A UNS S32654, A STAINLESS STEEL FOR CONDENSER TUBES

Jan Olsson, Avesta Sheffield AB, S-774 80 Avesta, Sweden

ABSTRACT
The new superaustenitic stainless steel UNS S32654 (Avesta Sheffield 654 SMO) is proposed for condenser tubes in desalination and power plants. A number of tests illustrating its corrosion resistance in a variety of chloride containing environments are being reviewed in terms of pitting, crevice corrosion and stress corrosion cracking. Relevant mechanical and physical properties are described. Health hazards and the impact on the environment are discussed as well as the cost benefits of using a stainless steel for this application.

INTRODUCTION
Stainless steel has frequently, and to a large extent successfully, been used for a number of applications in desalination plants since the early stages of this industry. High pressure piping in seawater reverse osmosis (SWRO) plants is one example, others are tube plates, linings, demisters and destillate troughs in multi stage flash (MSF) plants. Grades used have covered the whole range of conventional austenitic steels such as AISI 304L-316L-317L but also more highly alloyed grades such as UNS N08904 (904L) and S31254 (254 SMO). However, its use for condenser tubes in large desalination units and connected power plants has been limited in spite of a number of successful installations in European and North American power plants, both nuclear and fossil fuel based [1-2]. The reasons have partly been a lack of confidence in the heat transfer due to the low thermal coefficients of stainless steels, but also doubts about their capability to withstand localised corrosion in the hostile environments associated with these plants in countries with a very hot climate, Arabian Gulf countries being one example. This paper describes a recent development within the stainless steel family, UNS S32654 (Avesta Sheffield 654 SMO) which offers far better resistance to localised corrosion, i.e. pitting, crevice corrosion and stress corrosion

B
cracking (SCC) than other stainless steels, including the highly alloyed grades 904L and 254 SMO mentioned above. The chemical compositions of 654 SMO and other grades used as reference material when positioning this grade in terms of corrosion resistance are described, as well as mechanical properties and relevant physical properties, thermal conductivity being one example.

METALLURGY
Chemical composition 654 SMO is a fully austenitic stainless steel with the chemical composition given in Table 1 which also gives the compositions of other stainless steels and some nickel base alloys used as reference material in some of the corrosion tests. Also included are the Pitting Resistance Equivalent (PRE) numbers (PRE=%Cr+3.3*%Mo+16*%N) which give a rough estimate of the relative pitting resistance of the alloys mentioned. 654 SMO contains approximately 3 % manganese which is added to increase the solubility of nitrogen. Mechanical properties The mechanical properties of 654 SMO and some other alloys are given in Table 2. The high yield stress of 654 SMO is combined with a high elongation before rupture which implies good formability, almost on the same level as for a conventional austenitic grade such as 316L. The surface hardness is also high which gives this grade better resistance to certain types of erosion as described below. The high strength and surface hardness are consequences of the high nitrogen content of this grade. Physical properties Some physical properties are of importance when designing components for desalination plants, e.g thermal conductivity for heat transfer surfaces, modulus of elasticity (Young`s modulus) when determining the distance between the support plates in tube and shell heat exchangers, and the coefficient of linear expansion for equipment made of different materials, working at elevated temperatures. Some relevant physical properties are listed in Table 3.

C
The influence of the listed properties on the design and service performance of desalination and power plant components will to some extent be discussed below.

CORROSION RESISTANCE
Natural Seawater Natural seawater contains living organisms which form a biofilm on inert metal surfaces like passive stainless steel. This biofilm has a catalytic effect on the reduction of oxygen on the metal surface, resulting in more noble corrosion potentials than in sterile chloride solutions, close to 300 and just above 100 mV SCE respectively [3]. The consequence is a higher corrosivity of natural seawater. However, the activity of the biofilm is reduced if the temperature of the water is increased to temperatures above ambient and the maximum corrosivity of natural seawater is therefore achieved in the temperature range 25-35C. The resistance of some stainless steels and nickel base alloys to natural seawater at 25C was reported by LCCT seawater laboratories in 1993 [4]. Unwelded samples with crevice formers made of acrylic resin crevice washers, torqued to 8.5 Nm, were exposed for 60 days. Different surface conditions were investigated. Some results are summarized in Table 4. As expected, 254 SMO and C-276 performed much better than 316, but the only alloy fully resistant to all test conditions was 654 SMO. Chlorinated Seawater Unlike natural seawater, continuously chlorinated seawater is a sterile solution without any living organisms and conditions can easily be simulated by tests in artificial seawater or sodium chloride solutions. However, the samples must be polarized to a corrosion potential corresponding to that given by chlorination, normally in the range of 500 to 600 mV SCE [3]. The resistance to crevice corrosion has been determined accordingly by measuring the critical crevice corrosion temperatures, CCT, in a test developed to simulate conditions in plate heat exchangers [5]. Multiple crevice washers made of nitrile rubber, the same type of gasket material as that used in plate heat exchangers, are bolted to the samples with a torque giving a compression of 75 % of their original thickness. The specimens are then immersed in a sodium chloride solution, pH 8, and polarized to

D
500 mV SCE to simulate continuous chlorination. The test duration is 30 days at each temperature and the critical temperature is defined as the lowest temperature giving crevice corrosion. Results from tests in a solution with a chloride ion level of 25,000 ppm are reported in Table 5. These results are in line with failures reported for 254 SMO when used for plate heat exchangers exposed to continuously chlorinated seawater at 30-35C and they also indicate that 654 SMO can be considered as an alternative to Alloy C-276 and titanium, which both have a history of success for this application. In addition to these tests, simulating conditions in chlorinated seawater, a field test in real seawater has been performed at the seawater laboratory of the French Atomic Energy Institute (CEA) [6]. Welded samples were equipped with washers made of rubber-bound aramid fibre, bolted with a torque of 40 Nm to the samples. The samples were immersed in chlorinated seawater that had been heated to a temperature of 45C. The content of residual chlorine was 10 ppm and the test lasted for a period of 95 days. The results are summarized in Table 6. The test conditions were very severe with a chlorine level much higher than that which is normal in seawater systems. Only 654 SMO was resistant, thereby confirming the results from the laboratory tests. Deaerated Seawater The corrosivity of seawater is, as described above, determined by the presence of oxidizing agents, e.g. oxygen and chlorine, since these will raise the corrosion potential of the steel to a level where pitting or crevice corrosion might be induced. Deaeration of seawater and brine has a reverse effect, resulting in corrosion potentials well below -100 mV SCE, depending on temperature and level of oxygen. A 6Mo grade like 254 SMO will resist pitting and crevice corrosion at an oxygen level of at least 50 ppb at 120C according to Rogne and Johnsen [7]. 654 SMO should be even more resistant though no such results have been reported. Stress Corrosion Cracking Stress corrosion cracking (SCC) has often been called the Achilles` heel of stainless steels since its appearance is often unexpected and it causes severe damage of the component exposed to it. The most cost effective way of solving this problem is to use a ferritic-austenitic (duplex) stainless steel. However,

E
sometimes there is also a demand for better resistance to pitting and crevice corrosion and a highly alloyed austenitic grade could be the alternative. One way to simulate the most common reason for SCC, i.e. the evaporation of chloride containing water on a hot surface, is to perform a test according to the drop evaporation test (DET) method. In this test a dilute sodium chloride solution is dripped on a uniaxially stressed and heated sample. The SCC resistance is then measured as the threshold stress, giving SCC within 500 hours [8]. Threshold stresses for some stainless steels are given in Table 7. DET is a highly accelerated test and does not predict the service life in a practical situation. It should, however, give a realistic ranking order with regard to SCC. Judging from the results presented, 654 SMO will perform very well under severe evaporative conditions, e.g. as a construction material for the first stages of MSF chambers with an external wall temperature of just above 100C.

EROSION RESISTANCE
The erosion resistance of stainless steel is normally very good when exposed to high velocity seawater unless this contains solid particles which can cause mechanical abrasion. Also titanium is very resistant to high velocity seawater. However, it has been reported that condenser tubes made of titanium have suffered external erosion in Finnish and Swedish nuclear power plants [9]. This has been caused by high velocity steam, containing small condensate droplets, hitting the outer layers of titanium tubes with a speed of close to 200 m/s. Tests, simulating such conditions, have been performed on behalf of the Swedish State Power Board in order to establish suitable alternative materials. The first test showed that 254 SMO was superior to titanium and in 1991 approximately 50,000 metres of 254 SMO tubes were installed as a replacement for titanium at Oskarshamn II, a 600 MW plant on the Baltic Sea coast. However, the test also showed that the expected service life of the 254 SMO tubes could be inadequate for the remaining service time of other reactors and a second series of tests was performed including also 654 SMO. Specimens of different materials were attached to a rotating beam and the specimens hit a stream of droplets with a well defined size. The relative velocity of droplets is simulated by adjusting the rotation speed of the beam. The results from tests with a droplet size of 0.6 mm at two different velocities are shown in Table 8 [10].

F
This second series of tests showed that 654 SMO almost fulfilled the demand of approximately five times longer service life than the titanium grade used today, probably due to its higher surface hardness. Approximately 150,000 m of 654 SMO tubes have been ordered for the replacement of titanium tubes in Forsmark I and II (Baltic Sea, 970 MW each) and Ringhals IV (915 MW, North Sea) during 1995, starting with F I in early June.

HEAT TRANSFER
One important aspect of materials used for surface condensers and other heat exchangers is their heat transfer capability. Copper base alloys, both brass and copper nickel, have high coefficients of thermal conductivity while some other materials, e.g. titanium and stainless steel, are inferior in this respect. However, a comprehensive review of data on heat transfer was presented by Maurer in 1980 showing that other factors were far more important than the coefficient of thermal conductivity , water side films and water side fouling being two examples [11]. These findings have also been confirmed by measurements in Swedish nuclear power plants, see Table 9 [12]. The total heat transfer was increased by 17 % when using titanium tubes and soft ball cleaning instead of Al-brass despite the more than 80 % lower coefficient of heat transfer (17 and 100 W/mC respectively). Similar results, though even more favourable, were reported for another Swedish nuclear power station, Ringhals, where an improvement of 45 % was recorded. The difference was not discussed in the report but the Oskarshamn plant is located on the Baltic Sea coast with brackish water as coolant while Ringhals is on the west coast, using real seawater, which might cause differences in water side fouling. The experience from the installations of close to 600,000 metres of 254 SMO tubes as replacement for copper base alloys (brass and copper nickel) in Finnish power plants is the same, i.e. there was no drop in total heat transfer [13]. Also Siemens KWU has reported an almost 100 % "changeover" from copper base alloys to stainless steel and titanium for power plant condenser tubes with an increase from 2.6 millions of metres of stainless steel tubes in 1980 to more than 14 millions in 1995 without any reporting of inferior heat transfer [14].

G
HEALTH HAZARDS AND ENVIRONMENTAL IMPACT
The water from desalination plants is to a large extent used for drinking, cooking and food processing and should consequently contain no harmful levels of heavy metal ions such as chromium and nickel which, together with iron, are the main constituents of stainless steel. There have been some media references indicating that the leakage of metal ions from stainless steel equipment for food cooking and processing might be high enough to imply a hazard to human health. However, several studies have shown that the release of nickel during such processes is extremely low. Schwenk reported in 1991 that some nickel can leach during the first weeks of service from new potable water pipes made of 304 stainless steel before the passive layer is in balance with the surrounding environment [15]. However, the leaching stops after a few weeks only, indicating that nickel migration is a matter of concern only during commissioning. His results were confirmed by tests at Swinden Laboratories presented at Innovation Stainless Steel in 1993 [16]. In a more recent test, performed by the National Food Agency of Denmark, 22 different electric kettles with either stainless steel (5 models) or nickel (10 models) or chromium plated (7 models) heating elements were investigated [17]. The release of nickel varied between the different brands of kettles and also between different cooking periods. One of the "stainless" kettles released significant amounts of nickel, i.e. from 30 to 115 g/l, which was most probably due to other nickel plated parts of the kettle, since the release of chromium from this kettle was below the detection limit of 0.5 g/l. The others had variations between "below the detection limit" and 4 g/l. Most of the "nickel plated kettles" released more nickel than is considered acceptable from an allergy point of view, i.e. 50 g/l. The release of chromium was generally very low, for stainless steel the maximum value measured was 1.5 g/l. The conclusion that can be drawn from these studies is that there is no health hazard involved when using highly alloyed stainless steels for condenser tubes and other critical components in desalination plants. There has also been some concern about the environmental impact of copper ions released from the huge surface of copper alloys exposed to seawater in desalination plants. This was indicated in a paper by

H
Shams El Din et al recently [18] and it was also a question addressed at the 2nd International Marine Corrosion Forum on Malta in April 1995 [19]. The level of copper in uncontaminated seawater is on tenths of ppb level, North Atlantic seawater has around 0.1 ppb [20]. However, only 1 % of this, i.e. 0.001 ppb, is available as toxic copper ions. A level of 0.003 ppb has been found to have acute effects on the larvae of plankton. The uniform corrosion rate of copper alloys depends on seawater velocity and presence of contaminants, hydrogen sulfide being one example, but rates of around 0.01-0.1 mm/year are not uncommon. The uniform corrosion rate of highly alloyed stainless steel is hardly measurable and does not exceed 0.001 mm/year [21]. This means that a large MSF plant such as Taweelah B, with a heat transfer area of more than 500,000 m2, will release at least 50 tonnes (maybe several hundreds of tonnes under unfavourable conditions) of copper to the surrounding sea during one year of service. The corresponding figure for a stainless steel such as 654 SMO would be far less than 50 kilos.

ECONOMY
Another important aspect of a new material is its influence upon the cost or the total economy of a project. It is almost impossible to give long lasting accurate figures due to fluctuating material prices. For example the price of titanium has gone down during recent years while stainless steel prices have increased. However, the following case can be regarded as a guideline for estimating the cost for the fabrication of equipment made of a highly alloyed stainless steel like 654 SMO. The case is based on calculations made by a Swedish fabricator specialising in big tube and shell heat exchangers with experience from designing and fabrication in highly alloyed stainless steel and titanium [22]. Component: Dimensions: Heat transfer area: Design temperature: Design pressure: Code: Tubes: No of tubes: Ejector condenser Diameter 325*1899 mm 11.3 m2 160C (shell and tube side) 5 bars ASME Div. XII OD 19* 0.7, 0.5, and 0.9 for Ti, 654 SMO, 100 pieces

an

The calculations are shown in Table 10.

I
This type of equipment will be more costly when made of a highly alloyed stainless steel or titanium than a conventional stainless steel such as 316L. However, type 316L is normally not resistant enough for this type of service and there is a pronounced benefit in using more resistant materials, shown already at the time of the first replacement which is bound to come if 316L is being used. There are also technical and economic advantages in specifying a highly alloyed austenitic stainless steel for larger condensers. Less sophisticated and less costly cathodic protection (CP) systems can be installed if the equipment contains less resistant materials demanding CP. Austenitic stainless steels are not susceptible to hydrogen embrittlement while ferritic stainless steels and titanium can suffer hydrogen embrittlement if CP giving too a low potential is being applied. The higher modulus of elasticity implies less demand for support plates in long tube heat exchangers. This is normally not so critical for new fabrication but it can become an important issue at the retubing of old condensers. When the first installation of 254 SMO condenser tubes was made at a Finnish power plant in 1982, the main reason for selecting a stainless steel was that this selection of material enabled a retubing only, without the need for extra support plates. If titanium had been selected it would have implied a total rebuilding of the condenser with additional costs and a longer shut down period [23].

CONCLUSIONS
1. The corrosion resistance of 654 SMO is very good and able to cope with the hostile environments existing inside condenser tubes of MSF desalination plants and power plants. 2. The corrosion resistance of 654 SMO is good enough to cope with many other hostile brine and seawater environments existing in desalination plants, e.g. in general seawater piping, plate heat exchangers, solid MSF chamber shells, ejector condensers of ventilation systems, and high pressure piping in SWRO plants. 3. The erosion resistance of 654 SMO can be an advantage in applications where the material is exposed to high velocity streams. 4. There should be no concern about the heat transfer characteristics of a stainless steel such as 654 SMO.

J
5. The use of a highly alloyed stainless steel like 654 SMO will have no negative impact on neither water as food, nor on the environment.

REFERENCES
1. Redmond J.D., Fairhurst W., and H. Watanabe, High performance stainless steels: Ten years of successful performance proves their worth in seawater-cooled condensers. Molybdenum Mosaic, 1983. 2: pp.1-8. 2. Olsson J. and J. Redmond, Application of UNS S31254 austenitic stainless steel in power plants. NACE Corrosion/91, Paper No 505. 3. Walln B., Some factors affecting stainless steel corrosion in seawater. Avesta Sheffield Corrosion Handbook for Stainless Steel, 1994: pp. 33-40. 4. Kain R.M., Seawater testing to assess the crevice corrosion resistance of stainless steel and related alloys. Proc. 12th Int. Corrosion Congress, Houston (1993), NACE, Paper No 347. 5. Walln B. and E. Alfonsson, Seawater resistance of a second generation superaustenitic stainless steel. Proc. Innovation stainless steel 1993, Florence, Italy, 3:pp.21-28. 6. Fron D. and B. Walln, Effect of high concentrations of chloride and sulfide on stainless alloys in seawater. NACE Corrosion/93, Paper No 498. 7. Rogne T. and R. Johnsen, The effect of CO2 and oxygen on the corrosion properties of UNS S31254 and UNS S31803 in brine solution. NACE Corrosion/92, Paper No 295. 8. Arnvig P-E. and W. Wasielewska, Stress corrosion behaviour of highly alloyed stainless steels under severe evaporative conditions. Proc. The Danish metallurgical society`s winter meeting 1993, Esbjerg, Denmark: pp.37-49. 9. Bodlund G., Steam side droplet erosion of titanium condensers. Swedish State Power Board report U(K) 1989/43, ISSN 1100-5130 (in Swedish).

K
10. Bodlund G, private communication in 1993 11. Maurer J., Development and application of new high technology stainless alloys for marine exposures. Proc. Advanced stainless steels for seawater applications, Piacenza, Italy 1980, Climax Molybdenum Co.: pp. 12-29. 12. Multer I., A comparison of heat transfer between an Al-brass and Ticondenser. Joint ASME/IEE power generation conference, MilwaukeeWisconsin 1985. 13. Lecklin E., Imatran Voima OY, Finland, private communication in 1992. 14. Siemens strives for cost-effective materials selection. Stainless Steel Europe,1995, 3:pp.34-39. 15. Schwenk W., Nickel migration from Cr-Ni stainless steel exposed to potable water. British Corrosion Journal, 1991. 4:pp.245-249 16. Tupholme K.W., Dulieu D., Wilkinson J., and N.B. Ward, Stainless steels for the food industries. Proc. Innovation stainless steel 1993, Florence, Italy, 1:pp.49-54 17. Pedersen G.A. and J. Petersen, Investigation of the release of nickel, chromium and lead from electric kettles and release of trace elements from coffeemachines. National Food Agency of Denmark, Report No ILF 1995.1, 1995 (in Danish). 18. Shams El Din A.M., Wang L., and T.M.H. Saber, Behaviour of high strength molybdenum containing stainless steels in Arabian Gulf water. British Corrosion Journal, 1994. 1:pp.58-64. 19. Second international marine corrosion Forum. Nickel Development Institute, Malta 1995. 20. The metals and the environment. The Swedish Environmental Protection Agency, Report No 4135, 1993 (in Swedish). 21. Rogne T., Drugli J.M., and R. Johnsen, Testing for the initiation of the crevice corrosion of welded stainless steels in natural seawater. Materials Performance, 1987. 9:pp.29-34.

L
22. Nordin R., rnskldsviks Mekaniska Verkstad, private communication in 1995. 23. Laurin T. and J. Redmond, Finland power plant, first in Europe to use stainless steel seawater cooled condenser. Molybdenum Mosaic, 1982. 4:pp.1-5.

M
Table 1. Typical chemical compositions and PRE-numbers of tested alloys
Alloy 316L 2205 904L 254 SMO 654 SMO 625 C-276 UNS S31603 S31803 N08904 S31254 S32654 N06625 N10276 C 0.04 0.02 0.01 0.01 0.01 0.05 0.02 Cr 17.3 22 20 20 24 21 15 Ni 11 5.5 25 18 22 60 60 Mo 2.2 3 4.5 6.1 7.3 9 16 N 0.06 0.17 0.06 0.20 0.5 Cu Cu Mn, Cu Nb W Others PRE 26 35 36 43 56 51 68

= Trademark registered by Avesta Sheffield AB

Table 2. Mechanical properties, min. values at room temperature

N
Table 3. Physical properties Alloy Young`s modulus, kN/ mm2 200 190 205 110 138 Thermal expansion, *10-6, 25-100C 16.5 15 12.2 9 17 Thermal conductivity, W/ mC 13.5 8.6 9.8 17 50

316L 654 SMO C-276 Ti, grade 2 CuNi 90/10

Table 4. Influence of different surface conditions on crevice corrosion in natural seawater, 25C.
Alloy 316 254 SMO 654 SMO C-276 As delivered 10/12, 0/12 0/12 0/12 1.7 mm Ground, 120 grit 2/2, 2/2, 0/2 2/2, 0.02 mm 1.6 mm 0.27 mm ditto + pickling 2/2, 0/2 0/2 0/2 2.2 mm

Note: 10/12, 1.7 mm means crevice corrosion on ten out of twelve possible crevice sites, max. depth 1.7 mm.

Table 5. Critical crevice corrosion temperatures (CCT) for some alloys in NaCl with 25,000 ppm of chlorides at 500 mV SCE.
Alloy 254 SMO 654 SMO 625 C-276 CCT, C <30 60 <30 65

O
Table 6. Crevice corrosion in chlorinated seawater with 10 ppm of Cl2, 45C. Alloy 254 SMO 654 SMO 625 C-276 Crevice corrosion 7/8, 0/8 8/8, 7/8, 0.47 mm 0.31 mm 0.98 mm

Table 7. Threshold stresses (TS) and time to failure (TTF) for stainless steels when tested in accordance with the DET.
Alloy 316 2205 254 SMO 654 SMO TS, % of Rp0.2 (200C) <10 40 90 >100 TTF, hrs 155, 158 471, >500 438, >500 >500, >500

Table 8. Number of droplets hitting the metal surface before first visible attack of erosion at two different droplet velocities.
Alloy Titanium Ti-modified 254 SMO 654 SMO Hardness, HV 162 211 200 255 175 m/s 130,000 315,000 480,000 630,000 200 m/s 90,000 240,000 260,000 420,000

P
Table 9. Condenser overall heat transfer in Oskarshamn nuclear power station.
Tube material Al-brass, no cleaning Ti, new tubes Ti, soft ball cleaning Ti, granulate ball cleaning Heat transfer, W/ m2*C 2,300 3,000 2,700 2,900

Table 10. Estimated cost in USD for a condenser made of stainless steel or titanium.
Type of cost Material Fabrication Others (NDT, design etc) Total 316L 3,400 12,000 5,500 20,900 654 SMO 6,900 12,500 5,500 24,900 Titanium 7,000 12,500 5,500 25,000