boiler tube materials, their failures and prevention

Webinar on

Reference /reading Material

INDIA BOILER DOT COM

B-2 Miraj Apartment, Near Natubhai Circle,

Inox Cinema Road, Race Course(W), Vadodara–390 007

Tel. 0265 – 2386658/ 6541352

Cell: 9099933061/9824277793

e-mail: [email protected] / [email protected]

mailto:[email protected]

mailto:[email protected]

Critical Analysis & Prevention of Boiler Pressure Parts Failure

1

Introduction:

In this section; we would discuss the analytical steps to diagnose the reliability of the pressure parts in the boiler system. This discussion is meant for the O&M personnel, and not for the failure analysis experts. In fact, while a tube failure analysis experts job is to investigate and identify the cause/(s) of a tube failure „Post Mortem‟, the O&M personnel‟s job is to identify the indications/ conditions prevailing in a running boiler and take preventive measures against the failure.

Again, in majority of these failure investigation reports; you may notice that the „Damage Mechanism‟ has been identified in a very conclusive manner. But the „correction‟ or the preventive measures would be seldom suggested.

The damage mechanisms are actually not too difficult to identify, particularly with the array of NDT and metallographic test facilities available these days. It‟s like finding out the cause of death after a person dies. But to identify what has led to that „cause of death‟ is a different challenge all together.

Various „Damage Mechanism‟s due to which a boiler tube fails can be broadly categorized as under:

1. Stress 2. Over Heating 3. Erosion 4. Corrosion 5. Fatigue 6. Combination of stress & corrosion

While the tube finally fails due to stress; in most cases the limitation is breached due to any, or, combination of the other mechanisms as mentioned above. Very rarely it would be a purely stress failure, unless, there is a design failure, or, wrong metallurgy has been used by mistake.

Successful, reliable operation of steam generation equipment requires the application of the best available methods to prevent overheating, erosion, unwarranted stress, scale and corrosion. When equipment failures do occur, it is important that the cause of the problem be correctly identified so that proper corrective steps can be taken to prevent a recurrence. An incorrect diagnosis of a failure can lead to improper corrective measures; thus, problems continue.

There are times when the reasons for failures are obscure. In these instances, considerable investigation may be required to uncover the causes. However, in most cases the problem area displays certain specific, telltale signs. When these characteristics are properly interpreted, the cause of a problem and the remedy become quite evident.

So the point is, “Why wait for a failure to occur first?”

In fact, in most case, when the similar type of failure in any particular location recurs a number of times, then only we start investigating, doing the post-mortem, calling the experts, finding the root cause, etc., etc.

But a boiler tube seldom fails immediately after it starts getting damaged.

In 99.99% cases, an imminent tube failure would start giving indication long before it happens. The various damage mechanisms like overheating, corrosion, erosion, fatigue, etc. are initiated in the pressure part due to some undesirable conditions taking place in the system where they are working. When the O&M personnel start identifying these conditions, they can take necessary precautions to stop the tube failure.


2

To predict on the reliability of the pressure parts and take preventive actions, the engineer needs to know:

What adverse conditions are present in the boiler system?

What are the metallurgies of the pressure parts and what are their limitations to such adverse conditions?

What are the indications which will suggest that the limitations could be breached?

Adverse Conditions in Boiler System:

The first two things that come to your mind when you think about the adverse conditions in a boiler system are the HEAT and PRESSURE. Let us discuss about the heat first. The heat inside the fuel fired furnace is immense! The furnace temperature in most of the conventional fossil fuel fired boilers is in the range of 1100 – 1200oC. Even in a FBC system, where the furnace temperature is controlled by the bed material, the furnace temperature remains in the range of 850 – 950oC. Whereas temperature limit for Oxidation characteristics of various metals used in Boiler

are:

Carbon Steel: 425oC (456oC)

LAS T11: 550oC

LAS T22: 580oC

The problem is not just heat. These are „Pressure Parts‟; therefore subjected to continuous pressure resulting stress in the materials.

When the metal is expected to be stressed, then the selection of size and thickness are made on the basis of minimum allowable stress of the material under the specified condition. The ASME section II, Part D [8] gives a trend curve on condition of tensile, yield stress and certain percentage of creep as under.

The above trend curve would suggest that above 700oF; i.e. 370oC temperature, the creep would become dominant factor for the CS and LAS materials used in boiler pressure parts.

The first thing we therefore would need to focus on is „Creep‟. What is creep?

Creep may be defined as a time-dependent deformation at elevated temperature and constant stress. It follows, then, that a failure from such a condition is


3

referred to as a creep failure or, occasionally, a stress rupture. The temperature at which creep begins depends on the alloy composition. It should be pointed out that the actual operating stress will, in part, dictate or determine the temperature at which creep begins.

The end of useful service life of the high-temperature components in a boiler (the super heater and re heater tubes and headers, for example) is usually a failure by a creep or stress-rupture mechanism. The root cause may not be elevated temperature, as fuel-ash corrosion or erosion may reduce the wall thickness so that the onset of creep and creep failures occur sooner than expected.

However, regardless of the cause, the failure will exhibit the characteristics of a creep or stress rupture. Indeed, the ASME Boiler and Pressure Vessel Code recognizes creep and creep deformation as high-temperature design limitations and provides allowable stresses for all alloys used in the creep range. One of the criteria used in the determination of these allowable stresses is 1% creep expansion, or deformation, in 100,000 hours of service. Thus, the code recognizes that over the operating life, some creep deformation is likely. And creep failures do display some deformation or tube swelling in the immediate region of the rupture.

The allowable stresses permitted by the various construction codes are based in part on time-dependent creep properties. For carbon steels, these time-dependent properties dominate the allowable stress above about 750ºF (400ºC), although creep begins to occur in carbon steels at about 700ºF (370ºC).

Figure below shows a schematic creep curve for a constant load; a plot of the change in length verses time. The weight or load on the specimen is held constant for the duration of the test. There are four portions of the curve that are of interest:

An initial steep rate that is at least partly of elastic origin, from point "0" to point "A".

This is followed by a region in which the elongation or deformation rate decreases with time, the so-called transient or primary creep, from region "A" to "B". The portion from point "0" to point "B" occurs fairly quickly.

The next portion of the creep curve is the area of engineering interest, where the creep rate is almost constant. The portion from "B" to "C" is


4

nearly linear and predictable. Depending on the load or stress, the time can be very long; two years in a test and several decades in service.

The fourth portion of the creep curve, beyond the constant-creep-rate or linear region, shows a rapidly increasing creep rate which culminates in failure. Even under constant-load test conditions, the effective stress may actually increase due to the damage that forms within the microstructure.

Without going into a detailed discussion of the atom movements involved in creep deformation, suffice it to say that creep deformation occurs by grain-boundary sliding. That is, adjacent grains or crystals move as a unit relative to each other. Thus, one of the microstructural features of a creep failure is little or no obvious deformation to individual grains along the fracture edge.

The first two stages will not leave any microstructural evidence of creep damage. Somewhere along the linear portion of the curve, the first microstructural evidence of damage appears as individual voids or pores. These individual voids grow and link to form cracks several grains long, and finally failure occurs. The ultimate rupture is by a tensile overload when effective wall thickness is too thin to contain the steam pressure.

The Larson-Miller Parameter (LMP) can be used to determine the expected life of a component. Temperature and time are combined in the LMP, which can be expressed as LMP = (ºF + 460) (C + Log10 t) (10-3)

Where (ºF + 460) is the absolute temperature, C is a constant assumed to be 20 for carbon and low-alloy steels, and t is the time to failure in hours. The value of LMP can be plotted from Graphs for different material groups are available.

The LMP equation suggests that if any metal component, which is operating in the Creep regime, is exposed to higher temperature than which it is designed for, the life period would decrease by logarithm scale.

For example; for a 55oF (30.5oC) rise in metal temperature, creep life loss could be around 90%. Whereas, if the temperature increases by just 10oF (5.5oC), the creep life loss would be 60%!!

Though there are in-situ metallographic tests (Replica study) available to monitor the creep damage occurring in the components working in the creep regime, the metal temperature, gas temperature, attemperation spray quantity are more useful indicators to the O&M personnel to identify the possibility of creep (over heating) damages.

Let us now discuss about the second most adverse condition, which is PRESSURE.


5

The pressure parts like water wall, super heater, headers as well as the steam drums are all considered to be Thin Cylinder. When a thin cylinder is pressurized, it would try to rapture in two directions as indicated in the figure below, and the resultant stress in the component would be developed in these two directions only.

From the basic knowledge of Strength of Material, we can work out these two stresses on the basis of the internal pressure P, internal diameter D and the thickness T; and they are:

Longitudinal Stress: σL = (P x D) / 4 t

& Hoop Stress (Circumferential Stress): σc = (P x D) / 2 t

It is evident from the above equations that the Hoop Stress is two times larger than the longitudinal stress and would be primarily responsible for the failure if;

The allowable stress is reduced due to over heating Internal pressure increases due to local blockage, or The thickness is reduced due to erosion/ corrosion

………….And the rapture would be axial in such cases

In case the thickness loss is caused by water side corrosion; the internal diameter D increases while thickness T is reduced, resulting more significant increase in the stress. As compared to thickness loss due to erosion, or fire side corrosion, this would lead to a tube failure much faster.


6

That is the reason why the corrosive properties in water should also be considered as highly adverse condition, particularly in case of a boiler operating at high pressure and therefore at a high temperature.

So we can see that high temperature; high pressure, ash erosion, corrosion are the various adverse conditions in which the boiler pressure parts have to work and survive. That is why, the type of metal selection, and their capability to survive such conditions become very important to understand.

Boiler Tube Materials:

Though boiler pressure parts are made of metal, it would not be too unrealistic, if we label the Boiler material as “Fragile – Handle with Care”. We need to first understand the characteristics and limitations of different boiler material and various adverse conditions in the boiler which influences them to appreciate this apparently exaggerated remark.

Boiler Tube Materials:

Material Selection Aspects:

Plain low carbon and chrome-Moly type ferritic low alloy steels are mainly used for boiler

tubes of power plant boilers. These steels are used for tubes of almost all sizes of boilers

associated with power plants. Presently largest units of thermal power plants in our

country are of 500 MW capacity and here too boiler tubings are made of the above

mentioned carbon and low alloy steels only.

The boiler designers, considering cost, fabricability, space limitation and stability at high

temperature, have opted for the best possible creep resistance of carbon and chrome-Moly

type low alloy steels. It is to be noted that the bottom most position of the outermost

tube (called wrapper circuit) of the platen superheated assembly is made of A1S1 type 347

austenitic stainless steel. The large thermal expansion of the austenitic stainless steel is

not a limiting factor here as there is no interference in immediate vicinity of the wrapper

component (the same is not true for other vertical assemblies). At this location, in

addition to very high temperature, exist very high erosive attack of flue gas as it is

changing its flow from vertical to horizontal. With availability of necessary freedom for

thermal expansion, it became possible to use a creep and wear resistant material (i.e.

austenitic stainless steel) at a place where it is most needed.

It is not that factors associated with service conditions are not taken care of. The heat

exchanging assemblies have been so placed that flue gas flows uniformly without causing

excessive wear. Internal corrosion is totally left to the control of water chemistry. It is the

responsibility of boiler operators to see that „over heating‟ conditions do not develop (and

thereby creep conditions are avoided).


7

Standard Specifications

The low-carbon steels and the Cr-Mo type low alloy steels are mainly selected for the

boiler tubings of TPPs. At a few places, where service conditions are extremely severe and

the thermal-expansion-limiting space-restrictions do not exist, austenitic stainless steels

are also used. The steels generally used for boiler tubes are listed as per their „nominal

composition‟ below:

(A) Carbon Steels:

1. Low carbon (carbon below 0.2)

2. Medium carbon; (Carbon 0.30/0.35 max.)

(B) Low alloy Steels:

1. Moly type: (1/2 Mo)

2. Chrome-Moly type: (1/2 Cr-1/2 Mo), (1 Cr-1/2 Mo), (1.1/4 Cr-1/2 Mo), (2 Cr-1/2 Mo), (5) Cr – ½ Mo), (2.1/4 Cr-1 Mo), (3 Cr – 1 Mo), and (9 Cr- 1 Mo with Nb & V)

3. Chrome-Moly-Vanadium Type: (1/2 Cr-1/2 Mo – ¼ V).

(C) Austenitic Stainless Steels: (Type 304), (Type 316), (Type 321), (Type 347).

Materials for different pressure parts:

The actual material, size and thickness of a particular tubing system vary with the capacity of the boiler and the designer/manufacturer of the boilers. However the basic materials used for individual systems have not changed much. The reasons for having limited choice in selection of materials are already discussed. It is not to say that research and development has not made any difference. With advances in development the criteria for design parameters, acceptance standards, and approval methods have changed to a great degree. From materials point of view, abandoning the use of Vanadium bearing creep resistant steels is noteworthy. Some boiler manufacturers do use this type of steel but only at places where flue gas is out of contact. This is so because vanadium of steel and sodium (if present in flue gas) may form undesirable complex compounds which in molten stage are extremely corrosive even for highly alloyed stainless steels.

While WW and ECO systems are of carbon steel tubes, those of RH system are of low alloy steels. In case of SH system all varieties of steels are used depending on the temperature of the location.

The HEEs of WW, ECO, SH and RH systems are mainly made of seamless variety while larger pipes (ICPs, Headers) are made of boiler quality plates welded at the seam by ERW/SAW (submerged arc) practice.

The material specification of individual HEEs of the WW, ECO, SH and RH systems are listed in Table No. A(1) for a Power boiler up to 500MW .


8

TABLE NO. A (1)

Material specifications for Heat Exchanging Elements

Sr. No. Heat Exchanging Elements Specification

I ECONOMISER

1. Economiser Coils

SA 210 Gr. A1/C

II WATERWALL

1. Wall Panels

2. 2. Platen assemblies

SA 210 Gr. A1/ C

SA 210 Gr. A1/ C

III SUPERHEATER

1. Radiant roof Panel

2. SCW – side walls

3. SCW – front wall panel

4. SCW – Roof & rear wall panel

5. LTSH (Low temp.) Assembly

6. Platen S.H. Vertical assembly

7. Final S.H. Vertical assembly

SA 210 Gr. A1, SA 209 T1

SA 210 Gr. A1

-do-

-do-

SA 210 Gr. A1, SA 209 T1,

SA 213 T11

SA 213 T11, SA 213 T22,

SA 213 TP 347 H

SA 213 T22

IV REHEATER

1. Reheater coils

SA 209, T1, SA 213 T11,

SA 213 T22, SA 213 T91


9

Causes of Boiler Tube Failures:

Different Aspects of the Causes:

Understanding the cause of any failure is essential to ascertain whether the failure is “unavoidable” or is “avoidable” by preventive measures. In case of the BTFs a critical analysis of causes is highly valuable as most of the BTFs are avoidable. A critical study of the causes of the BTFs is further justified on account of the resultant non-availability of the power plant, generation losses (imagine a 210 MW power plant out for a minimum of four days if a BTF has necessitated boiler shut down) and associated industrial production losses.

Since a complete discussion of all causes of the BTF would be too voluminous, they are discussed only in brief in this literature survey.

Before going through the causes, an important aspect of these causes is discussed as it is important from a failure analysis (of the BTFs) point of view.

The (responsible for the BTFs) can be considered as comprising of three stages. First one is the Root cause (RC), and the last one is Damage Mechanism (DM). In between we have a stage of UC (Undesirable Condition)s as explained below:

For example consider that an out of alignment portion of a reheater tube has failed because of preferential erosion by flue gas. Here erosion is the damage mechanism which caused the failure but the root cause is the out-of-alignment of the portion of the tube which is responsible for the setting in of the damage mechanism.

In between these RC and the ultimate DMs there is an intermediate stage. A particular DM may set in because of either the tubes materials inferior corrosion resistance or development of excessive corrosive conditions. An apparently simple mistake of a mild steel tube at the location of a low alloy steel tube develops the first condition while presence of undesirable water chemistry develops the second condition. Thus between the RC and the DM is the stage of UC. A specific RC leads to the development of UC which sets in a DM ultimately causing the failure of the boiler tube.

Understanding of these stages is important from failure analysis point of view. Usually the failure investigator rarely finds evidence positively pointing toward a particular RC. He therefore generally follows a method of elimination. Once the investigator knows the UC (out of all possible conditions which can set in the DM) he can eliminate all RCs related to other possible conditions. This helps him a lot in the investigation. It is also important to know about the source of a RC. For example inadequate use of a water treatment chemical causes corrosion failure of a boiler tube. Here the RC (i.e. inadequate use of water treatment chemical) means „undesirable water chemistry‟. Thus a particular RC can be identified with an area of the boiler engineering and the same can be considered as source of the RC. Knowing the area of the boiler engineering of a RC” (i.e. a source of a RC) it becomes possible to assign responsibility of preventing similar failures to the concerned agency (unit/group/section/contractor/company). There can be different RCs (causing different DMs) belonging to one source.

Thus a source of cause (soc) is as important as the RC itself hence they are listed and classified in Table B (1).


10

TABLE NO. B (1)

The Classified List of the sources of the Causes Responsible

For the Boiler Tube failures

I. Design related

1) Improper material selection

2) Defective design of components and their assembly.

II. Fabrication related

1) Improper fabrication

2) Poor/Inspection and testing

III. Operation related

1) Undesirable water chemistry

2) Poor chemical cleaning

3) Undesirable flue gas chemistry

4) Undesirable coal and fuel oil

5) Improper checking

6) Failure of a component

IV. Repair related

1) Improper fabrication

2) Improper inspection and testing

3) Insufficient shut down and testing precautions.

A metallurgical investigator‟s first target, while undertaking a BTF analysis is to find the DM responsible for the failure. In this chapter on the theoretical background of the causes for BTFs this aspect is emphasized.

These days when a component (like a Boiler Tube) can fail in various manners and there are quite a good number of possible causes, the failure analysis find the Fault Tree Technique very useful in fault diagnoses. In a modified manner a fault tree for each of the DMs is prepared by interconnecting the DM with their UCs/RCs/SOCs. These modified fault trees are shown in block diagram type diagrams.

Identify prospective failure location and Prevent Failure:

Since every damage mechanism is preceded by a Undesirable Condition and since very rarely a damage mechanism is so aggressive that it leads to immediate failure, it gives us an opportunity to identify an Undesirable Condition in the system and detect the damage mechanism well in advance before it leads to failure. If we can replace a damaged tube before it fails and causes a forced stoppage, we can save a lot of money and increase the reliability of the system.

Contribution of Different Damage Mechanisms:

Following is list of all well-known DMs responsible for the BTFs.

1. Rupture (overload)

2. Corrosion


11

3. Erosion

4. Creep (over heating)

5. Fatigue and Thermal fatigue

6. Embrittlement

Most of the failures of steam-power plant equipments are due to over-heating, corrosion, fouling, poor workmanship, improper material and defective material.

It is beyond doubt that a large proportion of the BTFs is always found to be due to overheating. It was observed in a survey of cause wise tube failures (covering about 350 cases over a period of 5 years in USA) that overheating accounts for more than 50% of tube failures.

Even in developed countries in initial stages, when water chemistry measures were not well established corrosion type failures were more common.

In our country, the measures of controlling water chemistry are well established now and are well practiced at almost all power stations and hence the BTFs related to the corrosion DM are not very common. Till enough design experience was gained the fatigue (thermal fatigue in particularly) failures were quite common. Now a days such cases are rare.

The factor “Tube external wastage erosion” is equally dominant in the BTFs in our country. In fact, more than two third of the BTFs which occurred at one TPS in Gujarat, were due to erosion (Flue gas side).

The individual DMs are discussed in subsequent articles.

Over Heating (Creep):-

Logically there can be three conditions (i.e. UCs) which can cause overheating of boiler tubes:

(1) Tubes subjected to excessive heat flux.

(2) Insufficient heat extraction by the medium to be heated due to its insufficient flow

(3) Poor heat absorption by medium because of development of insulation by internal deposit.

These three conditions are called: (a) excessive heat exposure (b) flow starvation and (c) Internal Deposit. The RCs for these SCs are discussed below.

Excessive Heat Exposure:-

(a) Flame impingement – which in turn may be due to –

– 1. Burner distortion

– 2. Disturbance in location of burner with respect to water walls

- 3 undesirable flame shape

- 4 undesirable design/setting of burner throat.

(b) Partial restriction of flue gas at a location creating excessive flow at other location due to improper flue gas flow pattern.


12

(c) Excessive ash deposit on furnace walls (due to insufficient deslagging or soot removal) causing excessive heat flux at suspended assemblies (i.e. RHs and SHs).

Flow starvation:

Following things, at flow changing locations i.e. bends ect. can cause flow starvation:-

- a. Corrosion products

- b. Carry over salts

- c. Foreign materials like weld slag, sand, mud, rust (due to improper boiler cleaning or preservation).

From failure investigation point of view it is important to note that most of the time no evidence of chocking material is detected because it gets blown out when tube ruptures. Flow starvation also results if water level in the boiler drum remains low for a considerable time. In such cases quits a good number of water wall tubes get damaged.

In case of steam carrying systems i.e. (SHs/RHs) flow starvation may result from excessive extraction of auxiliary steam.

Internal Deposit:-

As a protective measure it is seen by boiler operators that magnetite layer (Fe3O4) of a limited thickness is maintained inside the boiler tubes. However undesirable water chemistry sometimes leads to excessive thickness of Fe3O4 which then acts as insulator to heat flow.

In addition, poor water chemistry may lead to excessive formation and deposition of oxides and corrosion products inside boiler tubes. Creep failures of water platen tubes because of deposition of copper salts is one of the examples.

Types of Overheating-

The overheating failures are of two types:

(1) Sudden rupture – failure in a very short period due to highly excessive heat flux condition – Short Time Overheating;

(2) Creep rupture – long period failure associated with a small degree of excessive temperature build-up – „Long Time Overheating‟.

Some call the first type as „overheating failures‟ and the second one as „creep failures‟.

Sudden ruptures are common in water-wall tubes where water boiling phenomenon takes place at an inside interface. It is believed that this type of failure (ruptures) are responsible for the onset of „unstable film boiling‟ (also known as departure from nucleate boiling i.e. DNB).


13

Nucleate Boiling

Film Boiling (DNB)

When water is boiler in a tube having uniform heat flux (Rate of heat transfer) along its length under conditions that produce a state of dynamic equilibrium, various points along the tube will be in contact with sub-cooled water, boiling water, low-quality steam, high-quality steam and superheated steam. A temperature gradient between the tube wall and the fluid within the tube provided the driving force for heat transfer at any point. Figure below indicates the effects that different heat fluxes have on tube-wall temperature. In the region where sub cooled water contacts the tube (at left in Figure) the conductance of the fluid film is relatively high; therefore, a small temperature difference sustains heat transfer at all heat flux levels.

Effect of Heat Flux on Tube Wall Temperature

However the conductance of a vapor film in steam of low quality of film boiling, a large temperature difference between the tube wall and the bulk fluid is required


14

to sustain a high heat flux across the film. The effect of the onset of film boiling on tube-wall temperature appears as sharp breaks in the curves for moderate, high and very high heat fluxes in Figure above. With increasing heat flux, the onset of unstable film boiling (also known as departure from nucleate boiling, or DNB) occurs at lower steam qualities and tube-wall temperatures reach higher peak values before stable film boiling, which requires a lower temperature difference to sustain a given heat flux, is established.

At very high heat fluxes, DNB occurs at low steam quality, and the temperature difference between tube wall and bulk fluid at a point slightly downstream from DNB is very high. Under these conditions, tube failure theoretically can occur by melting of the tube wall, although in reality the tube will rupture because metal losses its strength (and thereby loses its ability to contain pressure) before it melts. In design of fossil fuel boilers and nuclear reactors, DNB is an important consideration, because heat flux can quickly exceed the failure point (burnout point) at local regions in a tube if the tube does not receive an adequate supply of incoming feed water.

Factors leading to DNB in Natural Circulation:

In a natural circulation boiler, the circulation occurs due to the density difference in riser and downcomers. The downcomers contain water, whereas the risers contain a mixture of water and steam.

Assuming the density in riser and downcomer as ρr and ρd respectively, the height from the drum level to centerline of bottom ring header, the head difference forcing the water to circulate through the circuit becomes

ΔP = H ρd g - H ρr g = H g (ρd – ρr)

Therefore the circulation is maximum when ρr is minimum; which means the boiler is

generating steam at BMCR condition. The flow created by this natural circulation drives

the steam bubbles up leading to nucleate boiling.

It also indicates the vulnerability of the natural circulation boiler during start-up condition

since there is less amount of steam in the riser in the beginning. If we try to ramp-up the

boiler quickly (which we do more often than not), the heat flux can exceed critical heat

flux leading to DNB and subsequent film boiling.

If H in the above equation goes down, then it is also going to affect the circulation. That

tells us the importance of drum level.


15

Various factors leading to DNB can be listed as:

Fast Ramp up during cold start

Low drum level

Obstruction due to foreign object fouling the tube

Formation of scale inside the tube

Flame shifting towards one side wall

In superheaters and reheaters, which normally operate at temperatures that are 28 to 83 degrees C (50 to 150 degrees F) higher than the temperature of the steam inside the tubes, heat transfer is controlled primarily by the conductance of fluid films at the inner and outer surfaces. Although higher heat fluxes require higher tube-wall temperatures to sustain heat transfer, deposits have a greater effect on tube-wall temperatures and therefore on overheating.

These two types of overheating failures (usually) result in two distinctly different types of fracture features i.e. „thick-lip rupture‟ (in case of slow overheating – creep overheating) and „thin-lip rupture‟ (in case of sudden …………rupture).

Long Term Overheating

Thick-lip ruptures in steam-generator tubes occur mainly by stress rupture as a result of prolonged overheating at a temperature slightly above the maximum safe working temperature, for the tube material. Fracture surfaces of such ruptures are rough (crystalline) in microscopic appearance and usually are oxidized or hot-gas corroded because of exposure to a high-temperature corrosive environment following the rupture. The microscopic direction of fracture is normal to the tube surface (flat-face fracture) and parallel to the tube axis. Thick-lip rupture may or may not be accompanied by sight swelling of the tube in the region adjacent to the rupture, and usually there is only small amount of reduction in tube wall thickness at the fracture. Examination of a polished, transverse reaction through the tube at the center of the fracture usually reveals extensive secondary transverse cracking adjacent to the main fracture and when etched, the section will exhibit intergranular separation.


16

Thin lip conventional short term overheating rapture

Short Term Overheating with thick lip

Thin-lip ruptures usually are transgranular tensile fractures occurring at metal temperatures from 650 to 870 C (1200 to 1600 F). These elevated temperature tensile fractures exhibit microscopic and microscopic features that are characteristic of the tube alloy and the temperature at which rupture occurred. A tensile fracture results from rapid overheating to a temperature considerably above the safe working temperature for the tube material and is accompanied by considerable swelling of the tube in the regions adjacent to the rupture that have been exposed to the highest temperatures. Sometimes, steam escaping at high velocity through the rupture will impose a reaction force on the tube sufficient to bend it internally. The higher and more uniform the degree of overheating, the greater in the likelihood of lateral bending.

Ruptures caused by rapid overheating exhibit obvious tub-wall thinning adjacent to the rupture, often to a knife-edge at the fracture surface. Thinning also occurs in areas of swelling adjacent to ruptures.

Preventive Measures:

Preventing overheating failure means to approach perfection in all aspects. Prevention of overheating needs maintenance of „all parameters‟ religiously. Not only this, it needs perfect erection/fabrication.

During erection/fabrication it has to be seen that there is no burner/flame distortion/disturbance.

During operation and maintenance it has to be seen that following are prevented:

(1) Undesirable flow pattern of flue gas,

(2) Undesirable water chemistry,

(3) Undesirable water level in boiler drum etc.

Close control of boiler temperature within the designed range is very much essential for prevention of failures due to overheating. Here it is important to know that an increase of 100oF reduces rupture life of most of the steels to about one-tenth of their original value, the stress being constant. Therefore the boiler tube that has been properly designed for a life of 30 years may fail after only 3 years if temperature conditions develop that increase the temperature by 100oF.

Since highly ideal conditions are needed to avoid overheating it is found that whenever there is slippage in any parameter, this failure occurs as material selected for boiler tubes is just sufficient to meet design parameters only.


17

Corrosion

Basically the corrosion mechanism can set in under the following two conditions (i.e. SCs for this DM).

(1) Presence of a material having inferior corrosion resistance than expected and

(2) Development of excessive corrosive condition than expected

In boiler engineering the corrosion of boiler tubes is classified as follows:

(a) Water/steam side corrosion (i.e. internal corrosion).

(b) Furnace side corrosion (i.e. external corrosion).

The water side issues are not limited to corrosion only. There can be other problems like scaling and carry-over, which subsequently leads to failure. We are therefore going to have a detailed discussion on the problems related faulty water chemistry.


18

Problems caused by faulty Water Chemistry:

Boiler is a vessel wherein water is heated to convert into steam, with the application of heat of a burning fuel and/or the hot exhaust gas from some process. The heat potential available in steam is one of the well-known sources of energy to do mechanical work in any steam engine. It is essential to maintain uninterrupted flow of steam to the engines and hence the boiler has to be necessarily fed with adequate supply of water to keep up a safe level in the drum. In order to ensure long life and trouble free operation of a boiler it is important to keep the boiler water chemistry under control. Depending upon the working pressure and temperature of boiler, the quality of Feed Water for the boiler is to be maintained. However, due to continuous evaporation occurring in boiler, the concentration of impurities in the boiler goes on increasing. Therefore, it is necessary to carry out chemical injection in the boiler for forming precipitates of impurities, which are to be drained out at regular intervals. Due to this regular draining or blow down the quantity of water in boiler system has got to be continuously made-up by feeding high purity water which is obtained by appropriate treatment of raw water.

Raw Water Sources:

Though the water, in nature, is available in plenty, it can not be taken for use in a boiler without a proper chemical treatment. The physical and chemical properties of natural water, which widely vary, depending upon the source and strata on which it flows, makes it unsuitable for direct use in a Boiler. The natural water picks up minerals and salts from the earthen layer, which gets into the solution. Water, therefore, is a mixer of composite mineral salts in dissolved condition, in varying proportions, composition and degree. It gets polluted further with multifarious organic and inorganic impurities, due to disposal of industrial and domestic wastes. Decayed vegetation and marine lives also contribute to water contamination. Not only the dissolved salts in water, but also the presence of the coarse and un-dissolvable substances in suspended form, constituting mainly of silt and clay matters, generally termed as turbidity, make it unsuitable for direct use, without any treatment. Natural water, found on earth, contains siliceous matters, in dissolved as well as in colloidal forms.

Impurities:

Water as a Solvent

Pure water (H20) is colorless, tasteless, and odorless. It is composed of hydrogen and oxygen. Because water becomes contaminated by the substances with which it comes into contact, it is not available for use in its pure state. To some degree, water can dissolve every naturally occurring substance on the earth. Because of this property, water has been termed a "universal solvent." Although beneficial to mankind, the solvency power of water can pose a major threat to industrial equipment. Corrosion reactions cause the slow dissolution of metals by water. Deposition reactions, which produce scale on heat transfer surfaces, represent a change in the solvency power of water as its temperature is varied. The control of corrosion and scale is a major focus of water treatment technology.


19

Necessity of Water Treatment for Boilers:

Waters used as boiler feed water, such as industrial water, underground water, river water and sea water; usually contain various substances such as suspended solids, dissolved solids and gases. The amounts of these substances vary largely depending on the sources of raw waters.

The use of such water without the proper treatment may result in problems, such as scale, corrosion and carryover, in boilers and the auxiliary equipments.

The Mechanism of Scaling and Corrosion:

Scaling:

Deposition is a major problem in the operation of steam generating equipment. The accumulation of material on boiler surfaces can cause overheating and/or corrosion. Both of these conditions frequently result in unscheduled downtime.

Deposits

Common feedwater contaminants that can form boiler deposits include calcium, magnesium, iron, copper, aluminum, silica, and (to a lesser extent) silt and oil. Most deposits can be classified as one of two types

scale that crystallized directly onto tube surfaces

sludge deposits that precipitated elsewhere and were transported to the metal surface by the flowing water

Scale is formed by salts that have limited solubility but are not totally insoluble in boiler water. These salts reach the deposit site in a soluble form and precipitate when concentrated by evaporation. The precipitates formed usually have a fairly homogeneous composition and crystal structure.

High heat transfer rates cause high evaporation rates, which concentrate the remaining water in the area of evaporation. A number of different scale-forming compounds can precipitate from the concentrated water. The nature of the scale formed depends on the chemical composition of the concentrated water. Normal deposit constituents are calcium, magnesium, silica, aluminum, iron, and (in some cases) sodium.


20

The exact combinations in which they exist vary from boiler to boiler, and from location to location within a boiler. Scale may form as calcium silicate in one boiler and as sodium iron silicate in another.

Compared to some other precipitation reactions, such as the formation of calcium phosphate, the crystallization of scale is a slow process. As a result, the crystals formed are well defined, and a hard, dense, and highly insulating material is formed on the tube metal. Some forms of scale are so tenacious that they resist any type of removal-mechanical or chemical.

Sludge is the accumulation of solids that precipitate in the bulk boiler water or enter the boiler as suspended solids. Sludge deposits can be hard, dense, and tenacious. When exposed to high heat levels (e.g., when a boiler is drained hot), sludge deposits are often baked in place. Sludge deposits hardened in this way can be as troublesome as scale.

Once deposition starts, particles present in the circulating water can become bound to the deposit. Intraparticle binding does not need to occur between every particle in a deposit mass. Some nonbound particles can be captured in a network of bound particles.

Binding is often a function of surface charge and loss of water of hydration. Iron oxide, which exists in many hydrated and oxide forms, is particularly prone to bonding. Some silicates will do the same, and many oil contaminants are notorious deposit binders, due to polymerization and degradation reactions.

Since these scale components have the small thermal conductivities as shown in

Table below, the scale adhesion on the heating surface remarkably reduces the

thermal efficiency of boiler.

Substance Thermal conductivity (kcal/m·h·°C)

Silica scale 0.2–0.4

Calcium carbonate scale 0.4–0.6

Calcium sulfate scale 0.5–2.0

Calcium phosphate scale 0.5–0.7

Iron oxide (hematite) scale 3–5

Iron oxide (magnetite) scale 1

Carbon steel 40–60

Copper 320–360

When the part covered with the scale is locally superheated, the mechanical

strength of the tube material is reduced and the bursting may occur eventually.

Most scales formed on the heating surface are generally composed of several

mixed substances.

The thermal conductivity of such mixed scale is approximately 1 to 2 kcal/m·h·°C.

Figure below shows a model of heating surface covered with scale and the inner

skin temperature of the tube under the scale, obtained from the following

equation:


21

t2 = t4 + (1/α + D/ k2) x Q

Where,

t2 = inner skin temperature of tube (°C)

t4 = boiler water temperature (bulk)(°C)

α = heat transfer coefficient of boiling surface

(kcal/ m2·h·°C)

D = scale thickness (m)

k2 = thermal conductivity of scale (kcal/ m·h·°C)

Q = heat flux (kcal/m2·h)

Figure in the left shows the

relationship between the scale

thickness and the inner skin

temperature of tube at each heat

flux when a scale with thermal

conductivity of 2 kcal/m·h·°C

adheres on the boiler tube inside

with the pressure of 10 kgf/cm2.

Figure in the right shows the

relationship between the

temperature and the

allowable tensile stresses of

carbon steels using as boiler

tubes.

When the temperature

exceeds 350° C, the

allowable stress starts to

decrease and reaches the 50

to 60% of the original value

at 450° C. Therefore, the

tube wall temperature

should be kept below 450°C

for the safe boiler operation

and the scale thickness on

the tube has to be

controlled thinner than 1 to

2 mm as shown in previous

Figure.

JIS STB 410 is equivalent of ASTM grade SA210

GR A1

JIS STB 340 is equivalent of ASTM grade SA 192

In addition to causing material damage by insulating the heat transfer path from the boiler flame to the water


22

They roughen the tube surface and increase the drag coefficient in the boiler circuit. Reduced circulation in a generating tube contributes to accelerated deposition, overheating, and premature steam-water separation.

Corrosion:

Corrosion is one of the main causes of reduced reliability in steam generating systems. It is estimated that problems due to boiler system corrosion cost industry billions of Rupees per year.

Many corrosion problems occur in the hottest areas of the boiler-the water wall, screen, and superheater tubes. Other common problem areas include deaerators, feedwater heaters, and economizers.

Methods of corrosion control vary depending upon the type of corrosion encountered. The most common causes of corrosion are dissolved gases (primarily oxygen and carbon dioxide), under-deposit attack, low pH, and attack of areas weakened by mechanical stress, leading to stress and fatigue cracking.

These conditions may be controlled through the following procedures:

maintenance of proper pH and alkalinity levels

control of oxygen and boiler feed water contamination

reduction of mechanical stresses

operation within design specifications, especially for temperature and pressure

proper precautions during start-up and shutdown

effective monitoring and control

Corrosion Tendencies of Boiler System Components

Most industrial boiler and feedwater systems are constructed of carbon steel. Many have copper alloy and/or stainless steel feedwater heaters and condensers. Some have stainless steel superheater elements.

Proper treatment of boiler feedwater effectively protects against corrosion of feedwater heaters, economizers, and deaerators. The ASME Consensus for Industrial Boilers specifies maximum levels of contaminants for corrosion and deposition control in boiler systems.

The consensus is that feedwater oxygen, iron, and copper content should be very low (e.g., less than 7 ppb oxygen, 20 ppb iron, and 15 ppb copper for a 900 psig boiler) and that pH should be maintained between 8.5 and 9.5 for system corrosion protection.


23

In order to minimize boiler system corrosion, an understanding of the operational requirements for all critical system components is necessary.

Deaerators

Deaerators are used to heat feedwater and reduce oxygen and other dissolved gases to acceptable levels. Corrosion fatigue at or near welds is a major problem in deaerators. Most corrosion fatigue cracking has been reported to be the result of mechanical factors, such as manufacturing procedures, poor welds, and lack of stress-relieved welds. Operational problems such as water/steam hammer can also be a factor.

Effective corrosion control requires the following practices:

regular monitoring of operation

minimization of stresses during start-up

maintenance of stable temperature and pressure levels

control of dissolved oxygen and pH in the feedwater

regular out-of-service inspection using established nondestructive techniques

Other forms of corrosive attack in deaerators include stress corrosion cracking of the stainless steel tray chamber, inlet spray valve spring cracking, corrosion of vent condensers due to oxygen pitting, and erosion of the impingement baffles near the steam inlet connection.

Economizers

Economizer corrosion control involves procedures similar to those employed for protecting feedwater heaters.

Economizers help to improve boiler efficiency by extracting heat from flue gases discharged from the fireside of a boiler. Economizers can be classified as nonsteaming or steaming. In a steaming economizer, 5-20% of the incoming feedwater becomes steam. Steaming economizers are particularly sensitive to deposition from feedwater contaminants and resultant under-deposit corrosion. Erosion at tube bends is also a problem in steaming economizers.

Oxygen pitting, caused by the presence of oxygen and temperature increase, is a major problem in economizers; therefore, it is necessary to maintain essentially oxygen-free water in these units. The inlet is subject to severe pitting, because it is often the first area after the deaerator to be exposed to increased heat. Whenever possible, tubes in this area should be inspected closely for evidence of corrosion.

Economizer heat transfer surfaces are subject to corrosion product buildup and deposition of incoming metal oxides. These deposits can slough off during operational load and chemical changes.

Corrosion can also occur on the gas side of the economizer due to contaminants in the flue gas, forming low-pH compounds. Generally, economizers are arranged for downward flow of gas and upward flow of water. Tubes that form the heating surface may be smooth or provided with extended surfaces.


24

Superheaters

Superheater corrosion problems are caused by a number of mechanical and chemical conditions. One major problem is the oxidation of superheater metal due to high gas temperatures, usually occurring during transition periods, such as start-up and shutdown. Deposits due to carryover can contribute to the problem. Resulting failures usually occur in the bottom loops-the hottest areas of the superheater tubes.

Oxygen pitting, particularly in the pendant loop area, is another major corrosion problem in superheaters. It is caused when water is exposed to oxygen during downtime. Close temperature control helps to minimize this problem. In addition, a nitrogen blanket and chemical oxygen scavenger can be used to maintain oxygen-free conditions during downtime.

Low-Pressure Steam and Hot Water Heating Systems

Hot water boilers heat and circulate water at approximately 200°F. Steam heating boilers are used to generate steam at low pressures, such as 15 psig. Generally, these two basic heating systems are treated as closed systems, because makeup requirements are usually very low.

High-temperature hot water boilers operate at pressures of up to 500 psig, although the usual range is 35-350 psig. System pressure must be maintained above the saturation pressure of the heated water to maintain a liquid state. The most common way to do this is to pressurize the system with nitrogen. Normally, the makeup is of good quality (e.g., deionized or sodium zeolite softened water). Chemical treatment consists of sodium sulfite (to scavenge the oxygen), pH adjustment, and a synthetic polymer dispersant to control possible iron deposition.

The major problem in low-pressure heating systems is corrosion caused by dissolved oxygen and low pH. These systems are usually treated with an inhibitor (such as molybdate or nitrite) or with an oxygen scavenger (such as sodium sulfite), along with a synthetic polymer for deposit control. Sufficient treatment must be fed to water added to make up for system losses, which usually occur as a result of circulating pump leakage. Generally, 200-400 ppm P-alkalinity is maintained in the water for effective control of pH. Inhibitor requirements vary depending on the system.

Types of Corrosion

Corrosion control techniques vary according to the type of corrosion encountered. Major methods of corrosion control include maintenance of the proper pH, control of oxygen, control of deposits, and reduction of stresses through design and operational practices.

Galvanic Corrosion

Galvanic corrosion occurs when a metal or alloy is electrically coupled to a different metal or alloy.

The most common type of galvanic corrosion in a boiler system is caused by the contact of dissimilar metals, such as iron and copper. These differential cells can also be formed when deposits are present. Galvanic corrosion can occur at welds due to stresses in heat-affected zones or the use of different alloys in the welds. Anything that results in a difference in electrical potential at discrete surface locations can cause a galvanic reaction. Causes include:


25

scratches in a metal surface

differential stresses in a metal

differences in temperature

conductive deposits

Pitting of boiler tube banks has been encountered due to metallic copper deposits. Such deposits may form during acid cleaning procedures if the procedures do not completely compensate for the amount of copper oxides in the deposits or if a copper removal step is not included. Dissolved copper may be plated out on freshly cleaned surfaces, establishing anodic corrosion areas and forming pits, which are very similar to oxygen pits in form and appearance. This process is illustrated by the following reactions involving hydrochloric acid as the cleaning solvent.

Magnetite is dissolved and yields an acid solution containing both ferrous (Fe²+) and ferric (Fe³+) chlorides (ferric chlorides are very corrosive to steel and copper)

Fe3O4 + 8HCl FeCl2 + 2FeCl3 + 4H2O

Metallic or elemental copper in boiler deposits is dissolved in the hydrochloric acid solution by the following reaction:

FeCl3 + Cu CuCl + FeCl2

Once cuprous chloride is in solution, it is immediately redeposited as metallic copper on the steel surface according to the following reaction:

2CuCl + Fe FeCl2 + 2Cu0

Thus, hydrochloric acid cleaning can cause galvanic corrosion unless the copper is prevented from plating on the steel surface. A complexing agent is added to prevent the copper from redepositing. The following chemical reaction results:

FeCl3 + Cu + Complexing

Agent FeCl2 + CuCl

This can take place as a separate step or during acid cleaning. Both iron and the copper are removed from the boiler, and the boiler surfaces can then be passivated.

In most cases, the copper is localized in certain tube banks and causes random pitting. When deposits contain large quantities of copper oxide or metallic copper, special precautions are required to prevent the plating out of copper during cleaning operations.

Caustic Corrosion

Concentration of caustic (NaOH) can occur either as a result of steam blanketing (which allows salts to concentrate on boiler metal surfaces) or by localized boiling beneath porous deposits on tube surfaces.

Caustic corrosion (gouging) occurs when caustic is concentrated and dissolves the protective magnetite (Fe3O4) layer. Iron, in contact with the boiler water, forms magnetite and the protective layer is continuously restored. However, as long as a high caustic concentration exists, the magnetite is constantly dissolved, causing a loss of base metal and eventual failure.

Steam blanketing is a condition that occurs when a steam layer forms between the boiler water and the tube wall. Under this condition, insufficient water reaches


26

the tube surface for efficient heat transfer. The water that does reach the overheated boiler wall is rapidly vaporized, leaving behind a concentrated caustic solution, which is corrosive.

Porous metal oxide deposits also permit the development of high boiler water concentrations. Water flows into the deposit and heat applied to the tube causes the water to evaporate, leaving a very concentrated solution. Again, corrosion may occur.

Caustic attack creates irregular patterns, often referred to as gouges. Deposition may or may not be found in the affected area.

Boiler feedwater systems using demineralized or evaporated makeup or pure condensate may be protected from caustic attack through coordinated phosphate/pH control. Phosphate buffers the boiler water, reducing the chance of large pH changes due to the development of high caustic concentrations. Excess caustic combines with disodium phosphate and forms trisodium phosphate. Sufficient disodium phosphate must be available to combine with all of the free caustic in order to form trisodium phosphate.

Disodium phosphate neutralizes caustic by the following reaction:

Na2HPO4 + NaOH ↔ Na3PO4 + H2O

This results in the prevention of caustic buildup beneath deposits or within a crevice where leakage is occurring. Caustic corrosion and caustic embrittlement does not occur, because high caustic concentrations do not develop.

Different forms of phosphate consume or add caustic as the phosphate shifts to the proper form. For example, addition of monosodium phosphate consumes caustic as it reacts with caustic to form disodium phosphate in the boiler water according to the following reaction:

NaH2PO4 + NaOH ↔ Na2HPO4 + H2O

Conversely, addition of trisodium phosphate adds caustic, increasing boiler water pH:

Na3PO4 + H2O ↔ Na2HPO4 + NaOH


27

Control is achieved through feed of the proper type of phosphate to either raise or lower the pH while maintaining the proper phosphate level. Increasing blowdown lowers both phosphate and pH. Therefore, various combinations and feed rates of phosphate, blowdown adjustment, and caustic addition are used to maintain proper phosphate/pH levels.

Elevated temperatures at the boiler tube wall or deposits can result in some precipitation of phosphate. This effect, termed "phosphate hideout," usually occurs when loads increase. When the load is reduced, phosphate reappears.

Clean boiler water surfaces reduce potential concentration sites for caustic. Deposit control treatment programs, such as those based on chelants and synthetic polymers, can help provide clean surfaces.

Where steam blanketing is occurring, corrosion can take place even without the presence of caustic, due to the steam/magnetite reaction and the dissolution of magnetite. In such cases, operational changes or design modifications may be necessary to eliminate the cause of the problem.

Acidic Corrosion

Low makeup or feedwater pH can cause serious acid attack on metal surfaces in the preboiler and boiler system. Even if the original makeup or feedwater pH is not low, feedwater can become acidic from contamination of the system. Common causes include the following:

improper operation or control of demineralizer cation units

process contamination of condensate (e.g., sugar contamination in food processing plants)

cooling water contamination from condensers

Acid corrosion can also be caused by chemical cleaning operations. Overheating of the cleaning solution can cause breakdown of the inhibitor used, excessive exposure of metal to cleaning agent, and high cleaning agent concentration. Failure to neutralize acid solvents completely before start-up has also caused problems.


28

The acid chloride corrosion can take place through the following reaction under deposits or steam blanketing at low pH conditions:

MgCl2 + H2O → MgO + 2HCl

Fe3O4 +HCl → FeCl2 + FeCl3 +H2O

Fe + 2HCl → FeCl2 + 2H

In a boiler and feed water system, acidic attack can take the form of general thinning, or it can be localized at areas of high stress such as drum baffles, "U" bolts, acorn nuts, and tube ends.

Oxygen Attack

Without proper mechanical and chemical deaeration, oxygen in the feedwater will enter the boiler. Much is flashed off with the steam; the remainder can attack boiler metal. The point of attack varies with boiler design and feedwater distribution. Pitting is frequently visible in the feedwater distribution holes, at the steam drum waterline, and in downcomer tubes.

Oxygen is highly corrosive when present in hot water. Even small concentrations can cause serious problems. Because pits can penetrate deep into the metal, oxygen corrosion can result in rapid failure of feedwater lines, economizers, boiler tubes, and condensate lines. Additionally, iron oxide generated by the corrosion can produce iron deposits in the boiler.

Oxygen corrosion may be highly localized or may cover an extensive area. It is identified by well defined pits or a very pockmarked surface. The pits vary in shape, but are characterized by sharp edges at the surface. Active oxygen pits are distinguished by a reddish brown oxide cap (tubercle). Removal of this cap exposes black iron oxide within the pit.

Oxygen attack is an electrochemical process that can be described by the following reactions:

Anode:

Fe Fe2+ + 2e¯

Cathode:

½O2 + H2O + 2e¯ 2OH¯

Overall:

Fe + ½O2 + H2O Fe(OH)2


29

The influence of temperature is particularly important in feedwater heaters and economizers. A temperature rise provides enough additional energy to accelerate reactions at the metal surfaces, resulting in rapid and severe corrosion.

At 60°F and atmospheric pressure, the solubility of oxygen in water is approximately 8 ppm. Efficient mechanical deaeration reduces dissolved oxygen to 7 ppb or less. For complete protection from oxygen corrosion, a chemical scavenger is required following mechanical deaeration.

Major sources of oxygen in an operating system include poor deaerator operation, in-leakage of air on the suction side of pumps, the breathing action of receiving tanks, and leakage of undeaerated water used for pump seals.

The acceptable dissolved oxygen level for any system depends on many factors, such as feedwater temperature, pH, flow rate, dissolved solids content, and the metallurgy and physical condition of the system. Based on experience in thousands of systems, 3-10 ppb of feedwater oxygen is not significantly damaging to economizers. This is reflected in industry guidelines.

the ASME consensus is less than 7 ppb (ASME recommends chemical scavenging to "essentially zero" ppb)

TAPPI engineering guidelines are less than 7 ppb

EPRI fossil plant guidelines are less than 5 ppb dissolved oxygen

Mechanical Conditions Affecting Corrosion

Many corrosion problems are the result of mechanical and operational problems. The following practices help to minimize these corrosion problems:

election of corrosion-resistant metals

reduction of mechanical stress where possible (e.g., use of proper welding procedures and stress-relieving welds)

minimization of thermal and mechanical stresses during operation

operation within design load specifications, without over-firing, along with proper start-up and shutdown procedures

maintenance of clean systems, including the use of high-purity feedwater, effective and closely controlled chemical treatment, and acid cleaning when required

Where boiler tubes fail as a result of caustic embrittlement, circumferential cracking can be seen. In other components, cracks follow the lines of greatest stress. A microscopic examination of a properly prepared section of embrittled metal shows a characteristic pattern, with cracking progressing along defined paths or grain boundaries in the crystal structure of the metal. The cracks do not penetrate the crystals themselves, but travel between them; therefore, the term "intercrystalline cracking" is used.

Good engineering practice dictates that the boiler water be evaluated for embrittling characteristics. An embrittlement detector (described in Chapter 14) is used for this purpose.

If a boiler water possesses embrittling characteristics, steps must be taken to prevent attack of the boiler metal. Sodium nitrate is a standard treatment for inhibiting embrittlement in lower-pressure boiler systems. The inhibition of embrittlement requires a definite ratio of nitrate to the caustic alkalinity present in the boiler water. In higher-pressure boiler systems, where demineralized


30

makeup water is used, embrittling characteristics in boiler water can be prevented by the use of coordinated phosphate/pH treatment control, described previously under "Caustic Corrosion." This method prevents high concentrations of free sodium hydroxide from forming in the boiler, eliminating embrittling tendencies.

Embrittlement:

Hydrogen damage and Graphitization embrittlement mechanism cause metallurgical changes within the metal of the boiler tubes and affect their ability to sustain service loads (for which they are designed and used). These mechanisms make normally satisfactory tube material susceptible to brittle fracture.

Hydrogen Damage:

The source (SOC) for this type of damage is existence of improper water chemistry and root cause (RC) is either low or high pH. The low/ high pH damage is discussed in detail earlier.

Fe + 2NaOH → Na2FeO2 + 2H

Fe + 2HCl FeCl2 + 2H

4H+ + Fe3C → CH4 + 3Fe

The Methane gas generated collects in grain boundaries and form fissures as pressure builds up which eventually grow.

From fractographic examination point of view, it is important to note that cracking caused by hydrogen damage often resembles stress-corrosion-cracking, except that hydrogen damage failures may exhibit little or no crack branching. Mostly hydrogen damage is manifested by discontinuous intergranular cracking, often accompanied by decarburization.


31

The tubes damaged by this mechanism often rupture in a manner known as Window Fracture, in which a portion of the tube wall gets detached.

If the mechanism has not caused decarburization and grain boundary cracking in a boiler tube, its original ductility can be restored by simple low temperature backing treatment.

Maintaining proper pH and phosphate is a preventive measure against this type of BTF.

Graphitization:

Graphitization i.e. the phenomenon of decomposition of perlite into ferrite and graphite can embrittle the carbon and low-alloy steels (i.e. those used for boiler tubes) especially when the graphite particles form along a continuous zone.

Because graphitization involves prolonged heating (several thousand hours) at moderate temperature (exposure between 425o to 550oC) it seldom occurs in waterwall tubes and is very likely to occur in low temperature superheater (LTSH) and economizer assemblies.

Comparatively embrittlement through graphitization is found more common in carbon and ½% Mo type steel which are generally used in LTSH assemblies.

Caustic Embrittlement

Caustic embrittlement (caustic stress corrosion cracking), or intercrystalline cracking, has long been recognized as a serious form of boiler metal failure. Because chemical attack of the metal is normally undetectable, failure occurs suddenly-often with catastrophic results.

For caustic embrittlement to occur, three conditions must exist:

the boiler metal must have a high level of stress

a mechanism for the concentration of boiler water must be present

the boiler water must have embrittlement-producing characteristics

Where boiler tubes fail as a result of caustic embrittlement, circumferential cracking can be seen. In other components, cracks follow the lines of greatest stress. A microscopic examination of a properly prepared section of embrittled metal shows a characteristic pattern, with cracking progressing along defined paths or grain boundaries in the crystal structure of the metal. The cracks do not penetrate the crystals themselves, but travel between them; therefore, the term "intercrystalline cracking" is used.


32

Good engineering practice dictates that the boiler water be evaluated for embrittling characteristics. An embrittlement detector (described in Chapter 14) is used for this purpose.

If a boiler water possesses embrittling characteristics, steps must be taken to prevent attack of the boiler metal. Sodium nitrate is a standard treatment for inhibiting embrittlement in lower-pressure boiler systems. The inhibition of embrittlement requires a definite ratio of nitrate to the caustic alkalinity present in the boiler water. In higher-pressure boiler systems, where demineralized makeup water is used, embrittling characteristics in boiler water can be prevented by the use of coordinated phosphate/pH treatment control, described previously under "Caustic Corrosion." This method prevents high concentrations of free sodium hydroxide from forming in the boiler, eliminating embrittling tendencies.

Stress Corrosion Cracking (SCC)

Stress corrosion cracking (SCC) is the growth of cracks in a corrosive environment. It can lead to unexpected sudden failure of normally ductile metals subjected to a tensile stress, especially at elevated temperatures. SCC is highly chemically specific in that certain alloys are likely to undergo cracking only when exposed to a small number of chemical environments. The chemical environment that causes stress corrosion cracking for a given alloy is often one which is otherwise only mildly corrosive to that metal. Hence, metal parts with severe SCC can appear bright and shiny, while being filled with microscopic cracks. This factor makes it common for stress corrosion cracking to go undetected prior to failure. Stress corrosion cracking often progresses rapidly, and is more common among alloys than pure metals. The specific environment is of crucial importance, and only very small concentrations of certain highly active chemicals are needed to produce catastrophic cracking, often leading to devastating and unexpected failure.

SCC in boiler pressure part is most commonly is associated with austenitic (stainless steel) super heater materials and can lead to either trans-granular or inter-granular crack propagation in the tube wall. It occurs where a combination of


33

high-tensile stresses and a corrosive fluid are present. The damage results from cracks that propagate from the ID. The source of corrosive fluid may be carryover into the super heater from the steam drum or from contamination during boiler acid cleaning if the super heater is not properly protected. Features and cause of SCC

More often occurs in austenitic stainless steels Typical locations are those with potential for highest concentration of

contaminants, such as bends and low spots in straight tubing where condensate can form during shutdown.

High stress locations, such as bends, welds, tube attachments, supports or spacers, and near welds where a change of thickness occurs are susceptible

Carryover of chlorides or sulfates from the chemical cleaning of water walls Boiler water carryover Volatile carryover of sulfur containing compounds Introduction of high levels of caustic from de-super heating or attemperator

spray Condenser cooling water constituents from a condenser leak Ingress of flue gas environment into tube through primary failure, especially

in RH when vacuum is drawn

Flow Accelerated Corrosion (FAC)

FAC is a process whereby the normally protective magnetite (Fe3O4) layer on

carbon steel dissolves in a stream of flowing water (single phase FAC) or wet steam (two phase FAC). Historically, the terms “Flow assisted corrosion, Flow induced corrosion, and Erosion-corrosion” have been used to define FAC.

This process reduces or eliminates the oxide layer and leads to a rapid removal of the base material until the pipeline bursts. The FAC process can be as high as 3 mm/year.

FAC Mechanism

The FAC mechanism consists of iron dissolution in flowing water and can be described as the action of four simultaneous reactions:

Air


34

FIGURE-1: SCHEMATIC OF FAC MECHANISM

Oxidation of iron to soluble ferrous ions and magnetite at the internal interface between the steel and the oxide.

Fe + 2H2O = Fe2+ + 2OH- + H2 3Fe + 4H2O = Fe3O4 + 4H2

Diffusion of soluble species (iron and hydrogen) across the porous oxide and diffusion of hydrogen through the carbon steel.

Dissolution and reduction of magnetite at the external interface between the oxide and the water.

2Fe3O4 + N2H4 + 12H+ = 6Fe2++ 8H2O + N2

Transfer of soluble iron species towards the flowing water and transfer of hydrogen towards the air after diffusion through the steel.

The rate of metal loss depends on a number of parameters including the feedwater chemistry, the material composition, other materials in feedwater system, and the fluid hydrodynamics.

At low velocities, the flow is laminar and essentially parallel to the surface of the metal or to the adjacent streamlines. The velocity varies from essentially zero near to the oxide/water surface to a maximum at the centerline of the pressure vessel/tube. The growth of Fe3O4 at the oxide/steel interface matches the

dissolution. At higher velocities, the action of the friction between the water and the oxide induces irregular fluctuating radial and axial velocity components with flow. The fluid is mixed in a random manner and becomes turbulent. Thus the growth of Fe

3O

4 cannot match the flow-accelerated dissolution, exfoliation and

spallation and the oxide thickness reduces and becomes less protective.

The following factors influence FAC in fossil & HRSG plants and the overall solubility of these oxides.

Hydrodynamics- Velocity, geometry, temperature, and mass transfer

Water Chemistry (feedwater in conventional and HRSG plants, and LP evaporator in HRSG)- pH, ORP, oxygen, and reducing agent

Component material composition- Carbon steel, chromium, copper, and molybdenum.

ORP (Oxidation-reduction potential) is the most important factor for single-phase FAC. It is measured as voltage with reference to Ag/AgCl (sat. KCl) reference electrode. ORP is sensitive to the materials of construction, temperature, pH, partial pressure of oxygen in the flowing water, mass transport properties and flow rates. Thus ORP cannot be compared from unit to unit. ORP control the surface oxide that forms in feedwater or evaporator water and as it becomes more reducing the greater is the possibility for FAC. Changing to AVT(O) or OT essentially reduces the possibility of dissolution into the flowing water to very low


35

values, even in areas where FAC was severe with AVT(R). Therefore all ferrous feedwater systems should be operated on AVT(O) or OT.

pH of water is the second most important factor as it also affects the solubility of the surface Fe3O4. Generally a higher pH will reduce the amount of corrosion and

FAC.

FIGURE-2: CORROSION PRODUCT RELEASE AS A FUNCTION OF pH

Temperature influences several of the fluid properties: the pH of water or wet steam, the solubility of the oxide layer, the rates of the oxidation and reduction reactions, and the variables related to mass transfer. Solubility of magnetite rises

with increasing temperature then decreases with a steep drop to 3000C. FAC tends

to peak at temperatures in the range of 150-1800C.

FIGURE-3: SOLUBILITY OF MAGNETITE AS A FUNCTION OF TEMPERATURE AT VARIOUS AMMONIA CONCENTRATIONS


36

Velocity- FAC strongly depends on flow velocity. This is not simply determined by the bulk fluid velocity but also by the factors which influence the local velocity i.e., surface geometry, flow path geometry and turbulence.

Mass Transfer is the process of transporting material (essentially magnetite) from the surface to the bulk of the flowing water or water-steam flow. The local mass transfer coefficient depends in a complex manner on fluid velocity, fluid viscosity, flow geometry, pipe/tube surface roughness, steam quality and void fraction (for two phase flow) and temperature. Mass transfer is usually described by the dimensionless parameters: Reynolds, Schmidt and Sherwood numbers.

Geometry is the factor which locates where FAC will occur. Certain geometries affect mass transfer due to changes in local velocity and turbulence. FAC does not often occur in straight pipes or tubes, but is more often encountered at points of hydrodynamic disturbance. These include elbows, tight bends, reducer tees, locations downstream of flow control orifices and valves, and even fabrication discontinuities. The geometric enhancement of these features generally increases turbulence which increases the mass transfer coefficients.

Component material composition is important because even trace amounts of chromium, copper, and molybdenum can significantly reduce the solubility of magnetite and thus of FAC. Amounts of chromium as low as 0.1% will significantly reduce FAC. Often 1.25% Cr alloys are used for replacement of FAC damaged areas. These alloys are also used in HRSG evaporator circuits susceptible to FAC.

There are three possible choices for feed water treatment viz. AVT (R), AVT (O), and OT. The treatment chosen should match the unit, unit metallurgy particularly feed water, cooling water and possible contaminant ingress, but the aim of all the treatments should be to prevent boiler tube failures and provide protection to the steam turbine.

For all ferrous feed water system the generation and transport of corrosion products like magnetite, hematite, and ferric oxide hydrate occurs mainly due to corrosion and flow accelerated corrosion (FAC) of low pressure and high pressure feed water heaters, deaerators, economizer inlet tubing and piping, feed water piping and drain lines.

For mixed metallurgy system the generation and transport of corrosion products like cuprous and cupric oxides occur mainly due to the corrosion of low pressure and high pressure feed water heater tubes.

C. Carryover

Carryover in low pressure boiler Steam generated in boilers essentially includes no

dissolved solids, however, the dissolved solids in boiler water sometimes transfer

to the steam due to various causes.

This phenomenon is called “Carryover.” The following factors are the main causes

of carryover, and if the “Priming” or “Foaming” occurs, it accelerates the

carryover.


37

1. Factors related to the boiler water quality control: Excess concentration of the boiler water, the contamination of boiler water with oils and fats, the dissolution of silica to steam, etc.

2. Factors related to the operating control of boiler: Operation at the high water level, the rapid fluctuation of heat load, etc.

3. Factors related to the mechanical structure of boiler: Poor condition of the water and steam separator, etc.

The priming is an occurrence of abnormal violent evaporation caused by a rapid

increase of the heat load and so on. In the result, a large amount of boiler water

droplets and foams from the boiler water transfer to the steam line together with

the steam.

The foaming is a phenomenon which a large amount of foams are produced on the

boiler water surface by the water contamination with fats and oils or by an excess

concentration of dissolved solids in the boiler water. In that case, the bubbles

containing the dissolved solids transfer to the steam line.

As most of low pressure boilers have no superheater and steam turbine, the

problem due to carryover is not serious. However, the carryover leads to a

deterioration of the product quality when the steam directly contacts with the

products.

Fire Side Corrosion:

Two types of fire side corrosion are found to cause BTFs.

(a) High temperature fire side corrosion and,

(b) Low temperature fire side corrosion.

Coal Ash Corrosion

Corrosion with coal ash is not very common in our country. During combustion of coal, the minerals in the burning coal are exposed to high temperatures and to the strongly reducing effects of generated gases, such as carbon monoxides and hydrogen. Aluminum, Iron, Potassium, Sodium and Sulfur compounds are partly decomposed, releasing volatile alkali compounds and sulfur oxides (predominantly SO2, plus small amounts of SO3). The remaining portion of the mineral matter reacts to form glassy particles known as fly ash.

Coal-ash corrosion starts with the deposition of fly ash on surfaces that operate predominantly at temperatures from 540 to 705 C (about 1000 to 1300 F) – mainly, surfaces of super heater and reheater tubes. These deposits may be loose and powdery, or may be sintered or slag-type masses that are more adherent. Over an extended period of time, volatile alkali and sulfur compounds condense on the


38

fly ash and react with it to form complex alkali sulfates such as KAL (SO4)2 and Na3Fe(SO4)3 at the boundary between the metal and the deposit. The reactions that produce alkali sulfates are believed to depend in part on the catalytic oxidation of SO2 to SO3 in the outer layers of the fly-ash deposit. The exact chemical reactions between the tube metal and the alkali sulfates are not well defined; nevertheless, certain characteristics of coal-ash corrosion are known:

1) Rapid attack occurs at temperatures between the melting temperature of the sulfate mixture and the limit of thermal stability for the mixture.

2) Corrosion rate is a nonlinear function of metal temperature, being highest at temperatures from 675 to 730 C (about 1250 to 1350 F).

3) Corrosion is almost always associated with sintered or slag-type deposits.

4) The deposit consists of three distinct layers. The porous, outermost layer comprises the bulk of the deposit and is composed essentially of the same compounds as those found in fly ash. The innermost layer is a thin glassy substance composed primarily of corrosion products of iron. The middle layer, called the white layer, is whitish or yellowish in colour, often is fused, and is largely water-soluble, producing an acid solution.

5) Coal-ash corrosion can occur with any bituminous coal, but is more likely when the coal contains more than 3.5% sulfur and 0.25% chlorine.

6) None of the common tube materials is immune from attack, although the 18-8 austenitic stainless steels corrode at slower rates than lower-alloy grades.

Particles of fly ash deposit on super heater and reheater tubes in a characteristic pattern in relation to the direction of flue gas flow. The tube surfaces are corroded most heavily beneath the thickest portions of the deposit. When deposits are removed, shallow macro pitting can be seen. Eventually, the tube wall becomes thinned to the point where the material can no longer withstand the pressure within the tube, and the tube ruptures, as in the example that follows.

Oil Ash Corrosion

Fuel oil is used not as a main fuel but for many other reasons in power plant boilers. Undesirable oil ash can cause corrosion damage. During combustion of fuel oils, organic compounds (including those containing vanadium or sulfur) decompose and react with oxygen. The resulting volatile oxides are carried along in the flue gases. Sodium, which usually is present in the oil as a chloride, reacts with the sulfur oxides to form sulfates. Initially, vanadium pentoxide (V2O5) condenses as a semi fluid alag on furnace walls, boiler tubes and super heater tubes – in fact, virtually anywhere in the high-temperature region of the boiler. Sodium oxide reacts with the vanadium pentoxide to form complex compounds, especially vanddates (nNa2O.V2O5) and vanadylvanadates (nNa2O.V2O4.mV2O5). These complex compounds, some of which have melting temperatures as low as 249 C (480 F), not only foul tube surfaces but also actively corrode them.


39

Slag of equilibrium thickness (0.12 to 0.25 in.) has developed in experimental furnaces within periods of time as short as 100 hr. Slag insulates the tubes, resulting in an increase in the temperature of the slag, which in turn increases the rate of corrosion and also promotes further deposition of ash. Thicker slag deposits generally lead to greater corrosion, because slag temperatures are higher and more of the corrodent is present to react with tube materials. However, higher slag temperatures also make the slag more fluid so that it will flow more readily on vertical surfaces. Consequently, slag generally builds up in corners and on horizontal surfaces, such as at the bases of water walls and around tube supports in the super heater.

Oil-ash corrosion affects all the common structural alloys. The mechanism is believed to be catalytic oxidation of the metal by reaction with vanadium pentoxide or with the complex vanadates or vanadylvanadetes. Even highly corrosion-resistant materials such as 60 Cr-40Ni and 50Cr-50Ni cast alloys, which sometimes are used for super heater-tube supports, are not immune. In addition, non-metallic refractory materials used for furnace linings are attacked by vanadium slag; the mechanism of this attack appears to be a dissolution or slagging type of attack rather than the direct chemical attack that characteristically occurs with metals.

Fireside corrosion is not a common in our country as the probability of presence of vanadium is largely eliminated by abandoning the use of vanadium bearing steels. Vanadium (in steel or oil) in presence of sodium (in ash) can form undesirable complex compounds such as vanadates (n.Na20.V2O5) and vanadylvandates (nNa20.V2O4.mV2O5) which if melts, becomes excessively corrosive.

Low Temperature Corrosion

Low temperature fireside corrosive was common earlier but is rarely found now a days. In low-temperature zones of flue-gas passages, corrosion is caused chiefly by condensed water vapour containing dissolved SO3 and CO2. The dew point of sulfuric acid, which is the most active corrodent, ordinarily ranges from 120 to 150 C (about 250 to 300 F) for SO3 concentrations of 15 to 30 ppm, which are common for coal-fired boilers. The dew point of the acidic vapors depends on (a) the amount of moisture in the fuel and in the combustion air, (b) the quantity of excess air, (c) the amount of hydrogen in the fuel, and (d) the amount of steam used for soot blowing – all of which influence the amount of water vapour in the flue gas.


40

Cold End Acid Dew Point Corrosion

Condensation of acidic vapours is most prevalent in air heaters, precipitators, stack coolers, flues and stacks, and near the inlets of economizers in units without feed water heaters. Factors that increase the likelihood of acid condensation include; (a) low flue-gas flow, such as occurs during start-up and during periods of low-load operation; (b) excessively low flue-gas-exit temperatures during normal operation; (c) too great an amount of excess air of high humidity; and (d) very low atmospheric temperatures.

Low-alloy steels, particularly those containing copper, have been used successfully in economizers and air heaters that are prone to low temperature corrosion. An additional advantage of these alloys is their good resistance to pitting caused by excess oxygen in boiler water.

The deposits that form as a result of low-temperature corrosion may contain corrosion products, fly ash and the products of chemical reaction of the condensed acid with fly-ash constituents. Many of the compounds in these deposits are water soluble and thus can be removed conveniently by washing the affected areas with water. However, the deposits sometimes become difficult to remove, especially when they are allowed to accumulate until they completely plug passages or when they contain insoluble compounds such as calcium sulfate.

Low-temperature corrosion is encountered more often in oil-fired units than in coal-fired units, because the vanadium in oil-ash deposits is a catalyst for oxidation of SO2 to SO3, and because oil firing produces loss ash than coal firing. In coal-fired units, a substantial portion of the sulfur oxides is absorbed by fly-ash deposits in the high-temperature regions, where the sulfur oxides participate in coal-ash corrosion. Further-more coal ash, which is composed chiefly of basic compounds, partly neutralizers the condensed acids when if deposits on moist surfaces in low-temperature regions. Oil ash, which is mainly acidic, is incapable of neutralizing condensed acids.

Erosion:

It is believed that blowing/flowing/jetting/impinging steam or water is not erosive (unless it velocity becomes extremely high) and does not cause erosion of boiler tubes. It is fly ash particles entrained in steam (or water) jet/flow which is responsible for erosion damage.


41

Thus internal erosion of tubes by flowing media is ruled out (at least no such case is reported for boiler tubes) and it is only external material wastage through which the erosion damage causes tube failure. Therefore this subject is also referred to as “Tube External Wastage”.

Logically development of following three conditions (UCs) can set in erosion damage:

1. Poor erosion resistance of material, or

2. More abrasive/erosive constitutes of medium, or

3. Otherwise normal abrasive medium becoming damaging on account of undesirable flow velocity/pattern.

The first condition (UC) is out of consideration since the tube materials as it is have low erosion resistance and it will be a remote probability to have slippage of a tube with erosion resistance inferior than what is being used.

Various factors (RCs) develop the other two conditions (UCs) and cause BTFs. Contribution of erosion damage in BTFs Is significantly high in our country particularly because of the abrasive nature and amount of ash content of the coal available to the power plants. So many BTFs on account of external erosion occur in our country that the field engineering service section of M/s B.H.E.L. (the only boiler manufacturer of thermal power plants of India) keeps on identifying erosion prone areas on the basis of experience and data study and suggests necessary possible measures for preventing them. The external wastage through erosion may be subdivided as:

1. Damage through flue gas i.e. „Flue gas Erosion‟.

2. Damage through steam i.e. „steam Jet Erosion‟.

Flue Gas Erosion:

The root causes (RCs) for the Flue Gas Erosion failure are:

1. Non uniform gas flow distribution (causing turbulence)

2. Plugging in of gas path (resulting in higher gas velocities).

3. Excessive amount of abrasive ash in coal (high silica and Alumina).

4. Excessive dust loading on account of poor coal (higher ash lesser calorific value – higher fuel feeding to meet generation).

5. Undesirable excess air (higher velocity of flue gas then the safe limit)

6. Misalignment of tube components (preferential excessive erosion).

Abrasive nature and higher amount of ash content of coal are damaging but even ash with normal characteristics causes damage if factors develop such that gas velocity exceeds safe limit. This root cause is found predominant in the flue gas erosion type BTFs.


42

Partial fouling of gas passages in tube bank (either by excessive ash deposit or by misalignment of a tube component) can lead to erosion by forcing the flue gases to flow through smaller passages at higher velocity.

It is well established that higher particle concentration probabilities are much higher in second pass, particularly in low temperature super heater (LTSH) and economizer assemblies. This area is considered as most erosion failure prone.

Preventive Measures: There are two ways. First is to minimize damaging effect of nonuniform flow of flue gas by using suitable deflecting baffles or protective/sacrificial shields (cassette baffles), Second one is to adopt the practice of periodical thickness survey of erosion prone areas to detect tubes with critically low thickness and to arrange their replacement (before they fail).

Steam Erosion:

The RCs for steam erosion damage are:

1. Soot blower/deslagger erosion (jammed leaking blowers/deslaggers).

2. Steam cutting (impingement by escaping/leaking steam jet).

At initial stage of puncture the escaping steam jet is usually erosive enough to damage adjacent tubes on which it happens to be impinging. This type of steam cutting many a time causes multiple failures and proves to be highly damaging through mutual impingement.

Preventive measures: Operation and maintenance of the soot blowers is essential to avoid soot-blower erosion failures. With the knowledge of occurrence of puncture, shut down of boiler should follow immediately (many a time power load condition does not permit shut down and continued operation leads to multiple failure on account of steam cutting).

Mechanical Fatigue and Thermal Fatigue:

Mechanical fatigue solely arising from vibratory conditions or changes in mechanical stress cycle and leading to a BTF are not known and found in literature. Usually non availability of sufficient flexibility for thermal expansion and contraction for a component or system leads to fatigue failure in the component itself or in the fouling component (which restricts flexibility of the other components) depending on physical arrangement and assembly between the


43

fouling components. Even otherwise the expansion at operating temperature and contraction during shut down can cause cracking if the system is too rigid.

Earlier (when there was not enough design experience with the boiler manufacturers like M/S B.H.E.L.) crack formation in header to nipple welds was very common. Now a days compensating bends are provided before vertically suspended assembly tubes ending into header. Similarly faulty assembly design of „fixing arrangement‟ of thermal liner piping in desuperheater pipe (in earlier manufactured boilers with 200 MW units) used to cause failure of fixing bolts and supporting pipes by fatigue damage mechanism. Now a days a modified „fixing arrangement‟ is being used.

Welding two tube components with non-matching welding electrode (from thermal expansion and contraction point of view) can cause development of a fatigue crack at the weld joint.

It may be noted that all the above mentioned cases are related to design of assembly of fabrication. Thus for mechanical fatigue damage mechanism (DM) source (SOC) is at the design stage.

Regarding thermal fatigue failures i.e. failures due to starts/stops or variation in loading (generation level) condition, it is the opinion of the power plant engineers that this factor has to be taken care of right at the material selection stage (i.e. design stage) and the boiler tubes should have enough fatigue resistance to meet all these conditions.

Reliability Issues in HRSGs

The origin of problem:

The two paramount concerns of most purchasers of CCGT installations are

low installed cost and high fuel efficiency. Compared to highly rated large

GTs, HRSGs are perceived as low-risk equipment. Selection of the HRSG is

usually from the lowest bidder. To succeed in the very aggressive buyers‟

market, HRSG manufacturers have been compelled to concentrate on

developing low-cost HRSG designs which just meet the requirements of

purchasers‟ specifications and boiler design code.

High efficiency has driven rapid increases in GT exhaust flow and temperature

imposed on HRSGs. Comparative data of different generation GE model gives us

an idea how they have increased.


44

GE CCGT Model Exhaust Temperature (oC) Gas Flow (TPH)

5371 (PA) 485 446

6581 (B) 543 525

7121 (EA) 1070 536

9171 (E) 1484 543

9351 (FA) 2318 608

These have resulted in increase in the HRSG physical size, steam flow, pressure and especially temperature. These factors greatly increased the potential for early thermal fatigue damage in critical parts of the HRSG. CCGT start-up methods have been dominated by the concerns for the GT, the steam turbine and the demand for short start-up times from any condition. Little or no attention is given to the extremely damaging effects of rapid changes in the gas inlet temperature and flow to the HRSG during starts. Most large new combined cycle plants are designed under the assumption that they would be base loaded, or at least infrequently cycled. This basic assumption has proven to be far from actual operating modes for most new plants. Two-Shift cycling is differentiated from Seasonal Duty where plants are run essentially base load, but only for a few months of the year. New plants include those that are commissioned but not running or which were inspected close to the time of commissioning. Since combustion turbine ramp rates and startup procedures directly affect HRSG component temperature ramp rates, the push to rapid CT startups results in greater ramp rates in HRSG hot section components than was assumed in plant design analyses. Larger thermal stresses result with significant implications for fatigue life of affected components such as drums, thick section headers and tube-to-header welds. In additions, rapid thermal response results in more condensate accumulation during startups and a greater requirement for attermperation spray to control piping metal temperatures. These extreme conditions that are caused by cycling operations sometimes result in water hammer in affected piping systems, thermal quenching of hot component surfaces and in some instance leakage or failure of the pressure boundary at tube-to-header welds, riser piping to drums, crossover (connecting) piping and drain connections. Cold weather operations also provide a different challenge with the need to maintain temperature to prevent header failure from freezing conditions. While new plants have operated in general significantly less than originally assumed, most have pursued an aggressive approach to assure that HRSG component integrity is verified by periodic inspections; usually during scheduled outages when the GT maintenance has been scheduled. A thorough inspection of a large HRSG (for example, behind a Frame 7FA, or 501F/G GT) with reheater components typically requires about 2 days for a crew of 2 people. These inspections are more detailed than statutory “boiler” inspections and typically include the following activities: 1. visual inspection of HRSG gas path components: tubes, headers and their

supports, crossover piping, risers, drains, gas baffles, acoustic baffles and related structural components.


45

2. Ultrasonic testing (UT) of wall thickness for selected (high risk) tube, header and riser components, thereby establishing the condition of HRSG components early in life. Drum baffle plates and in some instances cyclone separator “can” thickness are also measured at some plants.

3. visual inspection of accessible HRSG water-side components (for large combined cycle plants this is generally limited to drum surfaces and internals) including: primary and secondary steam separation devices, feed water penetrations, instrument and blow down penetrations and baffle plates and their mechanical restraints (bolting and/or welds).

In addition to these routine activities, plants with a history of HRSG component damage may also schedule dye penetrant (PT) inspection of areas susceptible to certain types of cracking, radiographic testing (RT) of tube-to-header welds when there is a suspicion of weld defects or sub-surface cracks.

Some of these cycling related damage mechanisms (such as tube leaks and failures) have been significant enough to require (or cause) plant shutdown. Others have been detected during scheduled HRSG inspections. While much of the focus of good operations and maintenance practice is oriented toward controlling corrosion of susceptible materials, primarily the carbon steel components that comprise most of the HRSG surface area, one immediate observation is that corrosion – at least so far – is not a significant problem at most plants. This is less a consequence of excellent water chemistry control than it is simply too early to detect small amounts of corrosion shortly after commissioning. Flow accelerated corrosion (FAC) is a high-visibility issue which has been the cause of numerous fatalities at power plants over the years. FAC has not been detected at new reheat units although in general they have not operated long enough to experience significant wear, even for the highest risk locations. Experience from previous HRSG designs that have operated for longer periods (50,000 – 100,000 hrs) indicates that will likely change despite the best efforts of plant staffs to maintain water chemistry within targets. Those units with some cold end corrosion problems generally have a combination of design issues, fuel gas quality issues and often frequent exposure of susceptible surfaces to high ambient humidity with long periods of layup.

Tube and Header Leaks and Failures

The most significant damage that occurs in HRSGs is generally leaks and failures of pressure parts; specifically, tubes, headers and connecting piping. Tube failures are well known as dominant contributors to plant unreliability. While tube repairs are not lengthy procedures, they contribute substantially to the cost of cycling duty when they occur. Leaks and failures in larger components such as headers, major connecting piping and steam piping can require more lengthy outages with correspondingly greater costs. The most common tube damage mechanism is probably bowing which is attributable to a variety of sources including differential thermal stress, manufacturing variations in tube length, etc. From our inspections of new (pre-operational) units, we have observed that some slight tube bowing is sometimes present prior to operation. However, large displacements are not observed pre-service. Tube failures are less common, but have occurred at many large reheat HRSGs. The


46

root causes of these failures varies and depends on many factors including: material type, exposure to high temperatures (gas temperature) during startup followed by quenching from condensate accumulation or excess attemperation spray, waterhammer and stress corrosion cracking. Flow accelerated corrosion (FAC) has not been observed in these relatively new units to date. Condensate formation during startup is a well-known problem and plants experiencing repeated tube failures, extreme tube bowing and or related problems with attemperation spray equipment have sometimes installed temporary thermocouples to more accurately ascertain the temperature variations in reheater (and superheater) tubes. Some plants have also installed thermocouples to determine whether steam binding is occurring in HP Economizers that are poorly vented.

Kink in Reheater Tube – Cycling Unit

RH Tubes Thermocouples Identify Condensate

During Start-up/Shutdown

Bowed RH Tubes below Cold Reheat Inlet

following Water hammer Event

Low Cycle Fatigue Failure of T91 Reheater

Tube

Fatigue Failure of 304H Stainless Reheater

Tube Stub

Crack in P22 Reheater Header Weld

LCF in critical pressure parts of the HRSG Each design of HRSG has unique configurations and header dimensions which have

different LCF damage rates if subjected to similar transient conditions. Analysis of


47

all pressure parts of a complete large HRSG and LCF failures on other designs

highlight three areas which should be assessed on all designs of large HRSG:

a) HP superheaters and outlet manifold,

b) HP steam drum and evaporator circuit and

c) Low-temperature economizers a) HP superheater and HP outlet manifold Superheaters are subjected to a rapid increase in the GT exhaust temperature

during starts from any condition. On large units the GT exhaust temperature ramps

from 80°C up to 450°C in 5 min as the GT accelerates to synchronous speed. First

the steam flow into the superheater header can be 200°C higher than the

temperature of the outlet header of the superheater. Superheaters are designed

to stress limits set by creep considerations. When creep is significant, there is a

strong interaction between creep and LCF; thus, the creep life used at the end of

the planned life of the HRSG must be limited to not more than 60 per cent to leave

some allowance for LCF damage. For example, ASME Section III code case N47

indicates that, when the creep life factor is 0.6, then the LCF life factor is limited

to 0.052 (Fig. below).

Factors which determine the severity of thermal stresses developed by high heat

input to the outlet header include but are not limited to:

i. the header thickness, ii. the header diameter, iii. differences in the tube temperatures, iv. the tube diameter and pitch and v. the header temperature and transient steam temperature.

The only operational means available to minimize thermal stresses by hot steam

into the superheater is to hold the GT load at a nominal block load until steam

flow is established in the superheater and then to control the GT loading rate to

ensure that the temperature difference between the steam temperature and the

average header wall temperature remains at or below the initial difference on first

admission of steam.

Condensation occurs in superheater tubes during every purge of the HRSG prior to

GT ignition. This is because the GT exhaust temperature falls below the saturation


48

temperature. Quantities are substantial during hot and warm starts and a repeat

purge can fill the front panel tubes of the superheater. The condensate should be

removed from the lower headers at the peak rate at which it forms to prevent

pooling and flooding. The crucial importance of adequately sized and correctly

operated drains on superheaters has been overlooked on many large CCGT

installations. Many units have no blow-down vessel for HP high-temperature drains

from the superheater. Others have a blow-down vessel inadequately rated for the

flow, pressure and temperature of drains from the superheater during hot restart

purges. The drain installed on most designs of superheater was sized for

maintenance purposes and is too small for clearing the condensate at the rate

at which it collects.

Even where superheater drains are installed and connected to a blow-down tank, more often than not they are incorrectly used or not used at all during hot starts because many operators are unaware that conditions exist which cause substantial condensation in superheaters. In addition to forming a condensate during purge, there is a considerable chance of the formation of a condensate during synchronization of a large GT as the firing rate is reduced to control the speed for synchronization.

b) HP steam drum and evaporator circuit

Cold starts from ambient temperature are most damaging to the HP drum and

superheater headers because of initial condensation heating. Starts after weekend

shutdown can develop high thermal stresses if the GT is loaded too quickly.

The HP and saturation temperature step limits and large range ramp rates should

be determined to match the projected lifetime number and mix of starts. It is

important to keep the pressure as high as possible at all restarts by closing all

isolation valves. This is beneficial to the HP superheater outlet headers as well as

to the HP drum.

a) Low-temperature economizers Unless means are provided to pre-heat the feedwater before it enters, the low-

temperature economizer inlet headers are vulnerable to quench cooling during all

starts because the feedwater is not required for raising the drum level until long

after GT ignition, especially on cold starts. By the time that the drum swell

subsides and the feed flow is first required on any start, the gas temperature has

raised the low-temperature economizer inlet header and tubes perhaps 100 – l50 °C

above the feedwater inlet temperature. Depending on the tube arrangement at the

header attachments this thermal quench may be large enough to be an LCF concern

if starts become frequent.

Boiler and Steam Piping Damage Problems with boiler and steam piping is often associated with the reheat piping; particularly where attemperator sprays have been designed with too short downstream straight pipe lengths (less than 10 pipe diameters). Incomplete atomization of attemperator sprays impacts downstream piping surfaces as liquid droplets where it can cause significant thermal stresses. Water hammer is another phenomenon that has occurred at a number of combined cycle plants. It is often attributed to a combination of problems related to spray valve control, drainage of condensate or abrupt valve actuation.


49

Waterhammer is generally a destructive transient; casualties typically include adjacent piping supports with yielding of steam piping a common end result.

Leaking 16” Reheater Crossover Link Piping

Water hammer Damage to Cold Reheat Piping

and Supports

Cold End Corrosion Cold end corrosion is enhanced in many merchant plants because they are

operating for more time at part load than anticipated by design (and therefore

component surface temperatures may be below the acid dewpoint temperature

where acids will condense on tube surfaces and corrosion damage will occur). This

problem is aggravated by greater than anticipated sulfur content in fuel gas in

some locations. Older units have employed CO2 blasting to remove some of the

deposits, but the effectiveness is limited if the affected harp (typically the LP

Economizer or Feedwater Heater) has many tube rows.

Ammonium Bisulfite Accumulation on

Feedwater Heater Final Row

Sticky Deposits (pH = 3) on LP Economizer

Tubes

Moisture contained in gas turbine (GT) exhaust gas will condense on HRSG heat-

transfer surfaces when that metal is below the gas dew point temperature, which

typically ranges from 112F to 120F. Dew point corrosion usually is found in low-

pressure (LP) economizers and condensate heaters that receive water from a

relatively cool source; the condenser hot well, for example.

Piping, headers, and tubes in these HRSG components operate very close to the

water temperature; especially in upper and lower crawlspaces, where there is

relatively little gas flow and heat transfer. The inlet piping, headers, and tubes are

at the lowest temperature and most likely to show signs of attack.

Some OEMs specify alloy materials to protect against corrosion. Others provide a

recirculation system or external heat exchanger to increase the temperature of

incoming feed water. However, many HRSGs in service were built with carbon-

steel materials in areas where metal temperature is well below the gas dew point.

Dew point corrosion is easy to identify visually given access to the gas side of the

HRSG where feed water enters.


50

Compare the condition of inlet pipes, headers, and tubes to that of nearby piping,

headers, and tubes: If components at the feed water inlet exhibit greater material

wastage, dew point corrosion is the likely cause.

boiler tube materials, their failures and prevention

Documents