What Ceramic Manufacturers Need to Know about Dryout of Refractory Castables for Kilns and Furnaces
Water—physical and chemical—must be removed before refractory castables can be used in high-temperature applications.
The proper installation of refractory castables includes mixing, placing, curing and drying. Improper amounts of water used when mixing or poor densification during placement can lead to higher porosity, lower density, and reduced strengths. Improper curing can result in the formation of undesirable hydrate phases of the cement. Each of these issues can contribute to inferior performance of the final castable lining.
The dryout process is usually the last step before a refractory castable lining is put in service. The dryout should not be compromised because, just as with the other installation parameters, problems will result if proper care is not taken to remove the water from the castable lining.
Water Additions and Cement Hydrate Phases
Water is added to refractory castables to achieve the desired fluidity and cement hydration. Water fills gaps within a refractory castable and allows the material to move and flow during installation. Water reacts with the cement to form interlocking crystals that bond the materials together.
Two types of water remain in the refractory castable after installation and hardening: physical and chemical. Physical water is water remaining in the material that did not react with the cement but enabled the castable to flow. It is also called free water because it is in the pores of the castable, not bound in any hydrate phases. Chemical water is water tied up in the different cement hydrate and alumina or calcia hydrate phases.
According to Gitzen and Hart, approximately 75% of the casting water remaining after hydration is free water in the pores. The remaining 25% is chemically combined water. However, with developments in ultra-low-cement refractory castables, the amount of chemically combined water can be closer to 10% of the total water added. The purity and amount of cement also affect the amount of free or chemically bound water in the refractory castable.
Free water and chemically combined water must be removed before the refractory castable can be used in high-temperature applications. The water must be removed slowly because it will expand as it is heated and potentially yield internal pressures that can exceed the strength of the refractory castable and cause the lining to blow up or explosively spall. When water turns to steam, there is a volume expansion of approximately 1,600 times. This increase in volume is what causes high pressures inside the lining. The removal of water must be done in a controlled manner to minimize and dissipate the buildup of steam pressure.
When water reacts with calcium aluminate cement, many different types of hydrates can be formed. The types of hydrates formed during curing are determined mostly by the curing temperature.
Hipps and Brown have identified several critical dehydration points based on the steam volume (water vapor) produced by free water in the pores and from calcium aluminate cements in the early stages of heating. These hydrates are associated with relatively high volumes of steam that can relate to high internal pressures within the lining. The major dehydration points they identified are shown in Figure 1 as a function of the water vapor (steam volume). These hydrates will lose water over a temperature range, not at one specific temperature.
Hipps and Brown concluded that dryout schedules should be written to include holds just above these major dehydration points to allow steam pressure to dissipate. They suggest hold temperatures of 302°F (150°C), 572°F (300°C) and 1,049°F (565°C) during the dryout.1
Less significant dehydration points, in terms of steam volumes, observed by Hipps and Brown were alumina and calcia hydrates. The dehydration temperatures and reactions of these phases are:
300°F (149°C): dehydration of AH3 (gibbsite) → AH + 2H
570°F (299°C): dehydration of AH (boehmite) → A + H
750°F (399°C): dehydration of CH → C+ H
where C=CaO, A=Al2O3 and H=H2O. These phases will dehydrate over a temperature range as well.
Although the alumina hydrate phases have lower volumes of steam released compared to the phases in Figure 1, these hydrates can reduce permeability and contribute to explosive spalling during dryout. The combination of the C3AH6 and AH phases dehydrating in similar temperature ranges may contribute to instances of explosive spalling at hot face temperatures around 600°F (316°C). However, the 600°F (316°C) is air temperature, so these phases would be dehydrating relatively close to the surface and not 3-4 in. into the lining (a common spall depth). Therefore, explosive spalling that occurs around 600°F (316°C) is probably more related to vapor pressures from free water in the lining than from the dehydration of these phases.
Cement Level
The cement level of refractory castables can vary significantly depending on the product and its intended use. ASTM C 401-91 has classified alumina and alumina-silica castable refractories based on lime (CaO) content. The cement level is important to the dryout of refractory castables because it affects the rate of moisture loss, strength and permeability of the installed product.
A moisture loss study was performed on four different commercial castables with different calcium aluminate cement levels. Figure 2 shows moisture loss vs. temperature for these four different castables. More moisture is removed during the lower temperature stages (steeper slope of the lines) than in the higher temperature regions. More water is removed during the lower temperature range because the casting (free) water is removed in this range. Approximately 75% of the casting water remains in the pores as free water and not in the hydrates. The maximum amount of free water will come off at or near the boiling point of water: 212°F (100°C). This indicates that controlling the dryout at lower temperatures is critical because this is where the bulk of the water is removed.
The relationship between the rate of moisture removal and cement level was evaluated for these castables at three segments of the dryout schedule. Figure 3 shows the moisture loss rate vs. cement level for all three segments. In the temperature range of room temperature and 250°F (121°C), the moisture loss rate tended to decrease with increasing cement level. This is because there was more water tied up in hydrate phases for higher cement-containing mixes. Correlation between moisture loss rate and cement content is low because there are other fine additives (e.g., fume silica) in the lower cement products that affect the moisture loss rate. These fine additives contribute to the bond phase of low-cement castables and also decrease the permeability of the castable.
During the 300°F (149°C) hold, there is no correlation between these two variables because of the different mix compositions. However, the low-cement castables (LCC) and high-cement castables (HCC) lost about 20% of their total moisture during this hold. This shows that moisture removal is not primarily related to cement level, but is dependent on other characteristics of the refractory castable.
Between 300-800°F (149-427°C), the trend line suggests that increasing cement content yields a higher moisture loss rate. A logical conclusion is that higher cement mixes have more water bound up in cement hydrates that must be removed at higher temperatures. However, the loss rates for the LCC (4% cement) and HCC (30% cement) were very similar. This shows that LCCs are not necessarily easier to dryout because they contain less cement. Fine additives to ultra-low and low-cement refractory castables can have a dramatic effect on moisture retention. Any time moisture is retained, internal steam pressures can build and potentially result in explosive spalling.
Curing Conditions
The curing conditions of the refractory castable have an influence on the type of cement hydrates formed. Curing at temperatures below 70°F (21°C) will yield CAH10 as the primary cement hydrate phase. This phase is a gel-structure that is relatively weak and has lower permeability. In addition, this phase requires more water to form and thus leaves less free water in the castable to help create pore channels that aid in the moisture removal. Conversely, the CAH10 phase will have more chemical water to be removed during the dryout.
An additional problem occurs with the CAH10 phase when it is heated producing a volume shrinkage of approximately 50% (MacZura et al).2 The shrinkage results in some loss in strength because the cement bonds “pull away” from each other and from other particles within the castable. The higher the cement content in a castable, the greater this effect has on the final strength of the lining. The reduced strength, reduced permeability and higher amount of chemically combined water can make the installed castable lining more susceptible to explosive spalling because it may not be able to resist internal pressures caused by steam generation while drying.
Curing above 70°F (21°C) should be done as much as possible to minimize the negative effects of the CAH10 hydrate. The hydrates formed above 70°F (21°C) will yield higher strength and permeability, both of which help make the refractory castable more conducive to a successful dryout.
Gitzen and Hart have investigated the effects of the curing temperature on the different properties of a tabular alumina castable with 15% cement.3 Their tests show the relationship between curing temperature and transverse strength, permeability and explosion temperature. Generally, as curing temperature increased, so did the dried strength, permeability and explosion temperature between 40°F and 90°F. Ultra-low and low-cement castables minimize the impact of the effect of cement hydrate types formed at various curing temperatures because of the lower levels of cement used in these types of castables.
Permeability and Dryout Aids
The permeability of refractories, as defined by ASTM C 71, is the capacity of a refractory material to transmit fluids (liquids or gases). Permeability is important to the dryout process because it relates to how quickly and easily moisture is removed, which can minimize the pressure buildup within the installed castable.
The permeability of refractory castables can be increased in many ways. Two specific methods include the use of organic fibers and powdered metals.
Organic Fibers
Investigations were conducted to explore how permeability was affected by cement content, fiber content and installation consistency. The first conclusion is that higher cement levels related directly to higher permeability. The second conclusion is that, generally, pumped installation produces higher permeability than vibrated installation, primarily because more water is used. Interestingly, higher levels of organic fiber content, combined with vibration, produced the highest permeability.
Figure 5 shows that the addition of organic fibers had the biggest impact on increasing the permeability. The mixes with no organic fibers added had essentially zero permeability up to 800°F (426°C).
Powdered Metals
The addition of metal powder in the refractory castable will react and give off hydrogen gas, which creates channels within the castable matrix that have proven to be very effective in allowing moisture to escape as the material is being heated. The evolution of hydrogen gas can preclude the use of powdered metals. In addition, escaping gas can cause laminations and unacceptable surface appearance, but this is typically more of a cosmetic issue.
A dryout study compared organic fibers and powdered metals in an alumina-silica-silicon carbide-carbon blast furnace trough castable. The samples passed the test if they did not spall after being in the furnace for 15 min. The results of the dryout study show that mixes containing metal powder had a significantly higher spall resistance than a mix with organic fibers alone. These results are confirmed by actual field experience.
Types of Dryouts
One-sided dryout involves heating the lining from one direction to remove moisture from the castable. Water is driven from the hot face to the cold face of the refractory lining. One-sided dryout is commonly used in the field to dry castable linings such as ladles, reheat furnaces, blast furnace troughs, tundishes, cement kiln pre-heaters, aluminum furnaces, incinerators, etc.
Two-sided dryout involves heating the material from more than one direction (five or six sides). These types of schedules are used to dry pre-cast shapes. Water is driven from the hot face to the center of the piece. Unlike one-sided dryout, however, all of the moisture must be removed through the hot face. Therefore, the dryout schedules must be more conservative than those for one-sided dryouts.
The dryout schedule recommended is based on a number of factors. The length and ramp rates of the schedule depend on the material type, lining thickness, number of components in the lining, tons of material installed and restrictions of the unit being dried out. If the lining thickness varies within a unit to be dried out, the thickest section should be used to establish the schedule. The schedule is always based on the material or section of the unit that is most susceptible to steam spalling.
Successful Dryout
The dryout of refractory castables is one step in the installation process of a refractory lining. Many variables can affect the dryout. The type of refractory castable being installed, the curing conditions and the use of dryout aids are important because they can all affect the strength and permeability of the final product.
Dryout aids, such as organic fibers or powdered metal, should always be used when installing a refractory castable lining. The permeability and strength are related to the ability of the castable to release the water vapor formed during dryout and to resist the pressures created by the steam.
The dryout schedule selected is usually based on experience and takes into consideration the material type and lining configuration. A successful dryout does not happen by chance—it is planned.
For more information, contact the author at (412) 469-6120 or jbogan@thinkhwi.com, or visit https://thinkhwi.com.
References
- D.L. Hipps and J.J. Brown, “Internal Pressure Measurements for Control of Explosive Spalling in Refractory Castables,” American Ceramic Society Bulletin, 63, 7, 905-910, 1984.
- G. MacZura, L.D. Hart, R.P. Heilich and J.E. Kopanda, “Refractory Cements,” Ceramic Proceedings, Columbus, Ohio, 1983.
- W.H. Gitzen and L.D. Hart, “Explosive Spalling of Refractory Castables Bonded with Calcium Aluminate Cement,” American Ceramic Society Bulletin, 40, 8, 503-510, 1961.