Soon to be published .. NiDI Nickel Development Institute
GUIDE TO THE SELECTION AND USE OF HIGH PERFORMANCE STAINLESS STEELS
Metallurgy - Properties - Corrosion Performance - Fabrication
A Nickel Development Institute Reference Book Series No. XX XXXX
Page 1 9/3/99 TABLE OF CONTENTS INTRODUCTION ...... 3
CLASSIFICATION OF GRADES ...... 5
AUSTENITIC HIGH PERFORMANCE STAINLESS STEELS ...... 5 FERRmC HIGH PERFORMANCE STAINLESS STEELS ...... 9 DUPLEX HIGH PERFORMANCE STAINLESS STEELS ...... 10 PHYSICAL METALLURGY ...... 13 PHASE RELATIONS IN THE IRON-CHROMIUM-NICKEL SYSTEM ...... 13 SECONDARY PHASES ...... 17 KINETICS OF PHASE PRECIPITATION REACTIONS ...... 21 AUSTENITIC STAINLESS STEELS ...... 22 FERRmC STAINLESS STEELS ...... 24 DUPLEX STAINLESS STEELS ...... 26 MECHANICAL PROPERTIES ...... 28 AUSTENmC STAINLESS STEELS ...... 29 FERRITIC STAINLESS STEELS ...... 31 DUPLEX STAINLESS STEELS ...... 33 PHYSICAL PROPERTIES ...... 34
CORROSION RESISTANCE ...... 36 RESISTANCE TO INORGANIC ACIDS .37 RESISTANCE TO ORGANIC ACIDS .43 RESISTANCE TO ALKALIES AND ALKALINE SALTS .46 CHLORIDE- AND OTHER HALIDE-CONTAINING AQUEOUS ENVIRONMENTS .48 NEAR NEUTRAL ENVIRONMENTS - NATURAL WATERS AND BRINES .53 INFLUENCE OF MICROBIAL ACnVITY .57 OXIDIZING HALIDE ENVIRONMENTS - CHLORINATED COOLING WATERS AND BLEACH SOLUTIONS .58 ACIDIC ENVIRONMENTS CONTAINING HALIDES- FLUE GAS CONDENSATES .61 STRESS CORROSION CRACKING ...... 65 WATER AND BRINE ENVIRONMENTS ...... 66 SOUR OIL AND GAS ENVIRONMENTS ...... 70 HYDROGEN ENVIRONMENTS ...... 73 CORROSION ACCEPTANCE TESTS ...... 74
FABRICATION ...... 78 HOT WORKING ...... 79 COLD WORKING ...... 81 ANNEALING ...... 83 MACHINING ...... 86
WELDING ...... 87 AUSTENmC STAINLESS STEEL GRADES ...... 89 FERRITIC STAINLESS STEEL GRADES ...... 92 DUPLEX STAINLESS STEEL GRADES ...... 93 SURFACE CONDITION ...... 97
APPLICATIONS ...... 98
Page 2 9/3/99 Materials Workshop for Nuclear Power Plants Stainless Steel and Nickel-Base Alloys Presented by Mr. Curtis W. Kovach Dr. Nicole Paulus-Kinsman of the Nickel Development Institute (NiDI)
Workshop Schedule
8:15 am Welcome and Introduction
8:30 am Basics of Stainless Steel and Nickel-Base Alloys
9:30 am Corrosion Basics 10:30 am Welding Stainless Steel 11:30 am Lunch 1:00 pm High Strength Stainless Steels (Discussion)
2:00 pm Advanced Stainless Steels 3:00 pm Microbiologically Influenced Corrosion 3:45 pm Conclusion
The format for this workshop is flexible and open. Questions may be asked at any time and examples from the floor can be discussed during the session or one-on-one afterwards.
September 9, 1999 NIDI NUCLEAR POWER PLANT WORKSHOP
Conpany US Nuclear Regulatory Commission
Location: Rodville. Maryland Date: SeptLeSt9
High ATTENDEE INFORMATION N0CE Perfromance Mailing List Stainless Steel Book
Name: Phone:
Title:
Address:
Name: Phone:
Title:
Address:
Name: _ Phone
Title:
Address:
Name. Phone:
Title:
Address:
Name: Phone_
Title:
Address:
Name: Phone:
Title:
Address - -
Curtis Kovach RUJEDI Consultant NICKEL DEVELOPMENT INSTITUTE Technical Iarketing Resources, Inc. 3209 McKnight East Drive Head Office Pitsburgh, PA 15237-6423 U.S.A. 214 King Street West, Suite 5 10 Toronto, ON Canada M5H 3S6 Telephone 412-369-0377 Telephone 416-5917999 Fax 412-367-2353 Fax 416-591-7987 ckotuch8trnr-inc.cosn
Dr. Nicole Paulus K, I7~~I~bDIPo"D I Consultant NICKEL DEVELOPMENT INSTITUTE Technical Marketing Resources, Inc. 3209 McKnight East Drive 1Iead Office Pittsburgh, PA 15237.6423 214 King Street West, Suite 510 U.S.A. Toronto, ON Canada Mll 3S6 Telephone 412-369-0377 Telephone 416-591-7999 Fax 412-367-2353 Fax 416-591.7987 npaulus6tmrinc.com "1 1 "1- J L L NiDI NUCLEAR SERVICE WATER PIPING CASE STUDY No 15002 Table I 1994 OSKARSHAMN NUCLEAR System Description POWER PLANT, FIGEHOLV1, SWEDEN DESCRIPTION UNIT 1 UNIT 2 UNIT 3 Start Up 1972 1974 1985 Coastal nuclear power plants, which use brackish and System Testing 1970 1972 1983 often polluted water in their service water systems, face Power Output (MWc) 460 620 1.210 one of the most demanding service environments in the Regular System 700 1,000 1,700 Flow Rate (kg/s) industry. The Swedish utility OKG AKTIEBOLAG has Emergency System 25 400 700 this operating environment at their Oskarshamn Nuclear Flow Rate (kg/s) Power Plant in Figeholm, Sweden. The brackish, Main Condenser 20,000 26.000 45,000 polluted Baltic Sea water used in the service water system Flow Rate (kg/s) caused extensive corrosion of the original system materials. Material replacement, testing, and evaluation emergency service water system in all three plants is nor- have been on-going since 1978 giving OKG some of the mally only operated during shut-down periods. A most extensive operating experience with 6 Mo austenitic constant low flow rate is maintained through the system stainless steels, titanium and other high performance when it is not in operation to prevent water stagnation. replacement materials of any nuclear power plant in the The emergency systems in all three units were designed world. This case study reviews the problems experienced so that, even if the pumps stopped, the operating system with original system materials; replacement material eval- would never be empty. The return pipes go to the highest uation programs; and actual performance of the alloys in point in the system. service; therefore, providing valuable insight for utilities with equally severe operating environments. The open sections of the service water systems use brack- ish, polluted water from the Baltic Sea. The water enters the plant through culverts and is pumped through a short THE SYSTEM section of underground piping before entering the units. The closed portions of the service water system use demineralized water. Units I and 2 have open, once- The Oskarshamn Nuclear Power Plant has three operating through service water systems in all but the reactor units. The start-up dates for these units and other basic building, which has a closed system. The majority of the operating data for the service water system can be found piping and over 100 heat exchangers in these units are in in Table 1. The units all have open emergency service the open portion of the service water system. These are the water systems and have both closed and open once- older units with start-up dates in 1972 and 1974, through portions of their regular service water system. respectively. Corrosion problems became evident in both The condensers in all three units are fed by the open units within two to three years after commencement of portion of the service water system. commercial operation. The first problems arose in the Both the standard and emergency service water system main condensers. These were followed by problems with piping was originally rubber-lined carbon steel. The the heat exchangers, and then, with the piping. Unit 3, which began operation in 1985, has both closed and open any portion of the system. It is possible that the low water temper- sections in its service water system with eleven titanium plate heat ature during much of the year has prevented a MIC problem. The exchangers made by Alpha Laval as the interface between them. water source and description have not changed since initial testing Since this unit is fairly new and since titanium was initially installed of the system. in both the heat exchangers and the condenser, there have been no significant corrosion problems to date. Corrosion problems with the During July and August, crustaceans can accumulate in the system. pipe are anticipated when the rubber lining begins to crack or wear. At one time, a trial study was done in which hypochlorite was added to the service water system to prevent crustacean growth. Although this worked well, the use of hypochlorite additions was not adopted THE WATER by the plant because of environmental concerns. No additional chlo- rine or other chemicals are currently added to the water except The brackish, polluted water used in the open, once-through por- during system shut-down. A cathodic protection system is used to tions of the service water systems is drawn from the Baltic Sea. All prevent deterioration of the condenser tubesheets and waterboxes. three units were tested two to three years before the unit start-up date
Table 2 MATERIAL EVALUATION Water Description AND REPLACEMENT
Source Baltic Sea Oskarshamn selects materials by reviewing technical literature, lab- Type brackish, polluted 0 oratory tests, and actual in-plant experience. The laboratory based Temperature 0 - 20'C (32 - 72F) testing program is sponsored by Vattenfall, the Swedish State Power pH 7.5 -8.0 Board, with financial support from participating power stations. The Chloride level 4,000 ppm testing program simulates the corrosion and erosion problems Magnesium level 250 ppm common to the participating plants but at an accelerated rate. The Calcium 100 ppm materials included in the program are titanium, SAF 2507® Silica 0.8 ppm (UNS S32750) duplex stainless steel, 254 SMO® (UNS S31254) and 654 SM0® (UNS S32654) austenitic stainless steels. All the stainless steels are produced by Avesta Sheffield AB. The samples and may have had some stagnant water in them during that time. are evaluated on an annual basis. Oskarshamn also closely monitors the performance of those materials already in service during the reg- The basic characteristics of the water supply are outlined in Table 2. The high chloride content of the water and presence of pollution ularly scheduled inspection program. would be highly corrosive to many alloys. The increased solubility Initial material replacement began in 1978 when a large number of of oxygen at lower water temperatures and the increased corrosion aluminum brass heat exchanger tubes in Units I and 2 were replaced rates that normally result under these conditions are probably an with titanium. The replacement program for Units I and 2 has been important contributing factor to the plant's problems. The facility is on-going since that time. Subsequently, the rubber lining on the not aware of any microbiologically influenced corrosion (MIC) in carbon steel pipe began to crack (due to age) and wear away. As the
Table 3 Underground and In-Plant Replacement Materials
254 SMO 1984 1985 1986 1987 1988 1991 TOTAL COMPONENTS 1 . _ _ Pipe . 1 length i m 26 160 48 36 146'1' 42 458 ft. 85 524 157 118 479 138 1,500 diameter mm 206-306 60-408 206 206 60-168 306 in. 8-12 2-16 8 8 2-7 12 Elbows 907 2 15 12 9 75 4
Cone Pipe Fttings 2 l l - | _ T-pieees l - | 2 i Condenser Tubes length m -- 54,000 54,000 1t. - - 177.120 177,120 diameter mm - - 24 ft. - - - . - 0.94 TITANIUM 1978 1979 TOTAL COMPONENTS Condenser Tubes lenghth m 288.000 540,000 ft. 944,640 826,560 1,771,200 diameter mm 24 25 in. 0.94 0.98 carbon steel is exposed to the service water, it corrodes quickly, intermediate cooling system is in the reactor building. Unit 3 was making replacement of problem sections necessary. 254 SMO, a built with only eleven plate heat exchangers exposed directly to 6% molybdenum stainless steel, has been the replacement material service water. It also has several closed loop intermediate cooling of choice for pipe, tube and other system components since 1984 systems for localized components which use demineralized water. because of its corrosion resistance, availability and strength. Unit l's and Unit 2's heat exchanger tubes failed faster than Although titanium also provides the necessary corrosion resistance, anticipated because of the high water velocity. As of 1991, Unit 3's it is more difficult to weld than 254 SMO stainless. Increased inert titanium heat exchangers have not required replacement. gas shielding is required, and skilled titanium welders were difficult to obtain. The higher strength of 254 SMO stainless was also an The tubes and tubesheets in fifty heat exchangers in Units I and 2 important factor in its selection. have been replaced with titanium. To avoid the vibration problems which can be caused by the modulus difference, six support plates Table 3 provides a summary of the titanium and 254 SMO stainless were added to each "church window". The titanium has performed installed to date at Oskarshamn. It should also be noted that small well in this application, but, because of difficulties experienced in amounts of 2205 stainless steel have been used in the past. This welding titanium, 254 SMO stainless has been the replacement material is no longer used because of the problems experienced with material of choice for heat exchangers since 1984. The replacement crevice corrosion in some of the flanges. of aluminum brass heat exchanger tubes and tubesheets is an ongoing program. Und rgrovrnd anmd lui-Plcrn Piping Cond ensers
Since the service water enters the plant through culverts, there is The original condenser tubes installed in Unit I in 1969 and Unit 2 very little underground piping. Submerged pumps for the three units in 1971 were aluminum brass. Due to corrosion problems that pull the water from the culvert into short sections of underground started within two to three years of installation, the condensers were piping to bring it into the plant. The return pipes are submerged Im retubed in 1978 and 1979 with 544,000 m (1.77 million ft) of (3.28 ft) below the water surface to prevent foaming and splashing titanium. The condenser tubesheets in Units I and 2 are carbon steel of the brackish water. The original underground and in-plant piping clad with Type 304 (18Cr-lONi) stainless steel, and, in Unit 3, they (including the emergency service water system piping) were carbon are clad with titanium. The steam side of the outer most titanium steel with a rubber coating on both the outside and inside of the condenser tubes in Units I sand 2 began showing signs of steam pipe. It was installed in the following years: Unit 1, 1969; Unit 2, droplet erosion as early as 1988. This has resulted in tube leakage 1971; and Unit 3, 1982. As the rubber began to age and crack, and plugging of problem tubes. All the Swedish and Finnish power Oskarshamn began to experience corrosion problems severe enough to require pipe replacement. The same corrosion problems exist in plants with titanium condenser tubes have reported similar both the normal and emergency portions of the system. The rubber problems. Frequent inspections of the outer tubes are done. OKG cracking and corrosion problem has been particularly severe in the re-tubed the Unit 2 condenser with a total of 54,000 m (177,120 ft) crevices of flange joints. The corrosion originated almost entirely on of 254 SMO in 1991. Unit 3's original main condenser tubes are the inside of the pipe. titanium and have not had any problems to date. In 1984, Oskarshamn started replacing sections of the underground pipe and the in-plant pipe in Units I and 2. 254 SMO stainless and SYSTEMR INSPECTION AND titanium were selected as replacement materials. No signs of corro- CLEANING sion have been observed in subsequent pipe inspections, and Oskarshamn is pleased with the performance to date. Both materials Routine inspections are conducted by shift personnel as a part of have worked well, but, since they have found 254 SMO stainless to normal plant operation. Water flow rates are monitored in the be much easier to weld, it will be used for all future installations. central control room and by local flow meters in the plant. The 254 SMO stainless has been the primary in-plant piping Shift personnel look for leakage and other potential problems. replacement material. A summary of the 254 SMO stainless piping In addition, the plant has a regularly scheduled maintenance sizes, elbows, fittings and other components installed to date is program which began in 1978 and has not changed significantly shown in Table 3. since its inception. Cleaning and inspection of the system occurs All of the in-plant piping system components are obtainable in during the annual refuelling outage. Whenever the system is shut 254 SMO stainless. down, the system is flushed with "drinking" water, and ferrosulphate is added to any remaining brackish water to prevent deterioration. It should be noted that there is still a great deal of rubber-lined pipe During outages, the culverts are drained and cleaned using high in the plant. Replacement is occurring as necessary, when leaks pressure equipment. appear or when detached sections of the rubber lining adversely affect the water flow rate. Much of the piping is still rubber-lined carbon steel. This is not generally inspected other than for through-wall leaks or for partial Hoft Eicchvmgers detachment of the rubber lining which causes reduced flow through the pipe. If either problem exists, the section is replaced with The original aluminum brass heat exchanger tubes were installed in 254 SMO stainless. The sections which have already been replaced Units I and 2 in 1969 and 1971, respectively. Each of these units with 254 SMO stainless, are inspected visually every five years in has over fifty shell and tube heat exchangers cooled directly by the accordance with the requirements of a 20-year warranty provided by open portion of the service water system. The only closed loop the manufacturer. The inspection is done by Oskarshamn personnel. The aluminum brass seawater cooled heat exchangers in Units I and glass beads, or by using pickling paste or pickling acid. Carbon steel 2 are cleaned and eddy current tested during the annual refueling brushes are not allowed under any circumstances. Even common outage. Since the titanium heat exchanger tubes have performed grade stainless steel brushes are not suggested unless subsequent well, they are cleaned and generally only visually inspected. Unit 3 chemical cleaning of the weld is done. Brushing can potentially has a high pressure water strainer with 3 mm (0.12 in) filter holes cause iron contamination of the surface which may initiate pitting in installed behind the intake pumps to prevent crustaceans from an aggressive chloride environment. entering the system. Because of this, the titanium heat exchangers in Unit 3 are cleaned every five years. Suspended solids are not considered to be a problem in any of the units. Ball cleaning CONCLUSIONS equipment is run continuously to keep the main condenser free of biofouling. %hen the system is stopped, the condenser is flushed and The demanding service water environment that exists at the dried. Extensive testing of the titanium condenser tubes is required Oskarshamn plant led the materials decision makers to use and gain due to the erosion-corrosion problems that have occurred in the outer experience with highly corrosion resistant materials. This large scale, rows of tubes. Eddy current testing of these tubes is done annually. long-term experience with titanium and a 6Mo austenitic stainless The pumps are visually inspected every year. Normally, one out of steel in a highly demanding service water system environment provides valuable assistance to materials decision makers four pumps is replaced with an overhauled unit so that each pump is encountering similar problems. Titanium; the duplex stainless steel, overhauled an average of once every four years. There is no formal 2205; and the 6Mo austenitic stainless steel, 254 SMO, have all been program for inspecting valves, but, if a valve is temporarily removed used as replacement piping materials in the plant. Titanium has from the system because of other repair work, it is inspected. performed well in all applications except the outermost rows of the condenser tubes, but, due to problems associated with welding, Oskarshamn does not plan to use titanium in the future. Crevice FABRI CATU ON corrosion problems with 2205 eliminated it from future consideration. The preferred replacement material since 1984 has Since Oskarshamn currently uses only 254 SMO austenitic stainless been the 6Mo austenitic stainless steel alloy 254 SMO because of its steel, it will be the only material discussed. All the field and shop corrosion resistance, availability, relative ease of fabrication, and fabrication was done by an outside contractor. The welding proce- strength. Although 254 SMO is the only 6Mo austenitic to have been dure used by Oskarshamn was provided by the steel producer, tried, it is assumed that the other alloys in this family would also Avesta Sheffield AB. The following is a brief summary of infor- perform well. Two alloys that are currently being tested, SAF 2507 mation provided. Several welding techniques are suitable: Gas duplex stainless steel and 654 SMO austenitic stainless steel, may be Tungsten Arc Welding (GTAW or TIG); Gas Metal Arc Welding considered for future installations. (GNIAW or MIG); or Shielded Metal Arc Welding (SMAW). The weld pool should be protected from atmospheric oxidation by an The 6Mo austenitic stainless steels are readily available in all the inert gas cover. Preheating, hot spots, and post-weld heating are not common product forms needed in constructing a service water system. recommended and can be detrimental to the material properties. This makes it possible to replace all existing system components to A filler metal with a chemistry equivalent to Alloy 625 is provide maximum system cost effectiveness and efficiency. Since recommended. The Oskarshamn plant used Avesta P- 12 filler metal there are several 6Mo austenitic stainless steel alloy producers from which is produced by Avesta Welding. The interpass temperature of which to choose, materials specifiers can generally select from a the work piece should not exceed 212' F (100C). This welding variety of sources based on the properties and components required, procedure is similar to the procedures recommended by other quality, technical support, service, delivery, and price. producers of 6Mo austenitic stainless steels. Flange joints were used We ackinowledge the assistance of Al r Fredrik Baniekow of OKG instead of welding to join dissimilar materials. A&7IEBOLAG lio tas indispensable in this project'S success. AIr Any heat tint should be removed after welding by either grinding Barnek-ow prnwided the infonnation used in this case study and with a fine abrasive, by abrasive blasting with 75-100 micron soda-lime retviewed thefinal case stut'for accuracy.
...... OFFICES OF THE NICKEL DEVELOPMENT INSTITUTE
North America Japan Mehrauli-Badarpur Road Material has been prepared for the 214 King SLW., Suite 510, Toronto, - 113. 5-chome, Shimbashi, New Delhi 110 062, India general Information of the reader and Ontario. Canada MSH 3S6. Minato-ku, Tokyo, Japan Tel. 91 11 698 0360 should not be used or relied upon for Tel. 1 416 591 7999 Tel. 81 3 3436 7953 Fax. 91 11 698 9522 specific applications without first secur- Fax. 1 416 591 7987 Fax. 81 3 3436 7734 AustralastIng competent advice. While the materi- Fax.uro416pe917987 Fax. &3436South ric Australasfa al Is believed to be technically correct, Europe Central &South America P.O. Box 28. Blackburn South NiDI, Its members, staff end consultants 42 Weymnouth St. cto Instituto de Metais Nao Ferrosos Victoria 3130. Australia do not represent or warrant Its suitabili- London, England W1N 3LO Av.9 de Julho. 4015 Tel. 613 9878 7558t o n enerao spefic us an Tel. 44 171 493 7999 01407-100 Sao Paulo-S.P., Brasil Mobile 61 18 346 808 ty any general or specific use end Fax. 44 171 493 1555 Tel. 55 11 887 2033 Fax. 613 9894 3403 any kind n connection with theinfoma- European Technical Information Centre Fax. 5511 885 8124 South Korea ton herein. The Holloway, Alvechurch India Olympia Building -Room 811 Birmingham, England B48 70B co India Lead Zinc Information Centre 196-7 Jamsilbon-Dong, SongpaKu, Tel. 44152 758 4777 Jawahar Dhatu Bhawan Seoul 138 229, South Korea Fax.44 152 758 5562 39 Tughlaqabad Institutional Area Tel. 82 2 419 6465 Fax. 82 2 419 2088
Feb 96/2.5 I 61:
3- -~~~~~~~~-
= S - E S~~~~~~~~~~~~~~~~~-
D - -- e
w . ** *
- S -- 0
-- S.D
0 he material presented in this publication has been prepared for the general information of the reader and should not be used or relied on for specific applications without first securing competent advice. The Nickel Development Institute, its members, staff and consultants do not represent or warrant its suitability for any general or specific use and assume no liability or responsibility of any kind in connection with the information herein.
Reprinted from CHEMICAL ENGINEERING, January 1990, copyright 1990 by McGraw Hill, Inc. with all rights reserved. Additional reprints may be ordered by calling Chemical Engineering Reprint Department (212)512-3607 THE RIGHT METAL FOR HEAT EXCHANGER TUBES
Arthur h. Tuthill,* Tuthill Associates Inc. Attention to this checklist of selection factors will materially W^ When designing a heat exchang- er, an engineer first calculates reduce heat-exchanger failures km*the surface area needed to car- ry the heat load. Next, he or she devel- The impact of each factor is noted ual chlorine and manganese. It also ops the design to meet the standards of without it necessarily being related to includes pH, temperature and scaling the Tubular Exchanger Manufacturers that of other factors. In the tables, the tendency (Table 1). Assn. (TEMA) for heat exchangers, or impact of each factor is rated one of Water cleanliness-Design engi- other codes, and the company's stan- three ways: green means the tubing neers tend to assume that cooling wa- dards. He or she then makes compara- alloy or alloy group has given good ter will be clean. This occurs only if the tive cost estimates, factoring in knowl- performance under the stated condi- right screens and filters have been edge from experience, and selects the tions; yellow designates that the tub- installed and operators have made best tubing metal for the service. ing may give good performance, but sure that they work properly. Debris Most unexpected failures of heat ex- may require a closer study of the con- (such as sticks and stones) and sedi- changers can be traced to a factor that ditions at the site and relationships ment (such as sand and mud) that are had not been fully taken into account with other factors; and red signifies passed through or around the screens when the tube material was selected. In that the material has not performed and filters have been responsible for Tables 1, 2 and 3, these factors are well under the stipulated conditions, many tube failures. arranged according to, respectively, and special precautions are required to Long term, copper-alloy and stain- water quality, the character of opera- achieve good performance. less-steel tubes perform excellently in tion and maintenance, and exchanger clean water (i.e., free of sediment, de- design. Tube materials considered are Water quality bris and fouling organisms). Too often, copper alloys, stainless steels (Types Water quality encompasses: cleanli- however, sediment and debris find 304 and 316), and high alloys (6%molyb- ness, and the content of chloride, dis- their way into exchanger tubes. Corro- denum, superferritics* and titanium). solved oxygen and sulfides, and resid- sion under sediment is common with *Anew family of stainless steels high in chromium content (25-30%). tubes of these two materials. A high "Consultant to the Nickel Development institute, 15 Toronto St, Toronto, Ontario, Canada I
1. Water cleanlinets ; -Sedimfe rn~;:+f :\;;:;j I: \00 I:B iolgical Tuibe Clean Mud Sand0 : 0:f;00 D e~bris \ f\t unts me:\ t alz \ concentration of sand can abrade the CA0::;0~: f| - n ill zh zId -:lEDd X\::n -Si protective film on copper-alloy tubes.t ftia:s :s00:1-l czzEiz~ :7e-ED Such service, therefore, requires a 70- 30 CuNi-2Fe-2Mn alloy in the copper- CA = CO pper* a X SS St in esssteel alloy series, or stainless steel. A lodg- ment of sticks, stones and shell fragments creates downstream turbu- lence, which can cause pinholes in cop- per-alloy tubes. The obvious cure for these problems is to screen the water better. As a safeguard, the newer 85-15 CuNi with 0.5 Cr, C72200 alloy and stainless-steel tubes do well in withstanding the ef- fects of lodgment turbulence. Copper- alloy tubes effectively keep organisms from becoming attached, and are pre- ferred when environmental restric- tions on the use of biocides would re- quire frequent manual cleaning of stainless-steel and higher-alloy tubes. Chlorides- These provide a conve- nient framework for differentiating the stainless-steel alloys. Type 304 stainless steel resists crevice corrosion at chloride-ion concentrations of less than about 200 ppm, and Type 316 does so at levels up to about 1,000 ppm. (The chloride content of U.S. fresh water is typically less than 50 ppm, for which Type 304 stainless steel is therefore normally adequate.) The 42% molybdenum alloys suf- fer crevice attack at chloride-ion con- centrations from 2,000 to 3,000 ppm. Both 4'/2% molybdenum-alloy and du- plex (another new family of stainless steels) tubes have been subjected to sea water (20,000 ppm chloride con- tent) and have undergone substantial crevice corrosion beneath fouling. On the other hand, the 6% molybdenum and superferritic alloys, and titanium have proved resistant to crevice and beneath-sediment corrosion in salt water. Dissolved oxygen and sulfides - Copper-alloy and stainless-steel tubes perform best in water having enough oxygen (about 3-4 ppm) to keep fish alive. These tubes also do well in de- aerated water, such as that used in water-flood oil wells. Copper-alloy tubes do not stand up well in severely polluted water in which dissolved oxy- gen has been consumed in the decay tAll alloys form a protective film in corrosive liquids; on stainless steel, it is chromium oxide; on copper alloy, it is cuprous hydroxychloride. process and sulfides are present. ment in which bacteria thrive, promot- Tubes of higher-alloyed stainless steel Leaving an ing microbiologically induced corro- or titanium have served successfully exchanger full, sion. Corrosion will also take place un- in such water. der sediment. If an exchanger is to be Residual chlorine - Both copper-al- or even only left full for more than 2 or 3 days, loy and stainless-steel tubes have per- water should be pumped through it formed well in water containing up to 2 wet, invites once a day to displace the stagnant ppm residual chlorine, and have failed water. If an exchanger will be down in heavily chlorinated water. Although corrosion for at least a week, it should be drained the usual objective is to keep residual and blown dry. chlorine at about 0.5 ppm at the inlet Cleaning schedule - Heat ex- tubesheet, this level is sometimes ex- ever, copper-alloy and higher-alloyed changers should be periodically ceeded and the residual is normally tubes have fared well in such water. flushed out, opened and brush cleaned, higher at the point of injection. Scaling tendency - Copper-alloy to remove sediment and debris and to Acidity -In aerated water of pH and stainless-steel tubes perform well restore heat transfer capability. Ex- less than about 5, a protective film in both hard (scaling) and soft (nonscal- changers handling water high in bio- does not easily form on copper-alloy ing) water. The Langelier saturation logical foulants or sediment should be tubes, so they corrode and thin rapidly. index* is frequently used to distinguish cleaned weekly. Micro-organisms will In deaerated water of low pH, copper- scale-forming from corrosive (to carbon corrode stainless steel if they and oth- alloy tubes resist corrosion well. For steel) water. er deposits are not removed. high-pH water, copper-nickel or stain- Monthly or even quarterly mechan- less-steel tubes are preferred to admi- Operation and maintenance ical cleanings of copper-alloy and ralty- (71% copper, 28% brass, 1%tin) Among these factors are type of oper- stainless-steel tubes are adequate or aluminum-brass tubes, which tend ation (regular vs. intermittent) and the with most waters. The optimum in- to corrode under highly alkaline condi- frequency of cleaning (Table 2). terval can be critical because the pro- tective film on copper alloy tubes can be damaged by cleaning with metallic brushes and some types of abrasive blasting. Mechanical cleaning is sometimes put off for a year, and even longer, particularly if restoring an exchanger's heat-transfer rate is not critical to plant performance. Be cautioned, however, that corrosion beneath sediment frequently occurs when sediment removal is delayed by more than three months. Exchanger design The principal design factors that in- fluence tube performance are water velocity, tube diameter, shape (i.e., once-through or U-bend), orientation, venting, tubesheet material, channel (waterbox) material and channel inlet tions. Stainless-steel tubes have per- Character of operation- Lengthy arrangement (Table 3). formed well at a pH less than 5 and startups have been responsible for Fluid velocity - At velocities of greater than 9. many failures of copper-alloy and less than 3 ft/s, sediment deposit, de- Temperature-A protective film stainless-steel tubes. These occurred bris buildup and biological fouling in readily forms on copper alloys in warm because water had been left in, or it and on tubes can be excessive, result- water (inabout ten minutes at 60TF), but had only been partially drained from, ing in the need for frequent mechani- forms very slowly in cold water. It de- tubes for a long time. Such failures cal cleaning, which can cause copper- velops almost instantaneously on stain- have also been caused by extended alloy and stainless-tube tubes to fail less steel in both warm and cold water. outages from normal operations. prematurely. Manganese - Type 304 stainless steel Leaving an exchanger full, or even Copper alloys can be conveniently tubes have failed in fresh water having an only wet, invites corrosion. The water differentiated according to their wa- appreciable manganese content. How- will become foul, creating an environ- ter-velocity tolerance. Approximate 'This index indicates the tendency of a water solution to precipitate or dissolve calcium carbonate. It is calculated from total maximum design velocities for tubes dissolved solids, calcium concentration, total alkalinity, pH and solution temperature. of copper-base alloys and stainless steel in salt water are listed in Table 3. Maximum velocity is usually arrived at as a compromise between the cost of pumping and the advantage in heat transfer. The design velocity usually falls in the range of 6-8 ft/s, and may reach 12-15 ft/s when extra cooling demand is placed on an exchanger, such as in summer. Copper-nickel tubes stand up reason- ably well to variations in velocity, al- though some erosion and corrosion may occur at the inlet. The C72200 alloy resists inlet-end erosion and corrosion, as well as corrosion downstream of lodgments. The 2Fe-2Mn modification of the C71500 alloy resists inlet erosion and corrosion excellently. Copper-nick- el tubes, after their protective film has aged, withstand considerable velocity excursions without significant inlet erosion. Stainless-steel tubes perform best at high velocity and are useful up to velocities that induce cativation. Cop- per will tolerate somewhat higher ve- locities in fresh water, 10 ft/s being a common design velocity for copper tubes in air-conditioning condensers cooled with fresh water. Tube diameter - Tubes of large di- ameter are preferred for heat ex- changers because any solids that pass through screens will also flow through the tubes. By one rule of thumb, tube diameters should be at least twice the diameters of the screen openings. Tubes should not be less than Y2 inch if the water to the exchanger is not filtered. Once-through or U-bend - Be- cause U-tube bundles are difficult to clean, the water must be very clean (e.g., boiler feedwater quality), or at least well-filtered. Both stainless- steel and copper-alloy tubes are liable to corrode beneath sediment. U-tubes are particularly prone to such corro- sion if sediment and debris are not removed from their bends. Once- through and 2- to 4-pass bundles are easily cleaned by flushing, water lanc- ing, or brushing. Orientation- Heat exchangers, particularly condensers, are normally placed horizontally, with water flow- ing through the tubes. Both stainless- steel and copper-alloy tubes perform well in such exchangers. In those un- usual circumstances that require that the cooling-water flow on the shell- Tubesheet material-Tubesheets have occasionally caused the corro- side of a horizontal exchanger, sedi- of carbon steel, common copper alloy, sion failure of copper-alloy and stain- ment and debris cannot be kept from Muntz metal (a brass composed of 58- less-steel tubes. Coatings applied to building up around the support plates 61% copper, up to 1% lead, with the steel and cast-iron channels for the and lower tubes. This results in corro- remainder zinc), admiralty brass and purpose of reducing corrosion some- sion beneath the sediment. The build- aluminum bronze are anodic to cop- times deteriorate and spall, leading to up can be eliminated, or at least re- per-alloy tubes. The galvanic protec- tube corrosion and failure. Coated strained, by upstream filters, but tion that these tubesheet materials channels are also liable to deep local these will not prevent fouling organ- afford copper-alloy tubes does not corrosion at pinholes in coatings isms from thriving in low-flow areas. eliminate all inlet-end corrosion, but brought on by an adverse galvanic When the water, even clean water, often does help to keep it at a tolera- couple between the steel channel and must be on the shellside, the tubes ble level. the alloy tube and tubesheet. should be of high alloy. Stainless-steel tubesheets are rarely Copper-nickel and aluminum- A condenser is sometimes oriented used with copper-alloy tubes, because bronze (solid or weld-overlaid surface) vertically, with the cooling water on stainless steel is cathodic to copper channels are preferred with copper- the shellside, to improve heat trans- alloy. Installing stainless-steel or tita- alloy tubes, and may also be used with fer. This allows noncondensable gas- nium tubes in copper-alloy tubesheets high-alloy tubes. The adverse galvan- es to collect under the top tubesheet, has accelerated the corrosion of the ic effect is diminished in such installa- tions, because the area of the channel is larger than that of the tubesheet, and the channel and tubes are not TEMA stC3ndards contiguous. Channels of Type 316 stainless steel (solid or overlaid) are recommend tt at exchangers preferred with stainless-steel tubes. be installe CI level, but Channel inlet arrangement- Un- ar even flow, restricted water passages they should really be sloped and poor inlet-piping entry arrange- ments have caused numerous failures of copper-alloy tubes, as well as a few letting the temperature of the tube tubesheets. Stainless steel and titani- failures of stainless-steel tubes. Basi- walls in the gas pocket get almost as um polarize so readily that the entire cally, the entry and flow pattern in both hot as the incoming gas being con- inside surface of the tubes (not just the large and small channels should distrib- densed, evaporating the water and de- part adjacent to the tubesheet) must ute the water uniformly to all tubes positing scale on the hot tube sur- be considered as the cathode to the with as little swirling as possible. faces. Sediment also tends to build up copper-alloy tubesheet anode. If a review of the foregoing factors in the bottom of a vertical condenser. The normal cure for this anode- indicates that an exchanger's tubes In vertical installations, both Type cathode problem is to protect the cop- must be of a metal that is even more 304 and Type 316 stainless-steel tubes per-alloy tubesheet with an im- corrosion resistant than a copper alloy are prone to stress corrosion and pressed-current cathodic protection or stainless steel, an engineer fre- cracking just under the top tubesheet. unit. The unit's potential must be con- quently resorts to a more highly al- Remedies include: venting gas trolled to avoid hydrogen embrittle- loyed metal, such a 6-molybdenum through the top tubesheet; changing ment of ferritic stainless steel, or by- stainless steel, a superferritic or titani- to copper-alloy or copper alloy-stain- driding of titanium. An alternative is um. These materials provide outstand- less steel bimetallic tubing or to high- to resort to tubes of 6-molybdenum, ing resistance to corrosion in crevices alloy tubing, or orienting the ex- which resist hydrogen embrittlement and beneath sediment. a changer horizontally. Although the without the need to carefully control TEMA standards recommend that ex- an applied potential. The author changers be installed level, exchang- Tubesheets for austenitic stainless Arthur H. Tuthill, a materials and corrosion consultant (2903 ers should really be sloped slightly so steel should be of identical composi- Wakefield Dr., Blackburg, VA that they will drain completely when tion. Tubesheets of high-alloy austen- 24060; tel.: 703-953-2626), has had extensive experience in the shut down, avoiding corrosion where itic stainless steel are used with tubes fabrication, use and perfor- tubes sag between support plates and of ferritic stainless steel, because tu- mance of metals, particularly retain water even after being drained. besheets of matching composition are involving heat exchangers, pip- Venting - Exchangers are normal- not available. Solid or clad titanium ing and pumps, in a vde rane in metallurgical engineering from Carnegie-Mel- ly fitted with vent cocks so they can tubesheets are preferred with titani- lon University and a B.S. in chemical engineering be purged to clear gas pockets. Con- um tubes. from the University of Virginia, he is a licensed engineer (metallurgy). Among the many articles densers, particularly when chlorine is Channel material- Corrosion that he has had published is a five-part series in Chemical Engineerin: "Installed Cost of Corro- used as a biocide, tend to suffer corro- products from bare-steel and cast-iron sion-Resistant Piping (Mar. 3 and 31, Apr. 28, May sion when gases are not vented. channels (exchanger inlet chambers) 26 and June 23, 1986). I - a
* S
* - - 6 5
S.. - *- S
=
* A
- . *!0 :
00 0 * 0 . .
5. ! . 5
. 1.
. 00! 0. 0
E1 5 ! 5 a1!
00-. - .| 0 .0
£* * . .
- S
I . | 6 5!5~ 5EEEE~~~~~~~~~~0 5
|~~~~~ 0. * 0 -tS S - !: 1
! ~~~- S - ! I 00g _t Z m . , = , x x~~~~~~-am
.** S I- * 0 0 0 0 0
_i~~~~i ~~. - . 0 _!
_~~~~~ -S -. m 0
_~~~* - S *S ;:s _! E z I ~~~~~~~~~~~6 _~~~~~ 00 0 m
- . . * :.. 000 . 0 _~~~~~~~~~~~~~~~~~~~~~ S-5*0 T:I8
I m I I
* - -- U..: