Table of Contents Author Guidelines Submit a Manuscript
Journal of Engineering
Volume 2013 (2013), Article ID 719296, 8 pages
Research Article

Experimental Determination of Heat Transfer Coefficient in Stirred Vessel for Coal-Water Slurry Based on the Taguchi Method

1Mechanical Engineering, Annamalai University, Annamalai Nagar, Tamil Nadu 608 002, India
2Vice Principal, Sri Krishna college of Engineering and Technology, Coimbatore, India

Received 31 October 2012; Revised 3 May 2013; Accepted 8 May 2013

Academic Editor: You-Rong Li

Copyright © 2013 C. M. Raguraman et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Heat transfer in stirred vessels is important because process fluid temperature in the vessel is one of the most significant factors for controlling the outcome of process. In this study, the effects of some important design parameters for coal-water slurry in agitated vessel used in coal gasification such as stirrer speed, location of stirrer, D/d ratio, and coal-water ratio were investigated and optimized using the Taguchi method. The experiments were planned based on Taguchi’s orthogonal array with each trial performed under different levels of design parameter. Signal-to-noise (S/N) analysis and analysis of variance (ANOVA) were carried out in order to determine the effects of process parameter and optimal factor’s level settings. Finally, confirmation tests verified that the Taguchi method achieved optimization of heat transfer coefficient in agitated vessel.

1. Introduction

Research on heat transfer coefficient in agitated vessel is still critical and ongoing. Heat transfer in stirred vessel is important because process fluid temperature in the vessel is one of the most significant factors for controlling the outcome of process. Mechanically agitated vessels are widely used in mining, food, petroleum, chemical, pharmaceutical, pulp, and paper industries and are also used in coal gasification power plant [1]. The intensity of heat transfer during mixing of fluids like coal slurry depends on the type of the stirrer, the design of the vessel, and condition of the processes [2]. In this study the effects of some important parameters such as stirrer speed, location of stirrer, D/d ratio, and coal-water ratio were investigated and optimized.

Performing an experiment is more suitable for determination of the real performance characteristics of a system. However to prepare an experimental setup is very expensive and some systems cannot be constructed and tested in a laboratory. Also, preparing an experimental setup is a very time-consuming procedure because of the high trial numbers. Because of these difficulties, the modeling and then testing the system using numerical analysis, ANN (artificial neural network), or optimizing the trial numbers according to the Taguchi method is more appropriate and very popular nowadays [3].

Heat transfer rates in agitated vessel have been investigated for coal-water slurry in a flat bottom vessel equipped with flat-blade impeller making an angle of 45 degree to the axis of the shaft. Also the heat transfer coefficients of the flat-blade impeller parallel to the axis of the shaft results [4] were compared.

2. Experimental Detail

2.1. Experimental Setup

The schematic diagram of the experimental setup is shown in Figure 1. An agitator is driven by a vertically mounted motor which is used for stirring. The entire vessel is completely insulated. The inlet of the jacket is connected to an electrically operated boiler, which produces steam continuously at constant pressure. A pressure gauge is connected to the jacket for monitoring the jacket steam pressure. A vent is provided for releasing noncondensable gas and to maintain the pressure in jacket. A steam strainer is placed at the bottom of the vessel to collect the condensed steam.

Figure 1: Experimental setup.

A jacketed mild steel cylindrical vessel of 210 mm diameter, 6 mm thick, and 160 mm height with a flat bottom was used. Similarly the cylindrical inner vessel of 110 mm diameter, 6 mm thick and 110 mm height with similar flat bottom was used. This assures that the height of vessel is always equal to diameter of the vessel [4]. The impeller used in the jacketed vessel is having four flat blades (Figure 2) and is driven by a variac controlled motor. The impeller makes an angle of 45° to the axis of the shaft. The thickness of each blade is 1.6 mm and having a width of 17.5 mm. Outlets are given at the bottom side of the agitated vessel for collection of samples.

Figure 2: Details of Flat-blade impeller. All dimensions are in “mm.”

The cylindrical vessel of 210 mm diameter, 6 mm thick, and 160 mm height with a flat bottom was used which is named as jacketed vessel. Similarly the cylindrical inside vessel of 110 mm diameter, 6 mm thick and 110 mm height with same flat bottom was used. So the height of vessel is equal to diameter of the vessel [4]. The impeller used is four flat blade whose details as given in Figures 2(a)2(c).

2.2. Experimental Procedure

Initially the drain valve provided at the bottom of the jacketed vessel and gate valve 1 was closed. The vessel was first filled with a known volume of water and mix with coal particle (size: BSS 36). The initial temperature of the slurry (t1  °C) can be noted through the digital temperature indicator (Micro sensor Make, T-type), which was connected to a thermocouple (T-type, copper constant). The boiler was filled with required amount of water. Now the gate valve 2 was closed and the boiler was switched on. After steam was produced gate valve 3 was closed. The variac (Continuously variable auto transformer, Automatic Electric Ltd, Range 0 to 270 V) was used to vary the stirrer speed. When the boiler steam pressure reaches 2.5 kg/cm2 (gauge)-(Bourdon type, Gurn Make) open the gate valve 2 and allow the steam into the jacketed vessel. Now keep the pressure gauge at 2.0 kg/cm2 (gauge) by adjusting the gate valve 2 and start the stopwatch. When the temperature reaches the saturation temperature of the water (100°C) the time period ( sec) was noted and this period is called heating period (unsteady state). The above procedure was repeated for other sets of readings.

3. Preparation of Coal-Water Slurry

Coal gasification is a key process in integrated and coal gasification combined cycle (IGCC) power plant. Being one of the most competitive and promising coal gasification technologies, the slurry feed-type entrained-flow coal gasification process is shown in Figure 3, which has been extensively used in the foreign countries. Coal-water slurry (CWS) is a mixture of water and coal. First pulverize coal mix with water according to proper proportion. (coal : water = (40 to 70) : (60 to 30) and then add a little additive (about 1% of the total soiled weight; the coal-water two-phase flow is formed after strong agitation [6]. In an entrained-flow gasifier using coal-water slurry, it should be necessary to uniformly mix the coal-water slurry with oxygen to obtain the higher carbon conversion in short residence time (0.4–5 s).

Figure 3: Entrained-flow coal gasification process—slurry feed type [5].

The proximate and ultimate analysis of the Indian coal used in the preparation of the slurries is given in Table 1.

Table 1: Coal analysis details.

Additive is one of the key factors affecting CWS quality. Usually additive dosage accounts for about 1% of the total weight. Based on function, additives can be used like dispersant and stabilizing agents which are used in this project, that is, sodium carbonate and sodium salt of carboxyl methyl cellulose (Na-CMC), respectively. The best dosages of dispersant and stabilizers are 0.75% by wt. of solids and 0.1% by wt. of total solids [7].

The concentration of coal slurry could be maintained from 40% to 70% which is able to feed into the gasification without any feeding problem [8], so that the concentrations chosen are 40%, 50%, and 60%. The coal particle size used for coal-water slurry preparation is BSS 36 (0.422 mm).

4. Uncertainty Analysis

The objective of a measurement is to determine the value of the measurand, that is, the value of the particular quantity to be measured. A measurement therefore begins with an appropriate specification of the measurand, the method of measurement, and the measurement procedure. In general, the result of a measurement is only an approximation or estimate of the value of the measurand and thus is complete only when accompanied by a statement of the uncertainty of that estimate [9]. Through there are many factors in the measurement uncertainty, in this study, it is assumed that the major factors of resolution or detection of sensors and the variation of the measured data during repeated tests at the test condition. It defined the measurand, the output quantity as a function of the input quantities .

The uncertainties of the measured data were calculated by combining the type A and B [10]. The type A uncertainty was evaluated by statistical analysis of series of observation of the 50 times sampled data and the type B uncertainty was calculated by previous measurements, specifications from the manufacturer, hand-books, calibration certificates, and so forth [11]. The standard uncertainty was computed as root mean square error (RSM) of the type A and B uncertainties. Table 2 shows the uncertainty estimation of each measured parameter and 50 samples were collected in each test [11].

Table 2: Uncertainty estimation of variables.

5. Application of Taguchi Method

The quality engineering methods of Taguchi, employing design of experiments (DOE), is one of the most important statistical tools for designing high quality system at reduced cost. Taguchi methods provide an efficient and systematic way to optimize designs for performance, quality and cost. Optimization of process parameter is the key step in the Taguchi method to achieve high quality without increasing cost [12]. Taguchi’s parameter design method is a powerful tool for optimizing the performance characteristics of a process [13, 14]. The experiments were designed based on the orthogonal array technique. Using the Taguchi experimental design, an orthogonal array was selected for four factors and three levels for each factor. The orthogonal array selected for this study is the orthogonal array which is shown in Table 2.

Based on orthogonal arrays, the number of experiments that may cause an increase in time and cost can be reduced by means of the Taguchi technique. It uses a special design of orthogonal arrays to learn the whole parameter space with a small number of experiments [15, 16].

The selection of control factors and their levels is made on the basis of some preliminary trial experiments conducted in the laboratory and also from the literature review on the subject. The choice of three levels has been made because the effect of these factors on the performance characteristics can be estimated simultaneously while minimizing the number of test runs. An (34) standard orthogonal array [14] as shown in Table 3 was employed for the present investigation. This array is most suitable to provide the minimum degrees of freedom as 9 required for the experimental exploration.

Table 3: Factors and their specified levels.

The experimental layout for the process parameter using the (34) orthogonal array is shown in Table 4. The application steps in the Taguchi method are given in Figure 4.

Table 4: (34) standard orthogonal array.
Figure 4: Application steps in the Taguchi method.

6. Results and Discussion

The heat transfer coefficient was determined as per experiment runs. Three readings (corresponding to the three replications) are recorded for each experimental condition as shown in Table 5.

Table 5: Experimental layout using an (34) orthogonal array.

6.1. Statistical Analysis

Statistical analyses (over all loss function, S/N ratio, ANOVA) are carried out for a significance level of 0.05, that is, for a confidence level of 95%. The statistical errors in the experiment are normally distributed in normal probability plot as shown in Figure 5.

Figure 5: Normal probability plot (response is HT coefficient).
6.2. Overall Loss Function

A loss function is then defined to calculate the deviation between the experiment value and the desired value. The control factors that may contribute to reduce variation (improved quality) can be quickly identified by looking at the amount of variation present as a response. Taguchi recommends the use of the loss function to measure the deviation of the quality characteristics from the desired value. The four control variables belong to the higher-the-better quality characteristic. The loss function of the-higher-the-better quality characteristic can be expressed as follows.

The higher is better where is the loss function of the th quality characteristic in the experiment, is the number of tests, and is the experimental value of the th quality characteristic in the th experiment at the th test [9]. As a result, four quality characteristics corresponding to the four control variables and heat transfer coefficient are obtained using (1).

6.3. Signal-to-Noise Ratio

In Taguchi technique, the variation of the response is also examined using an appropriately chosen S/N ratio. Broadly speaking, the S/N ratio is the ratio of the mean (signal) to the standard deviation (noise). These S/N ratios, derived from the quadratic loss function, are expressed on a decibel (dB) scale. The formula used to compute the S/N ratio depends on the objective function. The overall loss function is further transformed into the signal-to-noise (S/N) ratio [9]. In the present study, heat transfer coefficient is a “the higher the better” type of quality characteristic since the goal is to maximize the heat transfer coefficient. The S/N ratio in th experiment can be expressed as follows:

The signal-to-noise ratios corresponding to overall loss function are computed using (2) for each of the nine experimental conditions and are reported in Table 5. Since the experimental design is orthogonal, the factor effects can be separated out in terms of the S/N ratio and in terms of the mean response. The average values of S/N ratios of the four control factors at each of the four levels are shown in Figure 6, and from which the levels corresponding to the highest S/N ratio values are chosen for each parameter representing the optimum condition. Here, the optimum condition corresponds to the maximization of the heat transfer coefficient. It is clear from Figure 6 that the optimum levels are stirrer speed (300 rpm), stirrer location (55.0 mm), D/d ratio (2), and C-w ratio (60 : 40), respectively.

Figure 6: Main effects plots for S/N ratios of coal-water slurry.

In addition to S/N analysis, main effects of the process parameters on the mean response are also analyzed. The mean response refers to the average value of the quality characteristic for each factor at different levels. Thus, the average values of the heat transfer coefficient for each factor at the three levels have been calculated and also indicated the same optimum level of the parameters as obtained in S/N ratio analysis.

6.4. Analysis of Variance (ANOVA)

ANOVA is a method most widely used for determining significant parameters on response and measuring their effects. Table 6 shows the computed results of the ANOVA with 95% confidence. In ANOVA, the ratio between the variance of the process parameter and the error variance called as -ratio determined whether the parameter has significant effect on the quality characteristic. This process is carried out by comparing the -ratio value of the parameter with the standard table value at the 5% significance level. If the -ratio value is greater than , the process parameter is considered significant. Depending on it, it can be seen that the effects of all the factors on heat transfer coefficient are significant. The last column of Table 6 indicates the percentage contribution (significance rate) of each process parameter of the total variation, indicating their degree of influence on the results. According to Table 6 D/d ratio has the most dominant effect on total variation and it is followed by stirrer speed , C-w ratio , and stirrer location . However, C-w ratio yields the most significant effect on the measured response as shown by the much higher -ratio (256.51) and also percent contribution (47.60). The percentages of contribution of each variable are shown in Figure 7.

Table 6: Results for quality characteristics and signal-to-noise ratio.
Figure 7: % of contribution of each design variables.
6.5. Prediction of Optimum Heat Transfer Coefficient

From the analysis of S/N ratio and the mean response characteristic, the optimum levels of the control factors are determined as stirrer speed3, stirrer location2, D/d ratio2, and C-w ratio3. Hence, the predicted mean of the quality characteristic (heat transfer coefficient) has been computed as shown by [17] as per (3): where = predicted optimum heat transfer coefficient and is the overall average of all the experimental data for heat transfer coefficient.

The calculated values of various response averages are  W/m2 K,  W/m2 K,  W/m2 K,  W/m2 K, and C-w  W/m2 K. So substituting these values in (3), the mean optimum value of the heat transfer has been predicted as  W/m2 K.

6.6. Confidence Interval (CI) for Predicting a Confirmation Experiment

A confirmation experiment is used to verify that the factors and levels chosen from an experiment cause a product or process to behave in a certain fashion. A selected number of tests are run under constant, specified conditions to observe results that, the experimenter hopes, are close to the predicted value. Thus a 95% confidence interval (CI) for the predicted mean of optimum heat transfer on a confirmation test is estimated using (4) [18]: where is the -ratio required 100 (1−) percent confidence interval, is DOF for error, is the error variance, is number of replications for confirmation experiment (=3), and is effective number of replication. is total number of experiments (=27 (9 × 3)) and is the total degrees of freedom (=8 (2 × 4)) associated with the estimate of mean optimum. From Table 6, the values are ,, and from standard statistical table in (15), the required -ratio for = 0.05 :  (0.05, 1, 18) = 4.41. Substituting these values in (4) and (5), the calculated confidence interval is . Thus the 95% confidence interval of the predicted optimal heat transfer coefficient is obtained as 707.99 Wm-2 K−1; that is, 695.52 W/m2 K < 707.99 W/m2 K < 720.46 W/m2 K.

6.7. Confirmation Experiment

In order to test the predicted result, confirmation experiment has been conducted by running other three replications at the optimal settings of the process parameters determining the analysis.

The results are shown in Table 7 and it is observed that the mean heat transfer coefficient obtained from the confirmation experiments is 712.87 W/m2K, which falls within the predicted 95% confidence interval.

Table 7: Analysis of variance (ANOVA) for heat transfer coefficient.

7. Comparison between Blade Geometry of Stirrer

The optimal parameter of heat transfer coefficient from the previous literature study in agitated vessel with flat blade impeller parallel to the axis of the shaft was determined and optimized as 258.23 W/m2 K based on the Taguchi method [4]. But in this study, the blade angle is changed making 45° angle to the axis of the shaft. As per comparison between these two types of blades, the flat-blade impeller making 45° angle to the axis of the shaft is performed as the best for getting higher heat transfer coefficient (712.87 W/m2 K) during the agitation using coal slurry.

8. Conclusions

The purpose of the investigation is to determine the heat transfer coefficient for stream augmented flows in agitated vessels with flat blade impeller making 45° angle to the axis of the shaft. The rate of heat transfer is influenced by a number of physical and geometrical factors such as vessel configuration, impeller type, and process fluid like coal slurry. This study has shown the application of the Taguchi method on the performance evaluation of heat transfer coefficient for agitated vessel using coal slurry in coal gasification power plant. The following conclusions may be drawn form the present investigation work. (i)Based on the ANOVA results, all control factors have significant effect on the quality characteristics statistically except stirrer location.(ii)D/d ratio has the most dominant effect on total variation and it is followed by stirrer speed , C-w ratio , and stirrer location .(iii)The optimal levels of the process parameters were found to be stirrer speed of 300 rpm, stirrer location of 55.0 mm, D/d ratio of 2, and C-w ratio of 60%. (iv)The optimized value of the heat transfer coefficient for a 95% interval has been predicated as 707.99 W/m2 K; that is, 695.52 W/m2 K < 707.99 W/m2 K < 720.46 W/m2 K.(v)From confirmation experiments, the mean value of the heat transfer corresponding to the optimum conditions was obtained as 712.87 W/m2 K (Table 8), which fell within the predicated value.(vi)The jacketed vessel was heated by waste steam which is exhaust from the steam turbine in the coal gasification power plant.(vii)The Taguchi method can successfully be applied to heat transfer investigation in agitated vessel using coal gasification plant to save energy, time, and material in experimentation.

Table 8: Results of confirmation experiments.


The authors are sincerely thankful to the authority of Department of Mechanical Engineering, Annamalai University, for extending their support in carrying out this project. The cooperation extended by all technical staffs of steam laboratory during experimentation is highly appreciated.,


  1. C. L. Rai, I. Devotta, and P. G. Rao, “Heat transfer to viscous Newtonian and non-Newtonian fluids using helical ribbon agitator,” Chemical Engineering Journal, vol. 79, no. 1, pp. 73–77, 2000. View at Publisher · View at Google Scholar · View at Scopus
  2. J. Karcz and F. Strȩk, “Heat transfer in jacketed agitated vessels equipped with non-standard baffles,” The Chemical Engineering Journal, vol. 58, no. 2, pp. 135–143, 1995. View at Google Scholar · View at Scopus
  3. A. M. Pinar, O. Uluer, and V. Kirmaci, “Optimization of counter flow Ranque-Hilsch vortex tube performance using Taguchi method,” International Journal of Refrigeration, vol. 32, no. 6, pp. 1487–1494, 2009. View at Publisher · View at Google Scholar · View at Scopus
  4. C. M. Raguraman, R. Ramkumar, L. Sivakumar, and A. Ragupathy, “An effect of blade geometry on heat transfer performance in stirred vessel-coal water slurry system using coal gasification,” International Journal of Engineering Science and Technology, vol. 2, no. 4, pp. 587–594, 2010. View at Google Scholar
  5. Y. C. Choi, X. Y. Li, T. J. Park, J. H. Kim, and J. G. Lee, “Numerical study on the coal gasification characteristics in an entrained flow coal gasifier,” Fuel, vol. 80, no. 15, pp. 2193–2201, 2001. View at Publisher · View at Google Scholar · View at Scopus
  6. Coal Water Slurry Technology—Tangshan Keynan Environmental Protection Technology and Equipment Co. Ltd., Japan.
  7. E. S. Mosa, A. M. Saleh, T. A. Taha, and A. M. El-Molla, “Effect of chemical additives on flow characteristics of coal slurries,” Physicochemical Problems of Mineral Processing, vol. 42, pp. 107–118, 2008. View at Google Scholar · View at Scopus
  8. T. J. Park, J. H. Kim, J. G. Lee, J. C. Hong, Y. K. Kim, and Y. C. Choi, “Experimental studies on the characteristics of Entrained flow coal gasifier,” Energy Conversion Research Department, Korea Institute of Energy Research, Jaejon, Korea.
  9. ISO/IEC Guide, Uncertainty of Measurement Part-3, Guide to the Expression of Uncertainty in Measurement, Joint committee for guides in Metrology (JCGM), 2008.
  10. C. G. Persson, Guide to the Expression of Uncertainty in Measurement (GUM) and Its Possible Use in Geo Data Quality Assessment, Q-KEN, Riga, Lativa, 2011.
  11. L. Jinkun, Y. Inyoung, Y. Sooseoki, and K. J. Su, “Uncertainty analysis and ANOVA for the measurement reliability estimation of altitude engine test,” Journal of Mechanical Science and Technology, vol. 21, no. 4, pp. 664–671, 2007. View at Publisher · View at Google Scholar · View at Scopus
  12. S. Kumanan, J. E. R. Dhas, and K. Gowthaman, “Determination of submerged arc welding process parameters using Taguchi method and regression analysis,” Indian Journal of Engineering and Materials Sciences, vol. 14, no. 3, pp. 177–183, 2007. View at Google Scholar · View at Scopus
  13. A. E. Lance, Quality by Design Taguchi Method and US Industry, Irwin Professional, Sidney, Australia; ASI Process, Houston, Tex, USA, 2nd edition, 1994.
  14. D. C. Montgomery, Design and Analysis of Experiments, John wiley, New York, NY, USA, 4th edition, 1997.
  15. W. H. Yang and Y. S. Tarng, “Design optimization of cutting parameters for turning operations based on the Taguchi method,” Journal of Materials Processing Technology, vol. 84, no. 1–3, pp. 122–129, 1998. View at Google Scholar · View at Scopus
  16. H. Oktem, T. Erzurumlu, and I. Uzman, “Application of Taguchi optimization technique in determining plastic injection molding process parameters for a thin-shell part,” Materials and Design, vol. 28, no. 4, pp. 1271–1278, 2007. View at Publisher · View at Google Scholar · View at Scopus
  17. V. Parashar, A. Rehman, J. L. Bhagoria et al., “Investigation and optimization of surface roughness for wire cut electro discharge machining of SS 304L using Taguchi dynamic experiments,” International Journal of Engineering Studies, vol. 1, pp. 257–267, 2009. View at Google Scholar
  18. P. J. Ross, Taguchi Techniques for Quality Engineering, McGraw-Hill, New York, NY, USA, 1989.