A SEDIMENT FLUSHING TECHNIQUE WITH A GRIT CHAMBER

Md. Noh, M.N.¹, Michiue, Masanori² and Hinokidani, Osamu³

¹Graduate Student, ²Professor and ³Associate Professor

Department of Civil Engineering, Tottori University,

Tottori-shi, Koyama-cho Minami 4-101, 680-8552 Japan

Tel: 81-857-31-5284 and Fax: 81-857-28-7899

Abstract: This study attempts to discover the sediment flushing performance through a sluice gate within the scope of one-dimensional sediment flushing model for a grit chamber under a steep slope condition. The experiment setup had been installed as to entertain a sediment flushing technique for a movable bed configuration. The idea is to obtain a water depth at the gate vicinity, which corresponds to a particular gate opening height so that the sediment-flushing performance can be monitored and measured. In this context, several discharge platforms had been chosen with the variation of water depth being imposed by alternating the gate opening height. Sedimentation process had been conducted by considering the critical sedimentation condition that was by allowing the maximum sedimentation to occur inside the reservoir. Upon achieving this, the flushing process would be commenced immediately. The results conclude that the x_max line was found to be slightly higher for the case of flexible bed compared to the case of fixed bed. 

Keyword: sluice gate, sediment flushing process and critical length of sub-critical zone x_max

1    INTRODUCTION

Hydropower generation has a big impact in modernizing our life nowadays. The electricity being supplied by this method enables cities, towns and villages to flourish and expand. This type of energy is clean without releasing any toxic pollution as compared to fossil base electricity supply. That is why in certain countries, the utilization of this method of energy production is more favorable. Beside hydropower, the storage water impounded by the reservoirs can also be used for irrigation, flood mitigation, water supply, navigation and recreation. Nevertheless, hydropower generation has it own shortfall also. In order to generate the turbine system, a lot of water is required; thus, at the same time conveys some sediment towards the downstream section that increases the rate of wear and tear of the turbine blade. To avoid this, normally a grit chamber or a sediment trap is constructed upstream of the turbine compartment so that sediments can be deposited along the trap zone or the low velocity zone. In relation to this, sediment flushing assignment will only be commenced on periodical basis once the amount of deposited materials has reached the maximum level.

2    OBJECTIVE

This study is aimed to justify the applicability of utilizing the critical length of sub-critical zone or x_max. Figure 1 shows the schematic diagram of the current sediment flushing technique. It had been observed under fixed bed condition that the length of sub-critical zone x did influence the degree of sediment flushing under steep slope condition¹⁾. The occurrence of hydraulic jump in between the two flow energy (sub-critical and supercritical flows) help the sediment movement by initiating them further downward through the mixing process. Provided that this length x is short then good sediment flushing can be obtained. If this length is long then poor flushing will occur. For such reason, the previous study had thoroughly investigated this length by estimating the longest length x for which good sediment flushing could occur. This length x had been called x_max. In the present study however, in-depth investigation on this length x_max has been taken by considering a flexible bed configuration.

Fig. 1    Flushing Phenomena

3    REVIEW OF PREVIOUS RESULTS

In order to test the applicability of this flushing technique it is important to understand first the concept of the critical length of sub-critical zone or x_max as mentioned in the previous study^{1), 3)}. The previous results suggested that if the length of sub-critical zone (x) is less than or equal to the critical length of sub-critical zone (x_max), then good sediment flushing could be obtained. Initially, this condition had been derived as equation 1.

                                   (1)

Where h_d: sequent depth of the hydraulic jump. However, with further investigation occupying identical experimental procedure i.e. rigid bed phenomena with different sizes of sands (0.1mm, 0.75mm and 2.0mm diameter), the variation of x_max could also be related as depicted in equation 2.

                                (2)

Fig. 2    Estimation of x_max

Where C_b: ratio between gate width and flume width and F_up: Froude number at supercritical zone. This x_max line defines between the good and the poor flushing criteria. The variation of this line under fixed bed study is being portrayed in Figure 2. Accordingly, any point that is located below or on this line can be defined as good flushing area while any point located above this line can be considered as poor flushing area. Figure 2 also indicates that the variation of sediment sizes is not a prime factor in determining the flushing performance. Instead, it had been observed that the time taken for sediment flushing was longer for larger sediment. Then, the length of sub-critical zone x can be calculated by using equation 3.

                           (3)

Where x: length of sub-critical zone, S: bed slope, h₁: water depth at the gate vicinity and h_up: water depth at supercritical zone. It had been shown before that the results between the experimental and the calculated values for this length were close to unity³⁾. In order to obtain good flushing performance, the next condition has to be governed as shown in equation 4.

                            (4)

Fig. 3    Plan view of experiment setup

The condition imposes by equation 4 will determine whether the prescribed water depth that corresponds to a particular length x is below or above the x_max line. Finally, in order to estimate the water depth at the gate vicinity, the following equations could be used

                      (5)

                              (6)

Fig. 4    Longitudinal view of experiment setup

Where C_a: discharge coefficient, a: gate opening height, g: gravitational acceleration, Q: water discharge and A_g: area of water body at the gate vicinity. Apparently, the values of C_a and h₁ have to be determined by the trial and error basis through simple iteration process.

4    EXPERIMENT SETUP AND PROCEDURE

Figure 3 and 4 show the plan and the longitudinal views of the experiment setup for this study. Both sedimentation and flushing processes had been performed with only one type of sediment i.e. 0.75mm average diameter. The amount of sand being used for this experiment had been fixed at 110 liters throughout. Two types of bed slopes had been employed that comprise of 1/15 and 1/25. For each case of bed slope, three different discharges had been set up which total up to 6 discharge platforms. All the gate materials had been made from acrylic. The experiment had been done by first determining the location of water depth on the x_max line (Figure 2). For each discharge platform, different locations would be set up that corresponds to the point below the line, near the line and above the line. This water depth could be achieved by adjusting the gate opening height up or down. All together 22 cases had been conducted and each was compared to each other in terms of the ability to flush out the sediments.

Table 1    Experiment results

The final sedimentation process was due when the sand had reached or moved near the gate position. Upon reaching this maximum sedimentation condition, then the gate would be opened for sediment flushing. For each case of water depth, the opening height of the sluice gate would be identical throughout the entire sediment flushing process. The opening height of the sluice gate had been chosen likewise as to correlate to the prescribed water depth within the gate vicinity as to relate it to the point below, near or above the x_max line.  Accordingly, the good and the poor sediment flushing performance could be observed as intended. For each interested time frame, the variation of flushing process had been measured by shutting down the sluice gate. On the average, at least 3 sets of bed profiles had been taken and recorded. Some examples for indicating the general bed profiles will be shown later so that the variations can be visualized easily.

5    FLUSHING PROCESS

Table 1 shows the hydraulic condition and the results obtained from all the cases. For the case of good sediment flushing process, the final shape of sedimentation or bed profile comprises of small amount of sand at the left and right side with no deposition at the mid-section i.e. case 1-1 with 0.2 liter remaining out of the initial value of 110 liters. This deposition profile occurred when the water depth at the gate vicinity was set to be below the critical flushing line. Figure 5 shows the top view for the typical final deposition shape. On the other hand, Figure 6 shows a typical example of poor flushing condition. The amount of sand left inside the reservoir for this case is higher with the mid-section still having some deposition remaining i.e. case 1-4 with 6.2 liters remaining. The final water depth had been discovered to be high and as the result had generated a low value of tractive force at the gate position which in turn unable to induce the sediments at the mid-section to flow out of the impounded section.

Fig. 5    Typical layout for good flushing

Case 1-1

Case 1-1 indicates the variation for good sediment flushing process and the profiles can be seen in Figure 7. The erosion process starts at the hydraulic jump location where the mixing process begins. As the time passes by, this hydraulic jump moves further downstream near to the gate location that is more or less related to the amount of sediment left inside the reservoir. The final shape shows that there is no deposition at the middle part of the reservoir after 80 minutes of flushing.

Case 1-4

Case 1-4 elaborates for the poor flushing process that can be referred from Figure 8. The bed profile is longer than the previous case after 20 minutes of flushing i.e. deposition shape begins at 200cm compared to 140cm for case 1-1. Obviously, the reason is due to the lower gate opening height being prescribed that not only induces the water depth at the gate vicinity to be higher but also imposes a smaller tractive force to initiate. The final sediment flushing shape indicates that some deposition still remains at the middle section even after 160 minutes of flushing.

Fig. 6    Typical layout for poor flushing

Fig. 7    Case 1-1

6    DISCUSSION

The results from all figures show that the performance of sediment flushing depends on the position of the water depth at the gate vicinity that can be pin pointed easily on the chart of the x_max line. This water depth position can be determined by applying those equations stated previously. Under this study, the x_max line has been found to be slightly higher than the previous results^{1), 3)}. This is due to the amount of sediment being used for such studies. The previous attempt only employs 50-milliliters sands for the sediment flushing process (fixed bed) while under this study it considers until 110 liters (flexible bed). In addition, in the present study the sedimentation process has been allowed to reach until the critical condition i.e. the deposition is allowed to reach the gate location. Such differences have made the results for the x_max line varies slightly. With this in mind, it is appropriate to shift the line to the new position for practical purposes. Figure 9 shows the new x_max line as has been observed concurrently. Hence, the new equation to specify this new condition is

                                (7)

Fig. 8    Case 1-4

Equation 7 varies from equation 2 by a decrease in the power factor from 2.25 to 2.0. This new factor has increased the position of x_max slightly higher due to the location of F_up as a denominator of the fraction. This new equation also indicates that good sediment flushing process can be performed even with a longer length of sub-critical zone associated with a higher water depth. Figure 9 also employs the results obtained by M. Michiue²⁾ et al. The sedimentation procedure had been conducted similar to the current study that was by allowing the critical sedimentation condition to occur before the opening operation of the sluice gate was undertaken. However, the sediment flushing procedure had been initiated slightly different that was by the draw down flushing technique. The extraction of their results indicates that the variation between the good and the poor flushing performance also follow the same trend.

Fig. 9    Adjustment of x_max

7    CONCLUDING REMARKS

Strictly speaking, the critical length of sub-critical zone  can be specified as the pivot point of sediment flushing process. This line differentiates between the good and the poor sediment flushing criteria. By employing such methodology, the procedure for good sediment flushing process becomes practical and convenient. For setting up the initial value, the water depth approximation at the gate vicinity can be done through trial and error analysis as mentioned earlier. The estimation of this water depth and also for the length of sub-critical zone should be undertaken by assuming the reservoir is flowing under free sedimentation condition i.e. no deposition.

The application of this sediment flushing technique is appropriate to be utilized for a grit chamber, sediment trap, sediment pond or settling basin especially those constructed at steep slope associated with the occurrence of hydraulic jump. For a reminder also, the entire water body that flows along this constriction ought to be allowed to pass only through the sluice gate. Thus, all the equations derived in this paper should only valid for such condition. Should there occur any flow condition that passes over the weir or any combination between the orifice and the weir flow, then the validity of those equations mentioned above will become trivial. 

Acknowledgements

The Authors would like to express his high appreciation towards Mr. T. Oda for helping preparing the experiment setup. Not to forget also to Mr. T. Tsuji and Mr. T. Hara for their diligence in setting up and performing the experiments, and Mr. T. Sunada for providing the previous data.

References

[1]    Md. Noh, M.N., Michiue, M. and Hinokidani, O: Experimental Study on Sediment Flushing through Sluice Gate for Small Reservoir. Proceeding of the 12^th Congress of APD-IAHR, Bangkok, Vol. 2, pp. 641-650, Nov. 12-16, 2000.

[2]    Michiue, M, Hinokidani, O. and Sunada, T.: One Dimensional Analysis of Sediment Flushing through Sluice Gate for A Grit Chamber, 51^st. JSCE, Chugoku Branch, pp. 245-246, 1999 (in Japanese).

[3]    Tsuji, T., Michiue, M., Hinokidani, O and Md. Noh, M.N.: Study on Sediment Flushing through Sluice Gate for A Grit Chamber, 52^nd. JSCE, Chugoku Branch, pp. 177-178, 2000 (in Japanese).