equation

AT = - £L_ in(l - x ) ... (19)AHv

where R = gas constant

T - temperature

AH = latent heat of vaporization of water at the

given pressure

total mol

solution

X = total mole fraction of all solutes in the

In either part the compounds whose solubility product:;

are exceeded are allowed to precipitate. The overall

concentration of the solute species in the water is obtained by

multiplying the feed concentrations by the appropriate

concentration factor. The net concentrations used in equations

(3) to (12) are obtained from the overall concentrations by

allowing for reduction in concentration due to precipitation of

the appropriate compounds whose solubility products are

exceeded. The values for the reduction in concentration due to

the formation of the various precipitates are at this stage

unknown and are solutions to the non-linear equations

represented by equilibria (2) to (12). An equation imposing

electroneutrality condition is added to these equations and

this set of equations is solved in the computer. Starting from

an approximate value for ionic strength, the calculations

refine the ionic strength value by iteration until convergence

is achieved. Calculations for the bulk concentration neglect-

any effect of precipitation within crevices.

CONCENTRATION OF SOLUTES IN STEAM GENERATOR WATER

Bulk Water

Three cases are considered.

1) Bruce Nuclear Generating Station, Ontario

Bruce NGS consists of four CANDU-PHW reactors, each of

nominal 750 MW(e )* capacity. Each reactor has eight steam

generators. Lake Huron water is used for condenser cooling. A

condenser leak rate per steam generator of 2,500 kg/h resulting

in 12.8 ;.g Na/kg in the feed wattr is considered.

2) Gentilly-2 Nuclear Generating Station, Quebec

Gentilly-2 NGS has a 600 MW(e) CANDU-PHW reactor with

four steam generators. St. Lawrence River water will be used

for condenser cooling. A condenser leak rate per steam

generator of 1,000 kg/h, which leads to 11.6 ug Na/kg in the

feed water, is treated.

3) Point Lepreau Nuclear Generating Station, New Brunswick

Point Lepreau NGS is similar to Gentilly-2, except

that it will use sea water for condenser cooling. A condenser

leak rate per steam generator of 5 kg/h giving rise to 63 yg

Na/kg in the feed water is considered.

The above leak rates are based on what is easily

detected by a sodium ion specific electrode and is possible to

locate.

The results of the calculations are shewn in Figures 1

to 4. Figures 1 to 3 show the buildup of Ca, Mg and Na

respectively, in the steam generator. The points at which a

compound starts precipitating is marked in the figures by a

vertical line with the appropriate label. Figure 4 shows the

change in pH for the three stations. The fresh waters

(Lake Huron and St. Lawrence River) are seen to be alkali-

forming and the sea water, acid-forming.

Table 3 gives the concentrations at steady state.

* electrical

Cirevices

Here the five cooling waters (Table 1) are considered

concentrated from their natural solute concentrations, rather

than from steady state concentrations in a steam generator.

Figures 5 to 8 show the buildup of the concentration of the

various ions as a function of concentration factor at a primary

temperature of 300°C. The appearance of a specific precipitate

is indicated at the corresponding concentration factor.

The concentration of each ion at the maximum

attainable concentration factor approaches the same value for

the fresh waters (except for Ottawa River water for which the

calculation could not be done up to the maximum concentration

factor because of some computational problems). Sea water

attains saturation with respect to each of the solute species

at 300°C.

Figure 9 shows the change in pH o as a 'function of

concentration factor in crevices. The sea water turns acid and

the fresh waters alkaline. Among the fresh waters, Lake Huron,

St. Lawrence River and Lake Ontario waters behave similarly

with respect to change in pH, with alkalinity decreasing in

that order. The Ottawa River water shows much higher

alkalinity than the other fresh waters at the same

concentration factor and the pHo^o shows an increasing trend

up to the highest concentration factor (10 ) considered.

Alkalinity results from the decomposition of the bicarbonate

ion (reaction 1)

HCO- •* CO2+ + OH~ ... (1)

Acidity results from the hydrolytic reactions (5a) to (7a).

Ca+2 + 2H2O + Ca(OH)2 + 2H+ ... (5a)

Mq+2 + H2O -> Mg(OH)+ + H+ ... (6a)

Mg + 2 + 2H2O - Mg(OH)2 + 2H+ ... (7a)

10

The calciu..i and magnesium concentrations in Ottawa River water

are much lower than those in the remaining fresh waters (Table

1). The Ottawa River water should, therefore, be much less

acidic, i.e. more alkaline than the other fresh waters.

Table 4 gives the maximum attainable concentrations

within crevices at a temperature of 300°C and Table 5 those at

275°C. The fresh waters all lead to alkaline solutions of high

chloride concentration, sodium and calcium chlorides being the

principal solutes. Sea water leads to an acid solution of high

chloride concentration, with sodium and magnesium chlorides as

the principal solutes. The concentrations at 275°C are lower

than those at 300°C.

POSSIBLE EFFECTS OF THE CONCENTRATION OF SOLUTES IN

STEAM GENERATORS

If the ionic species from the condenser cooling water

can concentrate to the maximum possible concentration within

crevices in the steam generator, corrosion may occur at an

accelerated rate at these sites. The types of corrosive attack

which have occurred in steam generators are caustic-induced

stress corrosion cracking of the tubes (Inconel-600 or

Incoloy-800), pitting and general metal wastage, also of the

tubes, usually caused by acid conditions, and corrosion of

other parts, e.g. the carbon steel tube support plates in the

steam generator.

The fresh waters all lead to alkaline solutions when

concentrated, but no cracking of Fe-Ni-Cr alloy tubes has been

observed in laboratory tests at the low concentrations of

alkali that are calculated to result from condenser leakage.

Thus, based on the current information on caustic cracking,

condenser in-leakage does not pose a threat, except perhaps

with Ottawa River water.

The acid chloride solution produced from sea water

could corrode carbon steel components, particularly the tube

11

support plates which form crevices around the tubes. Potter

and Mann have shown that carbon steel corrodes at a fast

and linear rate in 0.1 mol/kg FeCl solution at 300°C. A

similar situation is possible with concentrated sea water. The

phenomenon of tube denting* seen in some PWR steam generators

has been attributed to the rapid corrosion of the tube support

plates within crevices. Denting has been seen mostly with sea-

water cooled PWR's and it has been more severe in the hot leg

than in the cold leg. The calculations reported here do show

that the concentration of solutes at the lower temperature of

275°C is less than that at 300°C, which suggests a lower

corrosion rate in the cold leg than in the hot leg, as

indicated by the relative severity of denting at these

locations.

MODEL BOILER EXPERIMENTS

Experiments to study the effect of sea-water

in-leakage on corrosion were carried out in a model boiler.

The boiler had a single U-tube of Incoloy-800, the preferred

alloy for CANDU steam generators. A concentrating device (a

carbon steel umbrella), shown in Figure 10, was appended to one

leg of the tube to provide a crevice around the tube where

steam blanketing and concentration of chemicals could occur.

The heat required for the operation of the boiler was derived

from a high-temperature high-pressure loop. The water from the

loop flowed through the tube and boiled the secondary water in

the shell. The steam produced was condensed in a condenser

tubed with copper. The flow in the secondary system was

* Denting is the reduction in the diameter of the steamgenerator tubes within tube support plates.

12

maintained by natural convection (density differences). The

loop and the model boiler are shown schematically in Figure 11.

The operating conditions of the system are given in Table 6.

The sea water was injected into the boiler with the

make-up water. When phosphate treatment was used for the

boiler water the phosphates were pumped into the boiler. A

steady rate of blowdown was maintained. Analyses of samples of

the blowdown liquid, the make-up water and the feed water were

done periodically for pH, conductivity, chloride, phosphate,

sodium, calcium and magnesium.

Congruent Phosphate Treatment (CPT) and Sea-Water Injection !

During one phase of the experiments, the boiler water £

was treated according to the congruent phosphate treatment !\i

method, while injecting sea water into the boiler to simulate ;',

condenser leakage. The chemistry conditions were as follows: ?

pH =9.8

phosphate = 18 mg PO./kg water (average)

chloride (from sea water) = 17 to 55 mg Cl/kg (32 mg/kg average)

oxygen = 10 pg/kg

morpholine ~ 6 mg/kg

The observed pH of 9.8 at the morpholine concentration of

~6 mg/kg corresponds to a sodium-to-phosphate molar concen-

tration ratio of «2.4:1. (The desired range for this ratio is

2.2 to 2.6:1.) After 47 days, the tube was covered with a

thick black deposit composed of corrosion products (mainly

magnetite) and the products from sea-water salts (calcium

phosphate, magnesium hydroxide). No localized corrosion was

observed on the tube, even under crevices. The carbon steel

13

umbrella used as the concentrating device did not show any

localized or general corrosion.

All Volatile Treatment and Sea-Water Injection

During this phase the water was treated with

morpholine (a volatile amine), while sea-water injection was

maintained. This phase was started with a clean tube, i.e. no

prior deposit. The boiler chemistry conditions were:

pH = 9.5 to 9.6 (with morpholine)

chloride = 8 to 60 mg Cl/kg water (average 30 mg/kg)+ 2

Mg = 1.5 to 2.3 mg/kg as Mg+ 2

Ca = 1.4 to 4.3 mg/kg as Ca

oxygen = 5 to 10

The pH tended to drift towards lower values and it required

more morpholine than normally required to maintain the pH in )

the same range in the absence of sea water. About 12 mg/kg was

normally sufficient in the absence of sea water, but up to 25

mg/kg was required with sea-water injection. This behaviour

illustrates the acid-forming nature of sea water.

After 47 days the deposit formed in this phase was

thinner than that formed with CPT, but its chemical composition

was similar.

There was no evidence of localized attack on the tube,

but the carbon steel umbrella was severely corroded (Figure 1 2 ) .

The original wall thickness of 0.57 mm at the neck of the

umbrella was completely penetrated in 47 days. This allows an

estimate for the corrosion rate of greater than 900 mg/(dm'.day)

This high corrosion rate must have been due to the very high

chloride concentration and the acid pH that would have existed

in the crevice within the umbrella. This contrasts with the

lack of significant corrosion when congruent phosphate

14

treatment was used to counter the effects of sea water. The

non-volatile phosphate buffer system prevented the pH from

going to the acid side. The volatile amine, morpholine, which

is also only a weak base and hence mostly un-ionized at the

boiler temperature, was inadequate for this purpose.

Interpretation

The experiments described demonstrated that sea-water

in-leakage into boiler water treated with volatile amines would

result in accelerated corrosion of carbon steel components. If

the pH was maintained alkaline with a non-volatile buffering

reagent during the sea-water in-leakage, the rapid corrosion of

carbon steel and the related problems would be suppressed. The

buffering reagent may be sodium phosphates as in the congruent

phosphate treatment, borate or another suitable system.

SUMMARY AND CONCLUSIONS

Calculations show that concentration of the fresh

waters that are being used now or soon to be used for cooling

condensers at CANDU-PHWR stations lead to alkaline solutions of

high chloride concentration. Sea water, on concentration,

leads to highly concentrated chloride solutions of acid pH.

Model boiler experiments have shown that sea-water

in-leakage countered only by volatile treatment of the boiler

water would lead to rapid corrosion of carbon steel.

Accelerated corrosion of carbon steel has been the cause of

denting of tubes in some steam generators. A non-volatile

buffering reagent, such as the sodium phosphates, controls the

chemistry so that the corrosion of carbon steel is suppressed.(14 1

The current recommendation ' for secondary water

treatment for CANDU-steam generators at sea-water cooled

stations is to use AVT during normal conditions, but to switch

to a low concentration (2 to 10 mg PO /kg water) congruent

15

phosphate treatment with the sodium to phosphate molar concen-

tration ratio in the range (2.2:1) to (2.6:1) on detecting a

condenser leak. The phosphate treatment is to be maintained

until the leak is located and corrected, when the treatment is

to be returned to AVT. The findings from the work reported in

this paper, i . e . the inadequacy of AVT during sea-water

in-leakage for the protection of carbon steel components in the

steam generator and the desirabil i ty of using a non-volatile

buffering system for this purpose, support this water treatment

approach.

ACKNOWLEDGEMENT

The contribution of D.P. Martin in developing the

computer program used in this work has been very valuable.

16

REFERENCES

1. a) Stevens-Guille, P.D., "Steam Generator Tube Failures: h

World Survey of Water-Cooled Nuclear Power Reactors to

the End of 1971", AECL-4449, Atomic Energy of Canada

Limited (1973).

b) Stevens-Guille, P.D., "Steam Generator Tube Failures:

World Experience in Water-Cooled Nuclear Power Reactors

During 1972", AECL-4753, Atomic Energy of Canada Limited

(1974).

c) Stevens-Guille, P.D. and Hare, M.G., "Steam Generator

Tube Failures: World Experience in Water-Cooled Nuclear

Power Reactors in 1973", AKCL-5013, Atomic Energy of

Canada Limited (1975).

d) Hare, M.G. , "Steam Generator Tube Failures: World

Experience in Water-Cooled Nuclear Power Reactors in

1974", AECL-5242, Atomic Energy of Canada Limited (1975).

e) Hare, M.G., "Steam Generator Tube Failures: World

Experience in Water-Cooled Nuclear Power Reactors in

1975", AECL-5625, Atomic Energy of Canada Limited (1976).

2. Collier, J.G. and Kennedy, T.D.A., "Solute Concentration

in Highly Rated High Pressure Steam Generators.

I. A Model for the Concentration of Solute Within the

Pores of a Magnetite Deposit", AERE-R 7203, United

Kingdom Atomic Energy Authority, Harwell, UK (1972).

3. Cohen, P., "Heat and Mass Transfer in Porous Deposits

with Boiling", WARD-5836, Westinghouse Electric Corpor-

ation, Madison, Pa., USA (1972).

17

4. Macbeth, R.V., "Boiling on Surfaces Overlayed with a

Porous Deposit: Heat Transfer Rates Obtainable by

Capillary Action", AEEW-R 711, United Kingdom Atomic

Energy Authority, Winfrith, UK (1971).

5. Fisher, J.R. and Barnes, H.L., "The Ion-Product Constant

of Water to 350°C", J. Phys. Chem. 76, 90 (1972).

6. Marshall, W.L. and Jones, E.V., "Second Dissociation

Constant of Sulfuric Acid from 25° to 350° Evaluated

from Solubilities of Calcium Sulfate in Sulfuric Acid

Solutions", J. Phys. Chem. JO, 4028 (1966).

7. Helgeson, U.C., "Thermodynamics of Hydrothermal Systems

at Elevated Temperatures and Pressures", Amer. J. Science

267, 729 (1969) .

8. Marshall, W.L., "Thermodynamic Functions at Saturation of

Several Metal Sulfates in Aqueous Sulfuric and Deutero-

sulfuric Acids at Temperatures up to 350°C", J. Inorg.

Nucl. Chem. 37' 2 1 5 5 (1975).

9. Yeatts, L.B. and Marshall, W.L., "Aqueous Systems at High

Temperatures XVIII. Activity Coefficient Behaviour of

Calcium Hydroxide in Aqueous Sodium Nitrate to the

Critical Temperature of Water", J. Phys. Chem. T^, 2641

(1967).

10. Mellor, J.W., "A Comprehensive Treatise of Inorganic and

Theoretical Chemistry", Vol. Ill, Longman, Green & Co.,

New York, p 702 (1946).

\

18

11. Cohen, P., "The Chemistry of Water and Solutions at High

Temperatures for Application to Corrosion in Power

Systems", WARD-5788, Westinghouse Electric Corporation,

Advanced Reactors Division, Madison, Pa., USA, p 38

19 72) .

12. Liu, C-T and Lindsay, W.T., Jr., "Thermodynamics of

Sodium Chloride Solutions at High Temperatures", J.

Solution Chem. 1, 45 (1972).

13. Potter, E.C. and Mann, G.M.W., "The Fast Linear Growth

of Magnetite on Mild Steel in High-Temperature Aqueous

Conditions", British Corrosion Journal, 1, 26 (1965).

14. Hart, A.E. and LeSurf, J.E., "CANDU Steam Generator

Tubing Material - Service Experience and Allied Develop-

ment", AECL-5384, Atomic Energy of Canada Limited (1976).

19

TABLE 1

RAW WATER QUALITY DATA

Calcium

Magnesium

Sodium

Potassium

Sulphate

Chloride

Bicarbonate

OttawaRiver '

6.5

2.5

4

3.2

7

2.4

ConcentrationLakeHuron2

25-28

7-9

3-4

0.8-1.3

12-15

5-8

95-110

LakeOntario3

38-41

7.5-8

10-14

1-1.5

28

28-30

120

(mg/kg)St. Lawrence

River **

27

8

10

2

26

21

80

AtlanticOcean 5

400

1,272

10,556

380

2,649

18,980

140

1 Used at Nuclear Power Demonstration (NPD) Generating Station

2 Used at Bruce Generating Station and Douglas Point Generating Station

3 Used at Pickering Generating Station

11 To be used at Gentilly-2 Generating Station

5 To be used at Point Lepreau Generating Station

TABLE 2

EQUILIBRIUM CONSTANTS AND SOLUBILITY PRODUCTS

Reaction

H 20 = H + 0H~

HSO4 = H + S0 4

Mg(OH) = Mg +

CaSO4(s) = Ca *

Ca(OH)2(s) = Ca

CaCl2(s) = Ca

CH

Mg(OH)2(s) = Mg

+ 2MgSO (s) = Mg

4MgCl (s) = Mg + 2

+ 2

2 OH

2 OH

2 Cl"

(s) 2 Na + SO" 2

NaCl(s) = N a + + Cl"

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

(Ref.

5)

6)

7)

8)

9)

10)

11)

8)

10)

8)

12)

8

1

1

9

4

1

250(S = 0.K

.912xlO~

.05xl0~5

.OOxlo"

.311xlO~9

-9.270x10

.07x10

°C9848)

A

1

1

2

1

2

2

.45

.77

.14

.60

.14

.14

9.

4.

4.

2.

1.

275°(S = 1.

K

772xlO~

5SxlO~6

-650x10

290xl0~9

352x10

127:

3.

3.

.40x10

,462x10

11.72

9.

4,

.521xlO"6

.425

C112

1.

1.

2.

1.

2.

2.

2.

1,

2.

0,

1

A

33

77

15

60

15

15

,15

,60

.15

.62

.33

9.

1.

2.

5.

4.

300(S = 1K

120xlO~

94xl0~6

00xl0~

-10636x10

130xl0~i0

45'

1.

4.

4

4

2

.04X10"1

-8.634x10

.2"

.550xl0~6

.248

°C.287

1.

1.

2.

1.

2.

2.

2.

1.

2.

0,

1

A

20

73

16

60

1 6

o16

,16

,60

.16

.72

.20

1 Estimated from the solubility of calcium chloride at 260°C, given in Reference 10.

2 Estimated from the solubility of magnesium chloride at 181°C, given in Reference 10.

TABLE 3

STEADY STATE CONCENTRATIONS IN STEAM GENERATOR BULK WATER

Blowdown Rate = 0.1% of Steaming RateTemperature = 250°C

Condenser C o n c e n t r a t i o n (mmol/kg) EquivalentWater pH o

(leakage rate) Ca Mg Na K S O . C l 5 c

Sea water 0.059 0.311 2.73 0.058 0.164 3.184 5.42(5 kg/h)

St . Lawrence 0.510 0.014 0.507 C.060 0.040 0.691 10.92River(1000 kg/h)

Lake Huron 1.419 0.048 0.644 0.087 0.021 0.896 11.42(2 500 kg/h)

22

TABLE 4

MAXIMUM POSSIBLE CONCENTRATION IN CREVICE

Temperature - Primary: 300°C- Secondary: 250°C

Calcium

Magnesium

Sodium

Sulphate

Chloride

P H25°C

Fresh Water*(mol/kg)

14.36

1.72xlO"3

5.42

3,17xlO"5

34.15

10.41

Sea Water(mol/kg)

0.134

11.10

6.17

3.90xl0~3

28.56

3. 52

* Lakes Huron and Ontario, St. Lawrence River

23

TABLE 5

MAXIMUM POSSIBLE CONCENTRATION IN CREVICE

Temperature - Primary: 275°C- Secondary: 250°C

Fresh Water(mol/kg)

Sea Water(mol/kg)

Calcium

Magnesium

Lake Huron Lake Ontario

1.45

3.47x10-4

1.96

4.21x10-4

9.35x10

1.44

-3

Sodium

Sulphate

9.04

8.84x10,-5

8.28

6.86x10-5

9.06

1.45x10"

Chloride 11.95 12.20 11.91

pH 10.86 10.80 4.42

24

TABLE 6

OPERATING CONDITIONS OF THE MODEL BOILER SYSTEM

Primary Circuit

Temperature - boiler inlet: 3Q0°C

Pressure:

Flow rate:

PH:

Oxygen:

Purification flow:

Make-up water:

9.0 MPa

1.1 to 1.5 kg/s

10 to 10.5 {with LiOH)

<5 ug/kg {controlled bycatalyzed hydrazine)

«10% of the loop flow

distilled, deionized,deoxygenated water

Secondary Circuit

Temperature - steam:

- condensate:

- feed water:

Steaming rate:

Boiler pressure:

Blow down:

251°C

95°C

170°C

0.063 kg/s

4.1 MPa

=1% steaming rate

25

I - 2

•et

o

BUILDUP OF CALCIUM IN WATERTEMPERATURE: 250°CBLOWDOWN: O.I* OF STEAMING RATE

BRUCE - LAKE HURON ( 2500 ,kg/h )

G-2 - ST. LAWRENCE RIVER (1000 kg/h)

PT. LEPREAU - SEA (5 kg/h)

SOLID PRESENT - A: CaS04

B: Ca(0H)2

2 0 3 0T I M E ( H O U R S )

F I G 1 B U I L D U P O F C A L C I U M C O N C E N T R A T I O N I N S T E A MG E N E R A T O R W A T E R(CONDENSER LEAK RATES W I T H I N PARENTHESES)

26

- 3

où

oE

ai

oC_3

BUILDUP OF MAGNESIUM IN WATERTEMPERATURE: 250°CBLOWDOWN: 0. \% OF STEAMING RATE

PT. LEPREAU - SEA (5 kg/h)

G2 - ST. LAWRENCE RIVER ( 1000 kg/h)

10 - 6

BRUCE - LAKE HURON (2500 kg/h)

D - SOLID Mg(OH)2 PRESENT

_J I I I

0 0 502 0 3 0 4 0T I M E ( H O U R S )

F I G . . 2 B U I L D U P O F M A G N E S I U M C O N C E N T R A T I O N I N S T E A MG E N E R A T O R W A T E R(CONDENSER LEAK RATES W I T H I N PARENTHESES)

27

- 2

oc

UJ

oo

FIG.

BUILDUP OF SODIUM IN WATERTEMPERATURE: 250°CBLOWDOWN: 0. \% OF STEAMING RATE

I40 502 0 3 0

T I M E ( H O U R S )

3 B U I L D U P O F S O D I U M C O N C E N T R A T I O N I N S T E A MG E N E R A T O R W A T E R(CONDENSER LEAK RATES WITHIN PARENTHESES)

28

1 2. 0

0. 0

8 . 0 -

6 . 0 -

4. 0

2. 0

j/\^ BRUCE - LAKE

\. G2 - ST. LAWRENCE

PT. LEPREAU - i

TEMPERATURE: 250°BL0WD0WN: 0. \ai

1 i

.

HURON

RI

SEA

C; OF

VER

(5

( 2500

( !000

kg/h)

STEAMING

i

kg/h)

kg/h)

RATE

i

0 0 2 0 3 0 4 0 5 0

T I M E ( H O U R S )

F I G . . 4 C H A N G E I N p H ^ I N S T E A M G E N E R A T O R W A T E R

(CONDENSER LEAK RATES W I T H I N PARENTHESES)

29

ozo

- 1

-3

BUILDUP OF CALCIUM CONCENTRATIONTEMPERATURE - PRIMARY: 300'

- SECONDARY:

SOLID PRESENT

A: CaS04

B: Ca(OH),

-5

-6

ION

ST. LAWRENCE RIVER

LAKE HURON

OTTAWA RIVER

I I I0 1 2 3 4 5 6

log (CONCENTRATION FACTOR)

FIG.. 5 BUILDUP OF CALCIUM CONCENTRATION WITHIN CREVICES

3 0

2

0aa

JE

\

Ê -1

ION

t—

"* - 2(—

ICE

N

s -3

o

- 4<

I-5

-6

-

y

"IT

BUILDUP OF MAGNESIUM CONCENTRATION

SATURATION

y ^ - SECONDARY: 250°C

y SOLID PRISENT:

^ / D: Mg(0H)2

/ E: MgSO4

F: HgCI2

LAKE ONTARIO

ST. LAWRENCE s ' / r <

^ ^ ^ < ^ ^ a ' ^ k \ L A K E HURON

^^^KOTTAWA RIVER

>Q 01 1 1 1 1 1

0 1 2 3 4 5 6 7

log (CONCENTRATION FACTOR)

FIG.. 6 BUILDUP OF MAGNESIUM CONCENTRATION WITHIN CREVICES

31

ozo

-5

-60

I1

SATURATION(SEA)

LAKE O N T A R I O

ST. L A W R E N C E RIVER

LAKE HURON

OTTAWA RIVER

BUILDUP OF SODIUM CONCENTRATION

TEMPERATURE - PRIMARY: 300°C- SECONDARY: 25O°C

SOLID PRESENT

G: NaCl

I I I 12 3 4 5 6

log (CONCENTRATION FACTOR)

FIG.. 7 BUILDUP OF-SODIUM CONCENTRATION WITHIN CREVICES

32

e t

UJooo

-5 -

-6

SEA

. LAWRENCE RIVER

SATURATION

(SEA)

R I V E R

BUILDUP OF CHLORIDE CONCENTRATIONTEMPERATURE - PRIMARY: 3OO°C

- SECONDARY: 250°CSOLIDS PRESENT:

G: NaCII

F: M g C I

I 2 3 1 5 6l o g ( C O N C E N T R A T I O N FACTOR)

F I G . 8 BUILDUP OF CHLORIDE CONCENTRATION WITHIN CREVICES

vr> Q

OTTAWA RIVER

LAKE HURON

ST. LAWRENCE R IVER

TEMPERATURE - PRIMARY: 300°C- SECONDARY: 250°C

SEA

I I II 2 3 4 5 6

lo g ( C O N C E N T R A T I O N F A C T O R )

F I G . 9 C H A N G E IN p H 2 5 e c W I T H I N C R E V I C E S

34

HOSE CLAMP

UMBRELLA

BOILER TUBE

FIGURE 10 THE CONCENTRATING DEVICE (UMBRELLA)ON BOILER TUBE

MAKEUPTANK

CHEMICAL MAKE-UPDISTILLED WATER

TEST SECTION 9HEATER

CHEMICAL MAKE-UPDISTILLED WATER

CONDENSER

MODEL BOILER

6 KW HEATER

BLOWDOWN

PRIMARY

= SECONDARYAUXILIARY

MAKE-UP PUMP

SEA_WATER

""-S

MAKEUP

TANK

MAKE-UP PUMP

PHOS-PHATETANK

F I G U R E 1 1 S C H E M A T I C D I A G R A M O F M O D E L B O I L E R I N S T A L L A T I O N

36

FIGURE 12 THE CARBON STEEL UMBRELLA AFTER EXPOSURETO SEA WATER AND AVT FOR 47 DAYS.

The International Standard Seria} Number

ISSN 0067-0367

has been assigned to this series of reports.

To identify individual documents in the serieswe have assigned an AECL- number.

Please refer to the AECL- number when

requesting additional copies of this documentfrom

Scientific Document Distribution Office

Atomic Energy of Canada LimitedChalk River, Ontario, Canada

KOJ I JO

Price S3.00 per copy

1446-77