RECOVERY BOILER WATER AND FIRE SIDE DAMAGE MECHANISMS April 2015 Western Canada BLRBAC Craig Reid, M.A.Sc., P.Eng. Acuren Group Inc., Richmond, B.C.
GENERAL CORROSION FAC (Alias erosion) Feedwater Piping SULFIDATION Lower furnace OXIDATION Upper furnace
CRACKING MECHANICAL Simple overload Mechanical and Thermal fatigue ENVIRONMENTALLY ASSISTED (EAC) Stress Corrosion Corrosion Fatigue Combination of both
EAC ENVIRONMENTS Feedwater oxygen Fire side gases and smelt Metal temperature STRESS Residual stress from fabrication Cyclic stress from operation Residual + Cyclic
RESIDUAL STRESS Cold bending Welding Thermal expansion mismatch composite tubes A major driver for cracking Water side deaerators, economizer tubes, etc. Composite tubes floors (and wall tubes which think they are floor tubes)
THERMAL CYCLIC STRESS Due to temperature gradients Low frequency start ups and shutdowns Important for water side cracking Higher frequency combustion at primaries Corrosion fatigue cracking in composite tubes can enter the carbon steel inner layer
MECHANICAL CYCLIC STRESS Low frequency Superheater platen swinging due to sootblowing Higher frequency Resonant vibration due to gas flow Screen tubes and economizer tubes depending on anti-vibration
Water Side Cracking Corrosion fatigue, Stress assisted corrosion (SAC) cracking It s all about oxygen (and residual stress) deaerators, economizers, steam and mud drums, boiler floor and water wall tubes attemperators, desuperheaters condensate receivers steam piping where oxygenated condensate can dwell
Oxygen Thresholds Pitting means oxygen has been high enough to enable cracking if the stress state can drive it
1983 Deaerator Storage Tank Failure Head to shell weld failed Oxygen scavenger not added to the storage tank Oxygen pitting present
Cracking followed the heat affected zone in the shell
North Head (Intact) Typical cracked length 2 long X 0.2 deep
DEAERATOR STORAGE TANK FAILURE ANALYSIS Officially labeled as corrosion fatigue Welding residual stress identified as the prime cause no residual stress no failure Oxygen scavenger not added in storage tank so oxygen > 5 ppb
ECONOMIZERS the economizer inlet often experiences transient elevated oxygen at start up Scavenge oxygen by corroding economizer tubes Transient active strain is produced by temperature gradients during boiler start up Active strain is: added to residual stress concentrated at welded attachments e.g. fins Active strain + residual stress + > 5 ppb O2 = Water side cracking
Mill C water side cracking RB A RB B Hog PB Gas PB 1974 Start up 1991 Lo odour added second economizer 2012 First leaks water side cracking - 11 years 1991 Start up 2002 First leaks erosion thinning of tubes at lower headers 2004 Orifice change 2006 First water side cracking leaks 15 years 1978 Start up 2003 New economizer 15 years 2012 First water side cracking leaks 9 years 1985 Start up 2007 Wall tube leaks 22 years 2012 Tube leaks at drums (Gas fired PB has a very small economizer)
Originally Mill C - The oxygen sins of the fathers Feedwater from DA storage to economizers (O 2 scavenger only added at steam drum) Current Feedwater from DA storage 5 ppb O 2 O 2 scavenger added to feedwater line to recovery boilers (not yet to power boilers) Boiler fill Warm treated un-deaerated feedwater Feedwater from DA after boiler steaming
ECONOMIZER - new 1990
First Leaks at Header - 2002 turbulent flow (not FAC)
Orifice Change - 2004
2006 First Bottom Bend Leak The red circle indicates the leak. Bends cold bent. Residual stress at yield stress.
Close up view of the leak
Oxygen pitting
Close up of cracking in pitted surface arrows indicate crack tips.
Fluorescent magnetic particle testing revealed 40 transverse cracks 1/8" to ½" long.
Metallographic section showing cracking 0.094" deep.
2008 Straight Tube Cracking
Through wall crack
Additional Crack Growing diagonally across the fin
Additional Crack Half way through the wall
Mill C Hog Boiler Economizer water side crack leaking at shutdown
Eastern Mill Gas fired PB Gen Bank Tube
Mill C Gas Fired PB cracking due to residual stress from roller expanding
B&W RB, 750 psig design. 304L Composite wall tube installed in 1983 rebuild sample removed - 2013
Water Side Cracking cracking only occurs when critical values of both oxygen and active strain overlap in time. the total service time to the first leak depends on the number of start ups a year and the time during each start up when the critical combination of oxygen and strain occurs. Residual stress promotes cracking So stress relieve wherever practical
Gotaverken 1500 psig design
Bend 57 after grit blast cleaning. The surfaces displayed only superficial oxygen pitting and there was no evidence of under deposit corrosion.
Bend 57 after magnetic particle testing. Crack indications were found at the surface mandrel scarring.
The major crack extends from oxygen pitting. One adjacent crack had been blunted by pitting while another had not.
warm de-oxygenated water used for boiler fill in cross section the cracking is not classic water side cracking i.e. oxide filled wide cracks but cracks do initiate at superficial water side pits steaming rate increased from 585,000 lbs/hour in 1987 to 700,000 lbs/hour in 2012 speculate there are significant cyclic stresses during operation recommended all bends be replaced using stress relieved panels installed by welding membrane to membrane so no welding on the tubes
COMPOSITE TUBES Developed for resistance to sulfidation Cr increases sulfidation resistance Ni decreases sulfidation resistance but needed for welding and thermal expansion considerations Sanicro 304 Sanicro 38 625 overlaid 309 filler Sandvik 310/T22 383 filler Sanicro 67 ERNiCrFe-7 Cr Ni Mo Cu 18.5 10.5 20 38.5 2.6 1.7 23 58 9 23 13 25.5 21 27 31 3.5 1 30 60 29 60
PAPRICAN LAB TESTS 1% H 2 S in Nitrogen SA210 becomes unsuitable above about 400C metal T
PAPRICAN LAB TESTS
Stress Corrosion Cracking 1983 B&W Recovery Boiler, 1500 psig operating 304L Composite Floor Tube - 1993
1983 B&W RB Floor Tube Repairs 1993, 94, 95 1996 Floor replaced with Sanicro 38
1983 B&W Recovery Boiler Rebuild, 750 psig design 304L Composite Floor Tube - 2015
Sandvik Composite Tubes Sandvik Composite Tube Specification OD min delivered thickness total wall clad A210 A1 1 1/2 0.197 0.051 0.146 2 0.200 0.059 0.141 2 1/2 0.257 0.065 0.192 3 0.259 0.065 0.194
ASME ASME DESIGN tmin = (PD/(2S+P))+0.005D 2.5" 3" OD SA210 A1 2.37" 2.87" T = Tsat + ΔT Sources Recovery Boiler Ref Manual B&W 1700 psig RB Kvaerner radiant heat superheater Gotaverken 1500 psig RB ΔT > 50 F 83 F 122 F 3 F
Sandvik Composite Tubes Tsat + 122F Pre 1998 Allowable Stress psig Tsat (F) Tsat+ 122 S 2.5 inch 3 inch t min CA t min CA 600 486 608 15000 0.058 0.134 0.071 0.123 650 497 619 15000 0.062 0.130 0.075 0.119 700 505 627 15000 0.066 0.126 0.080 0.114 750 513 635 15000 0.070 0.122 0.084 0.110 800 520 642 15000 0.073 0.119 0.089 0.105 850 527 649 15000 0.077 0.115 0.093 0.101 900 534 656 14928 0.081 0.111 0.098 0.096 950 540 662 14856 0.085 0.107 0.103 0.091 1000 545 667 14796 0.089 0.103 0.108 0.086 1050 551 673 14724 0.093 0.099 0.113 0.081 1100 556 678 14664 0.098 0.094 0.118 0.076 1150 562 684 14592 0.102 0.090 0.123 0.071 1200 567 689 14532 0.106 0.086 0.128 0.066 1250 572 694 14472 0.110 0.082 0.133 0.061 1300 577 699 14412 0.114 0.078 0.138 0.056 1350 582 704 14288 0.119 0.073 0.144 0.050 1400 587 709 14148 0.124 0.068 0.150 0.044 1450 592 714 14008 0.128 0.064 0.156 0.038 1500 596 718 13896 0.133 0.059 0.161 0.033 1550 601 723 13756 0.138 0.054 0.167 0.027 1600 605 727 13644 0.143 0.049 0.173 0.021 1650 609 731 13532 0.148 0.044 0.179 0.015 1700 617 739 13308 0.154 0.038 0.187 0.007
Walls that think they are floors
Kvaerner/Tampella 1650 psig design 1250 psig operating 2007 Start up 304L composite wall tubes cracked up to carbon steel cut line no cracking in Sanicro 38
Front Wall PT crack monitoring areas
Cracks which enter the carbon steel PPRIC/ORNL Research Project Thermal fatigue alone cannot explain A corrosion component is also needed So Corrosion Fatigue Floor tubes only one case? Smelt spout opening tubes not uncommon Primary air port tubes rare Wall tubes not common but most mills do not look
B&W 1125 psig design 304 composite smelt runs
CE 1075 psig 1998 air system upgrade 304L composite tube wall installed 2007 VT 2008 Unusual rust bleed out from butt welds just above primaries PT cracking into carbon steel Excavated and weld repaired sample removal
46 tubes > 0.070 crack depth so into carbon steel deepest 0.125 so 0.063 into carbon steel
2007 sample tube
Cracking occurred at repair weld toe in less than 1 year Consistent with cyclic stress