1 TATA Scunthorpe Queen Anne- taphole suction hood optimisation by CFD simulation and physical modelling Author(s) Name(s) and Affiliations(s) Bob PETERS, Paul Wurth S.A. Luxembourg, Jean-Paul SIMOES, Paul Wurth S.A. Luxembourg, Sigurd RÖDL, VDEh Betriebsforschungsinstitut GmbH, Kersten MARX, VDEh Betriebsforschungsinstitut GmbH, Dave COLLINS, TATA UK, Scunthorpe works Stuart WOODLIFFE, TATA UK, Scunthorpe works Contact data Bob PETERS, Senior Project Manager, Paul Wurth S.A. Luxembourg, 32, rue d Alsace, L-1122 Luxembourg, Tel: +352 4970 2724, e-mail: bob.peters@paulwurth.com Summary Environmental requirements have become increasingly stringent during the past years in all industry branches. TATA Scunthorpe recently relined their Blast Furnace Queen Anne and completely redesigned their casthouse floor. Creating a safer working space and reducing the dust dispersion were main concerns driving the project. Changing the existing filter and fans were out of scope and only a limited suction capacity was available. Paul Wurth S.A. proposed modification of the existing ducting system and optimisation of the taphole hood. Together with BFI (VDEh Betriebsforschungsinstitut GmbH) the complexity of the construction was analysed and an optimal solution was developed. Measurements and observations in the plant, physical model tests, and calculations from computational fluid dynamics software were utilized to generate the basic information required to produce a layout of the secondary dedusting system. These efforts also contributed to analysis and optimisation of overall fume capture efficiency for the exhaust system. Key Words Blast Furnace, casthouse, dedusting, taphole hood Objectives A main topic of study for the casthouse redesign for blast furnace Queen Anne at TATA Steel in Scunthorpe was modification of the dedusting system. Observations in the plant showed that the initial dedusting installation was insufficient and that uncontrolled emissions were escaping the casthouse. From time to time the system behaviour resulted in unacceptable working conditions on the cast floor. Paul Wurth S.A., Luxembourg, was commissioned by TATA Steel to study a new dedusting facility, with a constraint of retaining the existing fan and filter units. Paul Wurth began collaborating with Dusseldorf, Germany based VDEh Betriebsforschungsinstitut GmbH (BFI) in order to survey the performance of the fume extraction system and to generate the basic information necessary for possible optimisation of the new casthouse dedusting system. The basic requirements for layout of the secondary dedusting system and optimisation of fume capture efficiency in the exhaust system were realized following: measurements and observations in the plant, numerical simulation with computational fluid dynamics software, and through use of a physical model. Major design sequences Plant observation In the first stage it was important to analyse the performance of the existing dedusting installation. This required elaboration of the extents of problems:
2 - Fan capacity - Layout of the extraction points and dedusting hoods - Influence of cross flows typical for blast furnace casthouses. During several days, a special survey was organised: fume propagation in the vicinity of the main runner was observed and documented by video cameras; air flow was made visible by using artificial smoke. that of Taphole 1. In operation, a lot of smoke was created during opening of Taphole 2. In spite of the fact that Taphole 2 was equipped with a side draft-hood (which worked quite well), a major portion of the fumes escaped and were not captured by the main hood. It appeared that the main hood was not working at all. While Taphole 2 was closed, cross flows were also visualised with artificial smoke. A slight drift of air from the left to right was observed. On the first day, work comprised of documenting the current practice. On the second day, false airflow in the vicinity of the main runner was visualized using artificial smoke. An alternating tapping technique was used, switching from one taphole to another to create an opportunity to analyse both tapholes during the same visit. Taphole 1 At Taphole 1 fume capture was fairly good when levels of fume generation were low, but it was insufficient during peak conditions. There was frequent, excessive fume spillage from the taphole area at times of peak fume generation. Typically, the fume flow rate reaches its maximum at the end of the cast, when part of the hot blast escapes through the taphole. Emissions during the observation were moderate while drilling. The flow of false air was visualised with artificial smoke. Cross flows in the casthouse can generally reduce the efficiency of an extractor hood at the tap hole. In this case no major cross winds were detected, see Figure 1a. Figure 1b: Fume and false air observation TH2 BFI uses advanced image analyses software ( Structural PIV ) for the quantification of fume flow rates. The Structural PIV methodology differs in a number of ways from the classical PIV (Particle Image Velocimetry) technique. One differentiator would be; where classical PIV requires a laser to create a light sheet, Structural PIV can utilize white light. A second difference is in the seeding method used. Classical PIV relies on use of tracer particles. Structural PIV on the other hand works with small structures inherent in smoke clouds. The analysis of images of smoke clouds was performed through ensemble correlation averaging, making it possible to receive a mean flow field with a sufficiently high signal-to noise ratio. Figure 1a: Fume and false air observation TH1 Taphole 2 As a result of alternating tapping during the survey, it was possible to study Taphole 2 on the same day as Taphole 1. Dedusting system geometry of Taphole 2 was totally different from Advantages of Structural PIV are that it can be used on full-scale objects in the plant, fewer safety precautions are required, and the equipment is less expensive. Figure 2a and 2b illustrate a determined flow field for the end of tapping process phase in two selected regions. The results were used as a boundary condition for the planned numerical simulations and model trials.
3 Figure 3a: Main hood in working position Figure 2a: Flow field determined by structural PIV Figure 3b: Main hood in taphole maintenance position Figure 2b: Flow field determined by structural PIV A further design challenge was to connect the hood to the ductwork routed below the castfloor. Between the three existing taphole machines very little space was available (see Figure 4). Construction constraints Compared to the previous hood design, the new one needed to be closer to the main iron runner, in order to allow maximum efficiency and also ensure access to the casthouse machinery. Paul Wurth developed a movable, rotating hood. This hood was designed to operate in a horizontal position during the dedusting process. It also had the ability to be lifted to an inclination required to allow access by heavy machinery used during taphole maintenance and cleaning activities (see Figures 3a + 3b) Figure 4: Area available for hood-main ductwork connection Figure 5 demonstrates the complexity of the connection ductwork, which had to be designed to fit between the existing structures and new machines implemented on the castfloor.
4 flow rates were fixed according to the planned flow rates for different operating cases. The complete geometry was also fed into the numerical model, and so, could be optimised based on the results provided by CFD simulation. Figure 5: Showing the complexity of the connection part Preliminary numerical calculations From a takeover point located outside the casthouse, the fume ducts were completely relocated and the routing was changed to run below the castfloor elevation. This strategy increased the working zone and improved the safe working conditions on the castfloor. Previously, the ductwork was fixed onto the steel work of the cast house building. Due to the lack of access, cleaning and inspection efforts were laborious and rarely realized. For a representative part of the newly designed ductwork the flow, temperature and concentration field was simulated using Computational Fluid Dynamics (CFD) software. Figure 6: Flow simulation of ductwork In the first design iteration a movable hood consisting of a plate with a central trumpetshaped channel was planned. Just above the tap hole an additional rectangular opening to capture the rising fume from the tap hole was planned. The fume capture was incomplete. Fumes escaped at the sides of the hood, through the gap near the movable cylindrical part of the hood and underneath the tuyere platform (see Figure 7). Applying the Finite-volume method to the geometry in question, the domain of integration was subdivided into a large number of control volumes (cells) resulting in the generation of a mesh. Differential equations for the conservation of mass momentum, energy and species were transformed by discretisation measures into linear equations for the individual cells. The variables in the domain of integration as a whole were ascertained iteratively by building the ultimate solution step-by-step from preceding results. The shape and number of the cells must be suitably established in order to come as close as possible to the exact solution of the differential equations. The planned ducting systems of Tapholes 1 and 2 were studied with numerical simulation. The Figure 7: Incomplete fume capture with first hood design To achieve better fume capture at the hood additional suction nozzles (slots) were added at the sides of the hood (see Figure 8).
5 Adding a slot in the main cylindrical duct considerably improved the situation (see Figure 11). Figure 8: Improved fume capture after with hood design modifications This iteration was still unsatisfactory, so an additional slot was added in the main horizontalcylindrical part of the hood to capture the fumes rising between the hood and the tuyere platform (see Figure 9). Figure 11: Calculated velocity field with additional slot Physical model Figure 9: Hood with additional slots in the main duct Based on the most efficient variant identified by CFD results, a representative section of the casthouse with the dedusting system was simulated using a three-dimensional isothermal model. As a prerequisite for the transfer of measuring results from the model to the large-scale design, a 1:5 scale was adopted for construction of a physical model and was observed for all components. Next, the area close to the taphole had to be modified because the cross section of the channel and the radius of the bend were too small. The velocities were far over the expected limits. These specific areas were re-evaluated and optimized as much as the surrounding structures would allow. Figure 12: Test rig The movable hood consisted of a plate with a central trumpet-shaped channel and two lateral slits. An additional rectangular opening was located for the capture of the fumes rising from the taphole. Figure 10: Areas of very high velocities
6 Heated ambient air was introduced in the model so that the influence of temperature gradients could be simulated qualitatively. To visualise flow, a special fog was added to the simulated fume gas. The fume flow rate was chosen so that fume propagation in the model was similar to the actual state in the plant. This was maintained as a constant for each process stage to enable comparison between the actual and optimised situations. Initial tests showed that some of the fumes were escaping under the tuyere platform to the sides. Additionally, some fumes were escaping between the cylindrical channel and the tuyere platform. As a result, further tests examined the influence of an additional slot in the cylindrical channel. The following trials replicated the final design studied by CFD and arrived at the same conclusion. The fumes captured in the gap between cylindrical channel and tuyere platform was satisfactory, but fumes still escaped under the tuyere platform to the sides. The performance of the suction channel near the taphole was not optimal. Fumes in the direct vicinity of the suction opening were captured. Meanwhile, fumes reaching the lower surface of the tuyere platform were escaping. Figure 13: Geometry of the old (a) and new (b) suction channel near the taphole Figure 14: New geometry of the suction channel on the hood. The trials showed that modification of the additional suction channel on the hood did not improve fume capture performance. Even if the five additional tubes above the taphole were closed the small trumpet drew mostly false air. Figure 12: test rig side view The physical model could be adapted very easily, enabling fine tuning of the suction capacity in particular areas. This was especially important where the area above the taphole spool had to be analysed and optimised. Through analysis, the suction channel near the taphole was modified in the model. As an alternative to the rectangular channel in the vicinity of the taphole, five tubes were installed in the region of the tuyere platform. Figure 13 shows the old and the new geometry in comparison. The new part is coloured in blue. The best fume capture was achieved when the small trumpet was closed, the rectangular slot in the cylinder was open, and the five tubes above the tap hole were open (see Figure 15). Figure 15: test rig side view Furthermore an additional trumpet shaped duct was integrated in the flat part of the main hood (see Figure 14).
7 Conclusion Observations in the plant, physical model tests, and calculations using computational fluid dynamics were all aids in creating the basic information required for optimisation of the Queen Anne casthouse dedusting system. The information published in this paper only summarizes results of the most significant tests and calculations. A hood and piping system was developed, which should be able to capture emissions to a large extent. Using both CFD and a physical test model were providing very similar results, and with these results an optimum design was achieved. It is expected that the proposed solution will clearly improve working conditions in the casthouse and will also contribute to a reduction of emissions to the environment.