Successful Real Time On-line Monitoring of High Temperature g g p - - PowerPoint PPT Presentation

Successful Real Time On-line Monitoring of High Temperature g g p Corrosion in Oxyfuel and Other Combustion Systems Combustion Systems

William M. Cox,

Corrosion Management Ltd., UK

Kevin Davis, Andrew Fry, David Swensen,

Reaction Engineering International, USA

Martin de Jong

DNV-KEMA, Netherlands

ABSTRACT

Real time high temperature corrosion monitoring is useful to investigate troublesome degradation mechanisms, to reduce the time required to undertake comparative materials testing, and to evaluate the effects of novel combustion environments. However, some do not believe monitoring of high temperature corrosion processes is possible using electrochemical

• instrumentation. Others have felt it should be feasible but have found

the results were unreliable the results were unreliable. This paper examines the development route of practical real-time high temperature corrosion rate determination, considers how the technology is consistent with traditional electrochemical theory and technology is consistent with traditional electrochemical theory and reviews results obtained in high temperature applications that demonstrate its use to characterize aggressive environments and fouling behavior in advanced combustion systems waste-to-energy fouling behavior in advanced combustion systems, waste-to-energy plants and oxy-fuel service in the UK, Europe, Asia and USA.

Relevant Applications pp

High temperature corrosion is an issue for

– fossil fuelled electric power generation systems t t b il – waste-to-energy boilers – gas turbines – reformers – reformers – fired heaters

tube metal temperatures typically ~350 to ~650ºC (660 to 1200ºF)

Drivers e s

There has been pressure in recent years for:

• improved energy efficiency
• CO2 reduction

This has resulted in developments in: p

– Low NOx combustion systems – Larger waste incineration plants Oxy fuel combustion – Oxy-fuel combustion These initiatives have considerable implications for tube metal corrosion behavior

Coupons p

Coupons provide useful data but: Coupons provide useful data, but:

– often short-term or lab tests – compromise test environment p – give only cumulative results – give only retrospective information – do not provide any control option – results vary from full-scale performance l t i b t t hi – plant excursions can be catastrophic

Results give sub optimal materials selection Results give sub-optimal materials selection

Relevant Electrochemical Theory

Anodic reaction: Anodic reaction:

Metal Me2+ + 2e-

Cathodic reaction: Cathodic reaction:

O2 + 4e-  2O2-

Combined:

2Me2+ + 2O2-  2MeO(dep)

Similarly, sulfur sulfides chlorine  chlorides

Other considerations

• Electrochemical processes are completely consistent

with established mechanical theories of high temperature corrosion protection p p

i.e. chrome oxide formers or nickel oxide formers, etc.

• “Electrolyte” can be molten but may be solid and

i d ti semi-conductive

• Corrosion is effectively just the reverse of

electrochemical materials extraction e ec oc e ca a e a s e ac o

Development Route p

• Early work on electrochemical potential (1930s)
• High temperature dc polarization studies (1960’s)

El t h i l i d (1970 )

• Electrochemical impedance (1970s)
• Electrochemical noise (1985)
• Full-scale Plant installations
• Full-scale Plant installations

Advantages of EN approach g pp

Avoids need for a reference electrode

• Avoids need for a reference electrode
• Does not require external polarization
• Evaluates only spontaneous transients

Evaluates only spontaneous transients generated by corrosion process itself

• Allows temperature control of sample
• Immediate response

EN is a DC Technique q

From Ohm’s Law:

V = IR

so:

V/I = R

and hence:

ΔV/ΔI = R(p)

Similarly:

V(n) / I(n) = R(n)

where: R(n) is equivalent to/analogous to R(p) is equivalent to/analogous to R ( ) q g (p) q g because in each case it is the ratio between the potential and current responses that gives the ‘R’ component

Specifically: Spec ca y:

Rn = Vn/In and from Stern Geary: y icorr = K . 1/Rn use Faraday’s Law to convert icorr to mpy

The value of K may/should be 1 but it can vary The value of K may/should be 1 but it can vary

Key Issue – Quantitative Rate y Q

1 Obt i ‘ t ti ’ l f th

• 1. Obtain a ‘representative’ sample of the

electrochemical activity

• 2. In fully immersed systems the cell factor may be

100%

• 3. In thin-film or semi-conductive conditions the cell

factor may be less than 100% (say 10% for example) factor may be less than 100% (say 10% for example)

• 4. In a boiler application the area factor may be (say)

25% (because the area under attack is only at 10.00 and 02 00 on the tube circumference) and 02.00 on the tube circumference)

• 5. Multiply the electrochemical response by the cell

factor and area factor to obtain the corrosion l / t ti t b il t b loss/penetration rate on a boiler tube

Typical Faults when using EN monitoring instrumentation EN monitoring instrumentation

• Dissimilar electrodes are used
• Only one electrode is used

y

• The ZRA is not properly balanced
• The influence of the cell factor is not

The influence of the cell factor is not understood

• The influence of the area factor is not

The influence of the area factor is not understood

• The circuitry is not adequately grounded

The circuitry is not adequately grounded

Instantaneous or Continuous?

EIM provides a ‘snap shot’

• EIM provides a ‘snap-shot’
• EIM works best under activation control
• EN provides continuous real-time output

– Allows on-line correlation with change in furnace environment/composition/metal temperature – works well under both activation and diffusion control – enables real-time verification of effectiveness of remedial action

Independent Verification p

• Waste to Energy (UK)
• Circulating Fluidized Bed (Taiwan)

Circulating Fluidized Bed (Taiwan)

• Waste to Energy (Netherlands)

P t h i l i (J )

• Petrochemical processing, (Japan)
• Oxy-Fuel (USA)

#1: Waste-to-Energy (UK) gy ( )

• Work done at Tyseley Waste-to-Energy
• Project supervised by CREED/Vivendi

j p y

• Two boilers handle 450,000 tpa MSW
• Superheater tubes lasted <18 months

p

• Probes installed in radiant and superheater

TWD Boiler Detail

Radiant Sensor

Superheater Sensor Sensor

Concerns

Damage had significant effect on maintenance cost

• Damage had significant effect on maintenance cost

and reduced the efficiency and availability of the plant.

• The mechanisms and causes of attack were not well

understood.

• Rate of damage varied widely on similar types of plant

d b ifi t l l diti and can be specific to local conditions

• Conventional measurement of the corrosion behaviour

is inherently difficult as it is normally retrospective and s e e t y d cu t as t s

• a y et ospect e a d
• nly possible over long operational periods.

Instrumentation

Superheater Corrosion Rate p

Superheater Trend Plot

• CTD. ENCR

20 per. Mov. Avg. (CTD. ENCR) 3.50E+01 4.00E+01 p g ( )

Average rate 3 97mm/y

2.50E+01 3.00E+01

Average rate 3.97mm/y

(tube wall thickness 4.8mm)

1 00E+01 1.50E+01 2.00E+01 0.00E+00 5.00E+00 1.00E+01 0 00 00 07/11/00 00:00 14/11/00 00:00 21/11/00 00:00 28/11/00 00:00 05/12/00 00:00

month 1 month 2

Stepped Temperature Traces pp p

Period: ~3 days y

#2: Circulating Fluidized Bed g

• Installation at a YFY Hsinwu Paper Mill
• Project supervision by ITRI, Taiwan
• 230 tph steam
• Designed to run on pulverized coal
• Fired instead on a high proportion of RDF,

tire chips, paper waste, sewage sludge

• Sensors in superheater and economizer
• High interest in fouling behavior

Circulating Fluid Bed Boiler g

Coal Replacement p

Fuel Coal Waste RDF - Sludge Fuel Type Coal Waste Tires RDF Diapers Sludge Inputs 40% 24% 13% 23%

• Coal and sludge contained Ca and Na

ash composition unknown

• Coal and sludge contained Ca and Na – ash composition unknown
• Tires typically contain Zn
• RDF ash composition unknown: Ca? Na?
• Overall sulfur-content ~1.5 wt%
• Some questions on fuel analysis

Superheater Sensor Installation p

Economizer Corrosion

Superheater Corrosion p

Fouling/Sootblowing g/ g

#3: Waste-to-Energy, Netherlands gy

• Installation at AEB Amsterdam
• Installation at AEB, Amsterdam
• Largest refuse-fired boiler in Europe

Installation supervised by KEMA

• Installation supervised by KEMA
• Purpose – comparative materials testing

4 b i t ll ti di t d h t

• 4-probe installation, radiant and superheater
• 2 materials on sensor – standard and

candidate candidate

• Evaluation of real-time transients in the

service environment service environment

AEB Waste to Energy Plant gy

Measurement Campaign at AEB p g

• Corrosion measurement campaign was

conducted at the high efficiency waste fired power plant in Amsterdam fired power plant in Amsterdam

• Part of the EU-project NextGenBioWaste
• Ultimate situation during campaign:

Ultimate situation during campaign: simultaneous surveillance by 4 corrosion probes N l i f

• No pre-selection of process parameters

New Methodology in Data Analyses: Surface Mapping Surface Mapping

36VLH84CP001 XQ50 P tert lucht voorw boven mbar 36VLH86CP003 XQ50 Tert lucht achtwnd boven mbar

Correlation of corrosion rate R2 and combustion air

36VLA31 CP021 XQ50 P prim air zone 2A. grat mbar 36VLA32CG001 XQ50 pos prim air contr damp Z % 36VLA33CF401 XQ50 F prim air grat zones 3A. Nm³/h 36VLA33CP031 XQ50 P prim air zone 3A. grat mbar 36VLA34CG001 XQ50 pos prim air contr damp Z % 36VLH20CF401 XQ50 F tertiary air Nm³/h 36VLH20CT001 XQ50 T tert air upstr of fan ° C 36VLA21 CF401XQ50F i i t 1 N ³/h 36VLA21 CP01 1 XQ50 P prim air zone 1 . grat r mbar 36VLA22CF401 XQ50 F prim air grat zone 2B Nm³/h 36VLA22CP022 XQ50 P prim air zone 2B. grat mbar 36VLA23CG001 XQ50 pos prim air CD z3A run2 % 36VLA24CF401 XQ50 F prim air grat zone 3B a Nm³/h 36VLA24CP035 XQ50 P prim air zone 3B. grat mbar 36VLA31 CG001 XQ50 pos prim air contr damp Z %

0,6-0,8 0,4-0,6 0 2 0 4

36VLA1 0CT001XQ50T prim air downstr of fan° C 36VLA1 1 CG001 XQ50 pos prim air contr damp Z % 36VLA1 1 CP021 XQ50 P prim air zone 2A grat r mbar 36VLA1 2CG001 XQ50 pos prim air contr damp Z % 36VLA1 3CF401 XQ50 F prim air grat zone 3A Nm³/h 36VLA1 3CP031 XQ50 P prim air zone 3A. grat mbar 36VLA1 4CG001 XQ50 pos prim air contr damp Z % 36VLA21 CF401 XQ50 F prim air grat zones 1 a Nm³/h 0,2-0,4

0-0,2

36VLA01 CF301XQ50LuftmengeZone1Nm³/h 36VLA02CF401 XQ50 F prim air upstr of fan Nm³/h 36VLA02CT001 XQ50 T suction prim air upstr ° C 36VLA04CF301 XQ50 Luftmenge Zone 3B Nm³/h 36VLA07CT001 XQ50 Master Ctrl Air preh. 2 ° C 36VLA08CT001 XQ50 Master Ctrl Air preh. 3 ° C 36VLA09CT001 XQ50 T prim air downstr of air ° C 36VLA1 0CT001 XQ50 T prim air downstr of fan C 36SBQ20CP001 XQ50 P 4 bar steam supp barg 36VLA01 CF301 XQ50 Luftmenge Zone 1 Nm /h

0-0,2 0,2-0,4 0,4-0,6 0,6-0,8

O2 mode Increased combustion air Steam mode

First pass deposits, superheater p p p

EN Probe Installation

KEMKOP coupon probes p p

Corrosion beneath deposits p

Radiant section roof

Lower portion of radiant section p

Tube Wall Corrosion

Boiler roof after cleaning

Radiant Section Tube Wall after cleaning after cleaning

Outcome:

• Client was able to obtain convenient

direct comparisons between candidate tube coating materials

• Client was able to identify transient

y

• perational conditions that increased

the risk of attack on the radiant and superheater tubes

#4: Petrochemicals Process, Japan

I t ll ti t Mi hi Pl t

• Installation at Mizushima Plant
• Installed in an OC (oxychlorination) Reactor

I ll i i d b MCC

• Installation supervised by MCC
• Purpose – identify when manifold attack took

l place

• Single-probe installation, dry fluid bed reactor
• Very high pressure and temperature
• Evaluation of real-time transients in the

i i t service environment

Typical manifold attack yp

OC Reactor Manifold Area

OC Reactor Manifold Sensor

EDC Plant – OC Reactor

R 11

t em p

R-11

250 300 0. 008 0. 009 0. 01

yr)

p ENCR

150 200

erat ure( ℃)

0. 005 0. 006 0. 007

R ate ( m m /y

50 100

Tem pe

0 001 0. 002 0. 003 0. 004

C orrosi

• n

/7 12: 00 6/8 0: 00 /8 12: 00 6/9 0: 00 /9 12: 00 /10 0: 00 10 12: 00 /11 0: 00 11 12: 00 /12 0: 00 12 12: 00 /13 0: 00 13 12: 00 /14 0: 00

0. 001

6/ 6 6/ 6 6/ 6/ 6/1 6/ 6/ 6/ 6/1 6/ 6/1 6/

Air feed Cat feed Reactor start

Outcome:

• Transient conditions, especially during

unit start-up and at load changes, etc., were identified that led to increased attack

• Remedial measures, mainly minor

adjustment to equipment operation, j q p p enabled the operators to avoid or minimize the periods of attack p

#5: Oxy-fuel Investigation, USA

• Study of oxy-fuel firing behavior in Pulverized

Study of oxy fuel firing behavior in Pulverized Coal fired power generation boilers

• Work conducted at REI in Salt Lake City

Work conducted at REI in Salt Lake City

• Investigation conducted on behalf of US DoE
• Purpose:

Purpose:

– evaluate impact of oxy-fuel on corrosion behavior – Evaluate comparative performance of different p p candidate alloy materials – T22, P91, 347H

• Radiant and Superheater sensors
• Detailed results reported in “Fire-Side Corrosion Rates of Heat Transfer

Surface Materials for Air- and Oxy-coal Combustion” by Andrew Fry last year

• Univ. of Utah 1.5MW Pilot Plant

Sensor probes and materials p

Example showing repeatability of comparative corrosion behavior comparative corrosion behavior

P i d Period: ~1 hour

“Bell Curve” effect on 347H

~600ºC ~500ºC

Bell Curve - Raask (page 327)

Bell Curve - Explanation p

EPRI (Viswanathan and Bakker)

( )

Summary

• Independent verification has confirmed that high

temperature EN instrumentation can provide accurate reliable real-time corrosion information

• Response is near-instantaneous to changes in service

combustion environment

• Oxy-Fuel work highlighted real-time responsiveness and

graphically confirmed previous ‘bell curve’ predictions

• Approach is equally applicable in molten salt or
• Approach is equally applicable in molten salt or
• xidation/sulfidation/choride attack conditions
• Technology is useful for corrosion investigation and

comparative materials testing

Plant Application pp

• The approach provides a powerful method of

evaluating current corrosion conditions

• It is especially useful for diagnosing and avoiding

It is especially useful for diagnosing and avoiding short-term excursions in plant operation that lead to premature corrosion failure

• It can be used advantageously to compare and

identify alternative candidate materials

• Perhaps its most exciting capability is the potential to
• Perhaps its most exciting capability is the potential to
• ffer real-time control and management of high

temperature processing systems

Now try and say EN doesn’t work!

Acknowledgements g

The authors would like to thank colleagues in each of their organizations, as well as to the various clients and collaborators who have supported projects and collaborators who have supported projects during the development period. In particular, thanks are due to Andrew Fry, Kevin p y Davis and David Swensen (REI), Arthur Stam (KEMA), Masazumi Miyazawa (MCC) and Adrian Metcalfe (C-M) for their input to the various projects Metcalfe (C M) for their input to the various projects discussed.

DoE Acknowledgement g

Some of the material presented in this paper is based upon work supported by the Department of Energy under Award Number DE-NT0005288 and was reported as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned i h R f h i ifi i l d i b d

• rights. Reference herein to any specific commercial product, process, or service by trade

name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. y g y

Thank ou for our attention Thank you for your attention