Manabu NOGUCHI*
Eiji ISHIKAWA**
Takashi KOGIN***
Suzue YONEDA****
Shigenari HAYASHI****
*
Technologies, R&D and Intellectual Property Division
**
Ebara Environmental Plant Co., Ltd.
***
Dai-ichi High Frequency Co., Ltd.
****
Faculty of Engineering, Hokkaido University,
To put the high-temperature erosion-corrosion-resistant alloys developed in the laboratory tests into practical use as thermal spray coatings, we examined the possibility of making them self-fusing alloys. Ni-Fe based developed alloys containing high concentrations of Fe produced granular protrusions during the remelting treatment, and smooth alloy surfaces like those of normal self-fusing alloys could not be obtained. Thus, we have conversely developed a thermal spray alloy coating that actively generates these granular protrusions and whose surface is entirely covered with a concavo-convex shape. The flow medium preferentially impinges on the granular protrusions of this convexity, which mitigates the erosion conditions of the sprayed base metal, resulting in a significant improvement in high-temperature erosion-corrosion-resistant properties. Demonstration tests have shown results that far exceed those of conventional alloys, and we are currently working to bring these alloys to market.
Keywords: Erosion-corrosion, Self-fusing alloy, Concavo-convex shape, Fluidized bed boiler, High temperature corrosion, Remelting treatment
A fluidized bed boiler is one of the biomass power plants expected as a renewable energy source. A heat transfer tube in the fluidized bed (hereinafter referred to as an in-bed tube), which is the main component of the fluidized bed boiler, is subject to high temperature erosion-corrosion damage due to erosion caused by the circulation of the flow medium and corrosion caused by chlorine and other substances derived from the raw material. Therefore, reducing the maintenance cost of the in-bed tube is expected to be one of the measures to reduce the operating cost of the plant. Therefore, we developed a high temperature erosion-corrosion-resistant alloy using a small-scale fluidized bed test apparatus and reported the results in our previous report1
）.
In the actual equipment, a thermal spray coating is applied to the surface of the in-bed tube, but in normal thermal spraying, peeling of the coating is inevitable. To prevent peeling, the actual equipment employs self-fluxing alloy spraying, which is a remelting treatment that metallurgically bonds the heat transfer tube (base material) and the coating. Self-fluxing alloying is essential for the practical application of the developed alloy. Therefore, we worked on the development of self-fluxing alloy technology for the developed alloy and developed a construction technology. Then, thermal spraying was applied to the in-bed tube of the actual equipment and the practicality was evaluated through a one-year demonstration test. This paper reports the results.
When remelting self-fluxing alloys, the entire thermal spray coating melts and flows off, if the temperature is too high, so the coating should be applied in a temperature range where the coating is in a semi-melting state. Generally, remelting is performed by visually checking the semi-melting state using a gas burner, but high-frequency induction heating was adopted as the remelting method because it enables high-quality coating through precise temperature control. The sprayed tube is rotated and the coating is melted as the high-frequency induction coil moves (Fig. 1). Although thermal spray-only coatings contain many voids, the remelting treatment improves the properties of the coating by bonding it to the base material and eliminating voids. The point of construction at this time is the optimization of the remelting treatment temperature.
Fig. 1 Remelting treatment by high frequency induction heating
Prior to the full-scale demonstration, an actual equipment exposure test was planned, including confirmation of the workability of the developed alloy and issues in actual use. Specifically, the developed alloy was thermally sprayed onto the surface of the thermocouple sheath tube and used in the actual equipment to identify unexpected risks. In addition to the Ni-based developed alloy and Ni-Fe based developed alloy, specimens were prepared for three types of SFNi4, a conventional material. Table 1 shows the composition of each alloy and Fig. 2 shows photographs of the appearance of the specimens. SFNi4, a conventional material, is characterized by its ease of application (wide semi-melting temperature range) and forms a very smooth coating. On the other hand, the Ni-based developed alloy could be applied without any problems, although the application difficulty increased slightly due to the slightly narrower semi-melting temperature range. The Ni-Fe based developed alloy, however, became more difficult to apply as the melting point increased due to the higher Fe content, and granular protrusions formed on the surface, resulting in a coating that was considered to be visually incompatible. We conducted the exposure test for six months in the hope that these protrusions would be removed early by the impingement of the flow medium in the actual equipment and would not cause any practical problems.
| Ni | Cr | Fe | Mo | Si | B | C | Cu | |
| SFNi4 Conventional material |
Remaining | 12-17 | ≦5 | ≦4 | 3.5-5.0 | 2.5-4.0 | 0.4-0.9 | ≦4 |
| Ni-based developed material |
Remaining | 20 | 4 | 1 | 1 | 5.5 | 0.5 | － |
| Ni-Fe based developed material |
Remaining | 20 | 30 | － | 1 | 5.5 | 0.5 | － |
Fig. 2 Photos of the appearance after the remelting treatment
Fig. 3 shows the results of the exposure test. As with the conventional material, no major changes were observed in the Ni-based developed alloy, and no concern regarding risk arose when applying it to actual equipment. We expected the Ni-Fe based developed alloy to lose the protrusions and have a smooth appearance, but the unevenness remained unchanged after exposure.
Fig. 3 Photos of the appearance of the actual equipment after six months of exposure
Although Ni-Fe based developed alloys are expected to have high environmental resistance based on the results of laboratory tests, solving the problem of the formation of granular protrusions on the surface has emerged as an issue for product commercialization.
In the previous report, we reported that the concavo-convex surface shape, caused by the difference in thinning rates between the precipitates in the sprayed alloy and the base metal, contributes to the improvement in erosion-corrosion resistance. Since it was confirmed that the granular protrusions were not easily removed even in the actual equipment, we considered using this concavo-convex shape. In other words, instead of suppressing granular protrusions, we intentionally made the entire surface uneven. Observations at the time of manufacture confirmed that granular protrusions were formed during solidification in the remelting treatment. In other words, the remelting treatment conditions were found to have an effect, so experiments were conducted by changing the temperature of the remelting treatment. Fig. 4 shows the results. The state of the granular protrusions changed with temperature, with lower temperatures resulting in higher densities. The sectional observation results show that the granular protrusions are mushroom-shaped protrusions that are integrated with the base metal and do not fall off easily. From these results, it was found that the amount produced can be controlled by the remelting temperature, and a control temperature range was set based on the density of the protrusions.
Fig. 4 Dependence of remelting temperature on the generation of granular protrusions
In the actual equipment, we used an in-bed tube sprayed with self-fluxing alloy over carbon steel, and we experienced rapid thinning, as the sprayed coating disappeared and the carbon steel was exposed. In the demonstration test, about 2 mm of SUS309 material was built up on carbon steel and a thermal spray coating was applied on top of the carbon steel, taking into account the risk of thinning. Fig. 5 shows photographs of the manufactured heat transfer tubes. In addition to the two developed materials, a conventional material (SFNi4) was used to manufacture a total of three types of in-bed tubes. The Ni-based developed alloy has an appearance similar to the conventional material, while the Ni-Fe based developed alloy has a dark brown appearance with granular protrusions. In addition, stripes can be seen in both cases, but these patterns are caused by the unevenness of the build-up layer applied as a base and have nothing to do with the quality of the thermal spray coating. Three panels were manufactured for each of these test devices and installed in the in-bed tube unit (Fig. 6). In the actual equipment, the thinning tendency varied depending on the location, so three panels were installed in locations assumed to have the same conditions. The conventional material and the Ni-Fe based developed alloy were placed symmetrically across the panel gap, and the Ni-based developed alloy was placed next to the conventional material.
Fig. 5 Photos of the appearance of the in-bed tube for demonstration
Fig. 6 Demonstration test panel mounting position
The demonstration test was conducted for about one year at a biomass power plant operating at a bed temperature of about 700 °C and a heat transfer tube surface temperature of about 300 °C. Evaluations were made by measuring the thickness with an ultrasonic thickness gauge before and after the test, and by observing the cross section after exposure.
Fig. 7 shows the evaluation results of the amount of thinning after the one-year demonstration test. These are the measurement results of a total of 92 points, 23 fixed points and 4 points in the circumferential direction, before and after the test using an ultrasonic thickness gauge. Although the amount of thinning varies greatly depending on the location, the tendency of thinning of the entire panel was that the Ni-based developed alloy increased the amount of thinning in both the maximum value and the average value compared to the conventional material, resulting in an inferior result. Since the surface of the Ni-Fe based developed alloy is uneven, the surface of the fixed-point measurement area was polished in advance to remove granular protrusions before the exposure test. The amount of thickness reduction in the polished area was less than that of the conventional material, and the average value was about half that of the conventional material.
Fig. 7 Amount of thickness reduction after the demonstration test
Figure 8 shows photos of the appearance of the conventional material and the Ni-Fe based developed alloy after exposure. The surface of the conventional material is smooth and there are no noticeable deposits. On the other hand, in the Ni-Fe based developed alloy, granular protrusions remained on the entire surface, although they appeared smaller in size. Observing the cross section after cutting (Fig. 9), we found that the surface was covered with two layers of deposits of about 30 μm around the granular protrusions. The sectional photographs show that a single layer of deposit was observed in the polished area where the protrusions had been removed. Fig. 10 shows the results of these EDS analyses. In the uneven area (Fig. 10-1), an outer layer composed mainly of oxides such as Si and Ca and an inner layer composed of oxides such as Ni, Fe and Cr were formed. On the other hand, in the polished area (Fig. 10-2), oxide layers such as Ni, Fe, and Cr were not observed, and salt components such as Na were present as deposits. In that form, many crack-like shapes were observed, which appeared during the solidification of the melt.
Fig. 8 Photos of the appearance after the demonstration test
Fig. 9 Photos of the Ni-Fe based developed alloy cross section
Fig. 10-1 Cross sectional EDS analysis results of Ni-Fe based developed alloy (uneven area)
Fig. 10-2 Cross sectional EDS analysis results of Ni-Fe based developed alloy (polished area)
The cross sectional observation confirmed that thinning had not progressed in the uneven area. These results suggest that during the period when the granular protrusions are present, the protrusions are preferentially thinned and thinning of the base metal is limited.
From the results of the demonstration test, it is judged that the Ni-based developed alloy has erosion-corrosion properties inferior to the conventional material in this environment. For the Ni-Fe based developed alloy, the thinning was suppressed by the granular protrusions, and the thickness of the polished area was about half that of the conventional material. Fig. 11 shows this. Based on past performance, conventional alloy life is about one year. In the Ni-Fe based developed alloy, there is almost no base metal thinning during the period when granular protrusions are present, and the protrusions are expected to remain for about one year. The thickness of the sprayed coating is estimated to be about half that of the conventional material, so if the coating thickness is equivalent to that of the conventional material, the life is expected to be about three years.
Fig. 11 Life evaluation of the developed alloy
A secondary effect of the granular protrusions may be to simplify the inspection process. Since it is not possible to identify the thinning area in the current in-bed tube from the outside, the thinning is usually repaired after measuring the thinning of the entire tube. In the case of the Ni-Fe based developed alloy, it is possible to identify the damaged area by the presence or absence of protrusions by looking at the appearance. This greatly reduces the inspection process prior to repair and is expected to effectively reduce maintenance costs.
The main feature of the Ni-Fe based developed alloy, which showed excellent results in the demonstration test, is the granular protrusions on the surface. The formation of these protrusions was not expected at first, but was the product of chance. The following is a discussion of how the shape of protrusions works for erosion-corrosion resistance and why protrusions are formed.
From the results of cross sectional observation (Figs. 9, 10-1, 10-2), the surface deposits of the thermal spray material varied depending on the presence or absence of protrusions. The polished area where the protrusions were removed had deposits that appeared to be salt components composed of Na, etc., and judging from the solidification traces, there is a possibility that these deposits existed as molten salt. It is unlikely that this molten salt layer, which exists as a liquid, has any protective function against the erosion of the continuously impinging flow medium. On the other hand, in the same Ni-Fe based developed alloy, an oxide layer consisting of Ni, Fe, Cr, etc., originating from the sprayed alloy was formed under the deposit layer (on the sprayed alloy side) consisting of Si, Ca, etc. originating from the flow medium in the area where the protrusions existed. In other words, the granular protrusions may have contributed to the stable growth of corrosion products.
In the previous report, we reported that the microscopic unevenness of the surface caused by the difference in thinning rate between the base metal and the precipitates contributed to the stable growth of corrosion products1
）. The macroscopic concavo-convex shape in this case is considered to have further enhanced this effect, with the convexity suppressing the impingement of the flow medium on the base metal and significantly mitigating the erosion conditions in the base metal. As a result, corrosion products grow stably on the surface of the base metal, thereby suppressing the corrosion rate, which in turn has led to the suppression of the thinning rate (Fig. 12).
Fig. 12 Mechanism of thinning suppression by granular protrusions
Granular protrusions were not generated in the Ni-based developed alloy, but only in the Ni-Fe based developed alloy. To investigate the reason for this, the thermal properties of both alloys were investigated by TG-DTA measurements. Figure 13 shows the thermal properties as the temperature rises. The endothermic peak around 1 050 °C corresponds to the Ni-Ni3B eutectic temperature. On the other hand, only the Ni-Fe based developed alloy has a peak around 1 140 °C, which corresponds to the melting of γ-Ni. In other words, this γ-Ni is thought to be related to the generation of protrusions. As a result of a solidification experiment using prototype alloys with different γ-Ni contents, it was found that the higher the γ-Ni content, the greater the number of protrusions2 ）
Fig. 13 TG-DTA measurement results of the developed alloy
As a form change during solidification, cases of volume expansion caused by the release of dissolved gases in the molten metal during solidification have been reported3 ）. In this study, we assumed that the same gas generation was the cause, and that γ-Ni crystallized from the molten metal, dissolved gas was released at this stage, and the generated air bubbles extruded the remaining molten metal, resulting in the generation of the concavo-convex shape (Fig. 14). To verify this assumption, the Ni-Fe based developed alloy was melted and solidified in air and in vacuum, respectively. As expected, protrusions were generated in air, but not in vacuum (Fig. 15). Based on this hypothesis, we have also confirmed that even in the Ni-based developed alloy where granular protrusions have not appeared in the past, protrusions are generated during solidification by adjusting the alloy composition so that γ-Ni is crystallized. From the above, we have succeeded in establishing a control technology that can explain the generation of granular protrusions without contradiction.
Fig. 14 Granular protrusion formation mechanism
Fig. 15 Verification of the granular protrusion formation mechanism
In order to put the alloy developed in the laboratory into practical use, we worked to make it a self-fluxing alloy. Contrary to the laboratory results, the Ni-based developed alloy was inferior to the conventional material in the actual equipment.
The Ni-Fe based developed alloy showed a peculiar property that, unlike normal self-fluxing alloys, granular protrusions were generated on the surface during the remelting treatment. Anticipating the improvement of high-temperature erosion-corrosion resistance through the concavo-convex surface shape, we have developed a technology to actively generate granular projections and form a concavo-convex surface shape over the entire surface. As a result of the demonstration test, we have succeeded in obtaining high-temperature erosion-corrosion resistance properties that far exceed those of conventional alloys, due to the function of the concavo-convex surface shape. It was also clarified that the generation of granular protrusions is due to the release of gas associated with the crystallization of γ-Ni. The newly developed Ni-Fe based developed alloys are currently being introduced to the market, and further application expansion beyond in-bed tubes is being considered.
This achievement was awarded the FY2021 Technology Award from the Japan Society of Corrosion Engineering in recognition of its industrial value and originality in the generation of granular protrusions. We would like to express our gratitude to Dr. Mohammad Emami, Hokkaido University, Hokkaido Research Organization, Dai-ichi High Frequency Co., Ltd., and everyone involved in the company for their great cooperation in carrying out this research. We would like to take this opportunity to thank you.
1) Manabu Noguchi, Eiji Ishikawa, Takashi Kogin, Suzue Yoneda, Shigenari Hayashi: Development of High-temperature Erosion-corrosion Resistant Thermal Spray Alloy for Biomass Fuel (Report No.1: Alloy Development through Laboratory Testing), Ebara Engineering Review, No. 263, P33-40 (2022)
2) Sena Tanaka, Shigenari Hayashi, Suzue Yoneda, Takashi Kogin, Eiji Ishikawa, Manabu Noguchi: Formation Mechanism of Concavo-convex Surface of Ni-based Self-fluxing Alloy Coating with High Fe Content for Fluidized Bed Boiler Heat Transfer Tubes, Proceedings of the 67th Japan Conference on Materials and Environments, B-318, P.197-198 (2020).
3) Yoji Awano, Kazufumi Morimoto, Yoshihiro Shimizu, Hiroyuki Takamiya: Shinkage Morphology in Al-Si System Casting Alloys, R&D Review of Toyota CRDL, Vol.27, No.1 P51-62, (1992).