Coatings in the electricity supply industry: past, present, and opportunities for the future

Abstract

The cost of electricity, the reliability of its supply, the impact on the environment, and the global sustainability, are all forcing the supply industry to make greater demands on the materials of construction of the generation, transmission, and distribution components. Not only is the performance being enhanced, but the requirements for durability, and the ability to identify potential failures, to estimate remaining life using non-intrusive methods, and to repair or replace quickly and effectively, are also subject to greater demands. Coatings play an important role in all of these aspects. In this paper, a very general summary of some of the more important issues is presented.

Introduction

The degradation of performance in the various parts of the electricity supply industry represents a significant cost penalty. The degradation may be associated with the loss of components, for example due to fracture induced by a time dependent deterioration process to the point where the component can no longer sustain the loads to which it is subjected in normal operation; or to a decay in performance, for example by the aerodynamic losses associated with roughening of an airfoil surface.

In some cases, either of these may be associated with a degradation of the surfaces of the components, and it is in these situations that surface coatings may be of benefit.

This is the subject of this presentation. Although the topic of this session is high-temperature coatings, in this presentation there will be some general discussion of processes which occur at lower temperatures, for completeness.

The lifetime of any component in an engineering system is determined by some degradation process, which may be as rapid as high-temperature oxidation or as slow as graphitization of mild steel. In a number of cases, the degradation is related to the component surface.

In some of these cases, the life-limiting process is obvious: loss of load-bearing section in general corrosion.

In other cases less so: degradation of strength by the loss of subsurface alloy elements.

And in others, indirect: fatigue failure of a turbine blade as a result of stress concentration due to pitting of the blade surface.

Degradation may not involve an actual failure of a component: solid particle erosion in steam turbines (SPE) greatly reduces the power produced by the turbine stage; oxide build-up on a heat exchanger changes both the heat-transfer coefficient and the temperature distribution.

The impact in both cases is economic, due to performance degradation.

Coatings are primarily used to restrict surface damage of components in practice, where other requirements prevent the substitution of an inherently resistant material.

Generally, materials selected for a particular duty will last an adequate length of time, provided the designer has taken a proper account of all the factors affecting life. The lifetime is determined by some degradation process, which may be mechanical — fatigue, in the case of materials serving a low-temperature function where the stress varies, or creep, for a material subjected to high stress at elevated temperatures—or one of the processes listed later. In some cases, the separation of these processes is not so clear—thus, pits on a turbine blade will act as stress raisers, and accelerate fatigue failure. Loss of section as a result of general corrosion or erosion will result in an increase in the stress experienced by a component and result in accelerated creep failure. Loss of load-bearing section, for example by internal penetration of corrosion or by near-surface depletion of strengthening alloy elements, will have the same effect. The surface degradation processes may also adversely affect the performance of components—an example is the solid particle erosion (SPE) of steam turbine blades by spalled oxide particles entrained in the steam, in which the most important economic effect is the degradation of the performance of the turbine, rather than any macroscopic failure of components. The thermal conductivity of the corrosion product is usually very significantly less than that of the metal, and in a component exposed to a thermal flux, the growth of the corrosion product will result in changes in the temperature distribution. For oxide growth on the inner surface of heat exchanger tubes, the effect is to raise the metal mean temperature for a fixed heat flux, and this can lead to an overheating creep rupture failure: the root cause, however, is the oxidation.

While the statement at the beginning of the last paragraph is true (no-one would build a machine if one knew at the beginning that it would have an inadequate life), the evolution of the products will inevitably lead to an effort to improve performance, and the usual way to do this is to improve the system design parameters, while keeping the overall materials and manufacturing specifications the same. Eventually, a point will be reached where the lifetime of some component will decrease to a level where the economical value of the overall system is compromised, and changes have to be made. So far as surface degradation processes are concerned, one initial option is to modify the surface while retaining the properties of the substrate, as part of the design bill-of-materials. However, more significant changes, which might be termed generational changes, in equipment will require radical changes in the materials, and in some of these, the solution is recognized from the beginning as requiring a materials system approach, in which a coating is an integral part rather than simply a band-aid.

A second issue is that while the design specifications as regards mechanical properties are limited by national or international standards, this is seldom (if ever) true for surface properties, and when a new system is put into service unexpected surface degradations may be encountered. A remedial action, which may involve the selective coating of components at risk, is needed.

Finally, the service parameters may change for one reason or another. An example relevant to power production is changes in fuel because of changes in fuel prices, or the appearance of new fuels at low costs, usually as a byproduct of some unrelated industry. Petroleum coke (Petcoke) and Orimulsion are two recent examples of this; increasing interest in the partial or total firing of biomass is another.

Generally speaking, coatings can be regarded as materials with greater resistance to the significant surface degradation process. In the case of corrosion, the coating itself has a greater resistance to the corrosive environment. In the case of erosion or wear, the coating is resistant to this aspect. However, the coating has a number of other important requirements. In brief, these are:

(1) The coating must have good adherence to the substrate, throughout the range of conditions that the component is exposed in service. Most obviously, it must tolerate without spalling from the surface the temperature variations that the component experiences in service; and the strains and strain variations that the component will normally impose on the coating. This latter is because the coating is normally thin in comparison to the substrate, and thus will be forced to match the substrate strains.

(2) A coating must not only be resistant to the condition for which it has been chosen, but any other condition to be experienced by the component. For example, in the case of a high-temperature coating, susceptibility to water damage of any kind when the system is down and the surface is cool is very undesirable.

(3) The conditions required for the coating process must be consistent with the system to be coated: a process requiring a high vacuum is not appropriate for coating a boiler water wall in place.

(4) The coating process must not damage the substrate. In general, this means that any process which involves heating the substrate to high temperatures needs to be assessed carefully.

(5) A very similar issue is that during the coating process or use, interdiffusion of any species between the coating and the substrate must also be assessed carefully. It may not in fact be harmful, but more frequently it is.

(6) The coating must not induce failure in the substrate. This means that the system should not allow a crack formed in the coating to propagate into the substrate, for example.

(7) Interface properties are very sensitive to the presence of minor impurities, since chemisorption (for example) can result in the formation of a complete monolayer of an element at an interface even if volumetrically the concentration is very low. These impurities may come from the substrate itself, the coating process, or the environment.

It is not the purpose of this paper to review coating methods in any detail, but it is useful just to list the principal coating types in relation to the methods of application.

Coatings resistant to high-temperature oxidation most frequently rely on the formation of a dense, adherent Al₂O₃ (alumina) layer at the interface between the coating and the environment. This means that the coating itself must have enough aluminum for alumina to be the preferred oxide, and to provide an adequate reservoir to reform the protective oxide if it is removed, for example by spalling or by mechanical abrasion, a reasonable number of times. One such coating is aluminum itself: the melting point is too low, but by use of appropriate methods, it can be deposited on a surface, for example by flame spraying, and then allowed to react with the surface to form an aluminide.

A slightly more sophisticated way of achieving the same result is to put the component in a pack containing either aluminum or an aluminum-containing alloy dispersed in an inert powder, together with an activator which is usually a halide. The aluminum is transferred from the pack to the substrate surface through the vapor phase by the halide which dissociates at the substrate surface, and diffuses into the substrate. This is called a diffusion aluminizing process, a pack aluminizing process, a pack cementation process or by a process name: alonizing is an example. Since in detail the actual deposition process involves the dissociation of a chemical vapor species, this technique can be regarded as one member of a larger group called chemical vapor deposition (CVD).

Alumina is relatively soluble in molten alkali salts; and in some circumstances where a molten salt is a component of the corrosive environment a coating which will form a dense adherent Cr₂O₃ (chromia) layer is preferable. This too can be applied by a pack cementation process, or by a CVD process where the chromium halide species is generated at some distance from the component to be coated.

These pack diffusion processes involve the substrate alloy itself in the formation of the coating, and in some cases the requirement that the chemistry of the alloy surface layers should not be modified significantly may be incompatible. The process usually involves holding the component in the pack for a considerable time at a relatively high temperature, and this too may not always be acceptable.

In a sense, electroplating can be regarded as a CVD process, and electroplating continues to be widely used. A recent example in a high-technology application is the widely used class of `platinum aluminide' coatings for gas turbines: typically, a layer of platinum is first electroplated onto the component surface, and the component is then aluminized, usually by a pack cementation process.

An alternative approach which does not involve the alloy itself in the coating process is the physical deposition of the coating composition on the substrate. One method of doing this is to evaporate the coating alloy from an ingot using an electron beam to form a melt pool; for more complex cases more than one such source may be used, either at the same time or sequentially. The vapor condenses on the component surface, which may be heated to improve the bond and the properties of the coating layer itself. This is called electron beam physical vapor deposition (EBPVD).

Another approach involves the deposition of a coating alloy in molten form onto the component surface using a flame or a plasma torch. The alloy may be supplied to the torch in a number of ways: the most usual are either as a wire or as a powder. The powder is partially melted by the carrier, and accelerated towards the substrate, where the semi-molten droplets `splat' onto the surface, generally cooling rapidly; once again the substrate may be heated. There is now a wide range of commercial techniques of this kind, which attempt to achieve higher impact velocities and thus higher densities. Vacuum plasma spray (VPS) or low-pressure plasma spray (LPPS) are two names for the same process, where the lowering of the pressure reduces the aerodynamic slowing down of the droplet as it approaches the surface. Other processes increase the velocity of the droplets by what amounts to a pulsed flame: the most important new technique of this type is the high-velocity oxy-fuel (HVOF) method. Whereas the low-pressure techniques cannot easily be applied to large components, such as boiler walls, the HVOF torch can, although the environment would be unpleasant for the workers.

The deposition and densification steps can be separated, for example by using laser remelting of surface deposits. Post-deposition optimization of the substrate/coating system (however applied) by heat-treatments, shot-peening, hot isostatic pressing (HIPing), partial remelting, etc., is often part of the overall process.

Thicker deposits are sometimes appropriate, and these can be achieved by weld-depositing material on the component surface (this is common in large chemical vessels); the attachment of a resistant alloy in sheet form by roll bonding, explosive bonding, HIPing (hot isostatic pressing); or, in the case of a tubular product, co-extruding a billet composed of the substrate alloy and the coating alloy. The last four of these require a relatively ductile coating material.

It is possible to use more than one of these methods to develop a composite coating system, for example by depositing a coating using EBPVD and then increasing the surface aluminum content of that by pack aluminizing. By using multiple sources and a shutter system, coatings with progressively changing compositions and characteristics can be manufactured: these are sometimes called `functionally gradient materials'.

This description is far from complete, but it will give some feeling for the choices available.

It would not be complete not to mention other cases where the properties of surfaces may be modified by coatings. Electrical and optical properties of surfaces are commonly modified, of course; and also thermal properties. One high-technology example is the application of diamond coatings to produce a surface that has a high thermal conductivity but a very low electrical conductivity; in principle this surface could also have a very low coefficient of sliding friction and a high wear resistance. These aspects of surface modification by coatings will be covered in some detail by other parts of this Conference, and will not be discussed here: this does not imply that they are not important for the utility industry. One special example of this class is the thermal barrier coating (TBC) which is increasingly used in the hot sections of advanced gas turbines.

The remainder of this paper will describe a few examples of surface degradation situations in the utility industry, and discuss the use—or otherwise!—of coating systems to combat the problems. With the exception of a brief mention of a problem in steam turbines, all of the aqueous corrosion problems will be ignored; this does not, however, imply that they are not important in our industry.