Selecting steel anchors for monolithic refractory linings

6 Sigma phase in refractory anchor materials
It is known from technical articles (e. g. [10]) that some stainless steels can degrade under various thermal conditions, e.g. 310ss forms a complex chromium-rich intermetallic compound, known as sigma phase in the temperature range of 500 °C to 900 °C. The amount of sigma phase formed increases with time if the steel is held at reasonable constant temperatures. The formation of this chromium complex is commonly held to be undesirable as it may change the metallurgical properties of the steel. Sigma phase has a variable composition depending on a range of factors including the base material from which it was formed and the particular temperature to which it was exposed. The presence of sigma phase will often lead to brittle failure at low temperatures, less than 200 °C (particularly during demolition), and can increase oxidation rates of the anchor material.
 
Figure 5 shows a section of a 253MA refractory anchor that failed in a brittle manner (during demolition) and Figure 6 is the metallurgical micrograph of the steel anchor showing the sigma phase globules and carbides along the grain boundaries. The anchor had been in service for approximately 12 months at temperatures of approximately 900 °C. The condition of the anchor surface was good with little visual sign of oxidization.
 
Lately DS alloy has been gaining more acceptance as a refrac­tory anchor material as it is reported by the manufacturer that the alloy is resistant to sigma phase in the 500 °C to 900 °C range. However, analysis of various anchors from service (Fig. 7)
has identified that DS alloy can also suffer from a chromium com­-
plex phase, similar to sigma phase. The phase composition was measured to be rich in chromium, silicon and molybdenum, all strong sigma and ferrite formers. The presence of significant molybdenum in the phase was interesting, as there was only trace amounts in the bulk material. However, brittle failure at ambient temperatures was not associated with the phase.
 
The chromium complex phase observed in these DS alloy anchors formed as a band at and near to the surface, suggesting an environmental effect. Subsequent intergranular oxidation that occurred within the chromium complex band has initiated fractures towards the inside of the anchors, reducing the load-carrying capacity of the anchor. The centre of the anchor has then fractured due to overload (creep or yielding) of those areas that were less affected by the formation of Cr complex and oxidation (Fig. 8).
 
Alloy DS contains chromium, the main element responsible for oxidation resistance. In order to fulfil its role in providing corrosion and oxidation resistance, chromium must be in solution in the matrix. The formation of a ferritic chromium rich phase has therefore led to a significant reduction in the local dissolved chromium content. The material denuded of chromium is then more susceptible to oxidation. The speed of formation of ferritic phases such as sigma is relatively slow especially in austenitic alloys such as Alloy DS.
 
While it is claimed that Alloy DS is resistant to sigma phase there are undoubtedly compromises made in its composition to enhance oxidation resistance on the one hand and reduce sigma formation on the other, these two requirements being, at least in part, contradictory.
 
The presence of sigma phase in refractory anchors has been blamed for the loss of structural integrity or failure of refractory linings. However, the loss of structural integrity is generally a low temperature phenomena associated with brittle failure of a critical component. In fact 310ss with sigma phase can have an equal or greater tensile strength when compared to the original condition.
 
7    Oxidation resistance
The oxidation resistance of steel alloys is very important as it has been one of the critical parameters used when selecting refractory anchors. When stainless steels are exposed to air a thin oxide film will form on the surface. At low temperatures this film takes the form of a thin protective layer but at high temperatures the oxide thickness increases considerably. It is reported that temperatures above the so-called “scaling temperature” (Table 2), then the oxide growth rate becomes unacceptably high.
The oxidation of steel anchors is critically important as the integrity of the anchor in a refractory lining determines the life of the lining, particularly at high operating temperatures. If the oxidation rate of the anchor is too high then a reduction in a refractory lining life can be expected. Steel anchors are relied upon more then ceramic anchors for lining integrity as it is well known that ceramic anchors are very unreliable as the material behaviour can be very brittle in character at temperatures less than 900 °C. This sudden brittle failure of ceramic anchors can result in a shorter lining life compared to using steel anchors. This means that steel anchors are widely used, and problems are encountered when steel anchors are pushed to the limit in temperature environments greater than 1000 °C. At these temperatures the oxidation of the steel must to be considered in determining the component life. The reason is that oxidation of the steel decreases the load carrying capacity both in static and dynamic environments.
 
However, this work shows that selecting the grade of steel for a refractory anchor based on the “scaling temperature” is very erroneous and can lead to poor material selection. This is due to the fact that oxidation of the steel is a function of temperature and time. It is very difficult to compare published data on the oxidation resistance of various alloy types as manufacturer’s results are generally presented in different units. It would be beneficial if industry had a standard test on oxidation resistance. In [11] the long term oxidation resistance for different alloys presented as a weight gain is shown. The data reveal that the long term oxidation of 310ss is vastly superior to 253MA (Fig. 9).
 
Takehiro Horio et al. [12] have investigated the oxidation of three types of metal anchors. The anchors evaluated were carbon steel (SS41) used for low‑temperature services, and stainless steel (type 304, 310S) used for high‑temperature services. It was found that the oxidation resistance of each material ­varies with environmental conditions, and the combination of CO2  (97.5  %) and H2O (2.5  %) was the most corrosive tested for any material. It was found that SS41 increased in weight by 23.5  % and 304 was oxidized to some extent under CO2 + H2O, but 310S was barely oxidized under any condition tested. It was also reported that anchors encased in refractory oxidize at a slower rate than air. We have also confirmed this through measurement of service samples.
 
Figure 10 shows the effect of increasing temperature on the metal lost in an air stream environment after 1008 hours exposure. It can be seen that the oxidation rate starts to increase exponentially above 1100 °C and the rate of oxidation for 253MA steel is significantly higher than 310ss. Thus it can be seen that selecting a refractory anchor based on scaling temperature alone is very misleading as the life of the anchor can be very short.
 
It has been proposed that anchor failure at the interface location is in part due to oxidation from hot gas tracking, i.e. when the process gas tracks between the hot face and insulation layer. It is known that the oxidation of steel is significantly different when exposed to air and encased in castable. If hot gas tracking is the mechanism for anchor failure then the oxidation of the steel at the interface zone can be expected to be significantly higher than the corrosion in the refractory. From our research it has been concluded that hot gas tracking does not initiate anchor failure but occurs after the anchor has deformed subsequently opening the interface, and allowing oxidation of the anchor to occur.
8    Impact energy (toughness) at the anchor
The failure of refractory anchors during demolition is a serious problem and identifying failure mechanisms can be difficult. When refractory concrete is being broken during demolition it can give the appearance that refractory anchors have already been broken. This in fact may not be the case. As previously mentioned the impact energy from jackhammering can result in sudden anchor failure due to brittleness of some alloys at low temperatures. Alloys like 310ss and 253MA are known to form sigma phase which can embrittle the steel at temperatures less than 200 °C. Thus it is not that the anchors have failed prior to demolition but the anchors have a thermal sensitivity that means they are susceptible to sudden failure at ambient temperatures when sufficient energy is imposed.
 
Figure 11 shows the Charpy impact energy at room temperature for Alloy DS, 310ss, Inconel 601, Haynes 160 and 253MA.
The data shows that Alloy DS and Inconel 601 have the ­
highest toughness at room temperature at 1000 hrs of operation. It should be noted that the exposure temperature is low and duration for the test samples is relatively short in some cases. Thus, conclusions drawn from the data must be carefully evaluated.
 
However, it is clear that some metals have a propensity to become embrittled at low temperatures quite rapidly. This is important as during demolition some anchors are likely to fail allowing the hot face to fall away, representing a significant safety concern.
 
9    Creep rupture
Creep rupture in anchors is associated with static structures where the load on the anchoring system can be low and reasonably constant. The load can either be self-weight of the ­refractory concrete layers and/or thermal strain loads. Our ­research has found that creep rupture failure of refractory anchors in static plant structures is a common failure mechanism.
 
Understanding creep failure means that better structural life prediction can be made and the probability of catastrophic failure can be reduced. Also at the design stage a more ­systematic approach can be carried out and the general “rule of thumb” can be avoided. Using the current approach which has been ­developed from experience does not properly account for ­lining structural conditions.
The creep rupture stress (CRS) is related to time and temperature by parametric Larsen Millar Parameter (LMP) data, which is available for most steel alloys used for refractory anchors, e.g. 310ss, Alloy DS and Inconel 601. The LMP curve for selected alloys is shown in Figure 12. Essentially, the lower the LMP for a particular stress, the lower the creep properties. The results are based on experimental data obtained from creep testing and care must be taken when using the data outside the experimental range.
 
The predicted CRS for 253MA and DS alloy refractory anchors after 30 000 hours at 1050 °C is 1.5 MPa and 4 MPa, respectively, assuming no corrosion of the steel. If corrosion, due to oxidation, of the anchor steel at 1050 °C is taken into consideration then the time to failure is estimated at approximately 13 800 hours for the 253MA steel and approximately 15 500 hours for the DS alloy anchor. Increasing anchor exposure temperature to 1100 °C can significantly reduce the creep life of the anchors from tens of thousands of hours to thousands of hours.
 
It is assumed that the oxidation of the steel progresses evenly along the anchor and at a slower rate than in air. However, process conditions can significantly vary the anchor corrosion rate, e. g. the presence of chlorides.
 
If the load on an anchor is increased by changing the material (hot face) density from 2300 kg/m3 to 3000 kg/m3, for example, then the stress on an anchor (253MA) will also increase by 30 %. This means the life of an anchor due to creep rupture
stress decreases from approximately 30 000 hours to approxi­mately 5000 hours, or if the refractory (hot face) thickness is increased by 7.7 %, i.  e. an extra 10 mm, it means the life on the anchor (253MA) will decrease from approximately
30 000 hours to approximately 19 000 hours.
 
These stress calculations assume simple ideal conditions due to material weight only. It is important to be aware that the anchor stress is the summation of thermal load and material weight. While this is an idealized calculation for overhead dense refractory lining (roof) the data is based on published creep rupture and oxidation rates. It also assumes that thermal strain load on an anchor is less than 10-4 s-1. The presence of chlorides or sulphates has not been considered in the oxidation rate of the steel.
 
It can be seen that selecting the base metal for refractory anchors should not be based on scaling temperature or “rule of thumb”, it needs to be based on oxidation rate, service conditions, service time and material weight. It is possible to predict the service life of refractory anchors if a systematic approach is taken and process and thermal load conditions are taken into account in the design phase. The accuracy of the results can be improved by collecting and analyzing samples from critical plant areas.
 
10 Conclusions
The failure of steel alloy anchors at the interface (zone between the insulation and hot face) is due to the prolonged low stresses on the anchor at high temperatures. There is no clear evidence to conclude that steel anchor failure is caused firstly by hot gas tracking induced oxidation of the anchor at the interface. Failure of the anchor will be due to either creep rupture or low thermal strain rate induced stress.
 
It is clear that the general “rule of thumb” that relates anchor spacing to lining thickness is fundamentally incorrect and should not be used. Selecting anchors based on published scaling temperature is also flawed and should not be used.
 
The correct selection of steel anchors for monolithic linings is critically important as the integrity of the anchor in a refractory lining is one factor in determining the life of the lining. To more accurately determine refractory anchor stress numerical analysis using ATENA needs to be carried out. Creep rupture stress and oxidation rates are two critical parameters that need to be taken into account.
 
If the oxidation rate of the anchor is too high then a reduction in a refractory lining life can be expected. This is particularly important as volatile salts can significantly accelerate the corrosion rate. While the oxidation rate of steels can be estimated from published data it does not reflect corrosion rates in refractory. It is recommended that corrosion rates be determined from in-service samples to more accurately predict anchor life.
 
Metallurgical literature has shown that 310ss forms a chromium complex (sigma phase) in the temperature range of 500 °C to 920 °C and the amount formed increases with time. The presence of sigma phase will often lead to brittle failure at low temperatures, less than 200 °C, and increased oxidation rates of the steel at high temperatures. Published data shows that the impact energy at ambient temperatures for 310ss affected by sigma phase formation can be reduced by as much as 80  %. However, the loss of structural integrity due to sigma phase in 310ss is a low temperature phenomena associated with a loss of toughness. The embrittlement does not occur at operating temperatures. This loss of toughness in 310ss can be incorrectly attributed to a loss of structural integrity or failure of refractory linings which is actually caused during removal of refractory.
 
If DS Alloy is selected as a refractory anchor material then creep rupture must be taken into consideration. DS Alloy has an inherently lower creep rupture stress than some other alloys.
 
It has been found that the use of plastic tips on refractory anchors does not prevent cracking of the hotface castable around the anchor tips. Plastic tips only increase the corrosion of the steel in that area when exposed to high temperatures.
 
It is clear that a more rigorous engineering approach should be applied when designing refractory structures. The designer needs to consider not only the load carrying capacity of the anchors for both static and dynamic environments but also the corrosion rate of the steel in the process environment.