Influence of Small Head Hydropower on Upstream Bed Topography

Transcription

1 ICSE6-232 Influence of Small Head Hydropower on Upstream Bed Topography Ronald MÖWS; Katinka KOLL Leichtweiß-Institut für Wasserbau, Technische Universität Braunschweig Beethovenstraße 51a, Braunschweig Leichtweiß-Institut für Wasserbau, Technische Universität Braunschweig Beethovenstraße 51a, Braunschweig ABSTRACT. The interest in utilisation of small head hydropower increased within the last years. Thousands of small weirs without any hydropower use exist throughout Europe. Most of these weirs cannot be removed due to the former straightening of rivers. Hydraulic model tests were conducted at the Leichtweiß-Institute for Hydraulic Engineering and Water Resources at the Technische Universität Braunschweig (LWI) to investigate the influence of building a hydrostatic pressure machine (HPM), a kind of water wheel, as a run-of-river station into a weir on the bed topography. The influence on an existing weir with a silted backwater and on a new weir without siltation was investigated. Due to the experimental setup, most of the discharge passed the weir and only a percentage was flowing through the HPM. In the experiments using an existing weir, only a negligible change in the upstream bed topography up to a discharge of 1.2 times the critical discharge of the sediment (Q crit ) was noticed. At 1.5 times Q crit the sediment upstream of the HPM eroded in a certain area. Upstream of the HPM the flow velocity rapidly decreased and a horizontal eddy developed. The eddy caused a deflection of the transported sediment, which passed the scour and directs it over the weir. The sediment did not pass the HPM. Doubling the width of the HPM caused lower flow velocities, but simultaneously an increase of size and depth of the scour. Feeding sediment caused a decrease in scour size, a gravel bar developed further upstream and an erosion channel occurred in the middle of the flume. In the experiments using the new weir situation, sediment was fed into the backwater area and a sediment body developed growing towards the weir. In the vicinity of the weir the shape of the body changed and very deep scour depth developed directly upstream of the HPM. The scour was refilled by transported sediment: The bed topography development was comparable to the development of the sediment feeding experiment at the existing weir. Key words Sediment, Scouring, Hydropower, Small weirs. I INTRODUCTION Small dams and weirs can be found in nearly all European river systems. These existing structures cannot be removed. The backwater area of these structures is often silted up to a certain level, depending on e.g. the sediment budget of the river, the weir type and the age of the weir. In order to fulfil the requirements of sediment continuity, given by the Water Framework Directive as well as the target of the Renewable Energy Directive, water wheels may give a contribution. In the context of the European research project Hylow a hydrostatic pressure machine (HPM), a kind of water wheel, was developed. The HPM was designed for the use of low head difference up to 2.5 m [Müller, 2009]. The installation of a HPM as a run-of-river power plant into a weir may influence the morphodynamic regime, resulting in sediment mobilisation, transport and deposition. The initiation of scouring processes may endanger the stability of the constructions and bed load could be transported through the machine and cause damages to the wheel. The HPM design allows the passage of small grain sizes. In mountain river systems high bed load transport rates with coarse grain sizes occur during flood events and may endanger the construction. This paper reports on laboratory experiments conducted in a flume with movable bed to investigate the impact of a HPM installed in a weir on the bed topography and the flow field. Results of bed surface scans and velocity measurements are presented considering the effect of the experimental discharge related to the critical discharge of the bed material, the position of the HPM and the width of the HPM. 965

2 Additionally a new weir situation was investigated, focussing on the sedimentation of transported material in the backwater area and the resulting bed forms due to the installation of a HPM. II EXPERIMENTS II.1 Setup The experiments were carried out in tilting flumes at the LWI. Two different flumes were used for the tests. The experiments using an existing weir were conducted in a 30 m long, 2 m wide (reduced to 1 m) and 0.8 m high one (flume 1) and tests using a new weir in a 20 m long, 0.9 m wide and 0.6 m high one (flume 2). The slope of both flumes was adjusted to 0.5 %. In both flumes the discharge can be regulated with a valve at the inlet and measured with an IDM. The water level can be regulated with a weir at the outflow of the flumes. A CNC controlled traverse is installed on top of each flume. The traverse of flume 1 is equipped with 3 ultrasonic probes in order to scan the bed topography and at flume 2 a laser scanner is installed. At the flume inlet a flow straightener was installed. And further downstream a weir with a height of 7 cm and a width of 81 cm (71 cm at flume 2) was installed. The HPM construction with a clearance of 15 cm was placed in flow direction on the left side of the weir (see Figure 1). The dimensions of the constructions correspond to a 1:1 model which is built in the Iskar River in Bulgaria. The HPM structure consists of two side walls including two bearings and a 10 cm wide water wheel with a hub diameter of 5 cm and straight blades of 7.5 cm length. The gaps between bottom and wheel and between bearings and wheel were about 1 mm. Figure 1: View of the weir with the HPM from downstream. Before the experiments could start, sediment was filled parallel to the flume bed into the backwater area upstream of the weir up to a certain level and smoothed. For the experiments using the existing weir the sediment surface reached up to 1 cm below the weir crest, thus a 6 cm high free layer resulted in front of the HPM. Within the new weir test the surface was smoothed at the height of the bottom of the inlet structure to the HPM in 7 cm below the weir crest (Figure 2). Figure 2: Side view on the experimental setup for the tests using an existing weir and a new weir. The flow direction is from left to right. The sediment had to fulfil the following criteria. It had to be small enough to enable mobile bed conditions and it had to be coarse enough to be transported without any bed form development like dunes. Additionally, the sediment should be uniform to avoid the development of an armour layer. Fine gravel with grain sizes from 2 to 6 mm and a medium diameter d m = 3.6 mm was chosen. II.2 Experimental procedure After preparing the bed surface, uniform flow conditions were adjusted for a certain discharge and held constant for 24 hours. The bed surface was scanned with ultrasonic probes to determine changes of the bed 966

3 topography at the beginning and after the end of the experiments. The resolution of the ultrasonic scan was Δx = 2 mm in longitudinal direction and Δy = 80 mm in lateral direction. The accuracy of the resulting height was ± 2 mm. An area of 498 cm length and 88 cm width was scanned starting 2 cm upstream of the weir. The distance to the flume walls was 6 cm on both sides. Flow velocities were measured with a micro propeller at the end of each experiment in cross-sections 5 and 50 cm upstream of the weir. The lateral distance of the measuring points was 10 cm and the vertical position was 2 cm above the weir crest. Sediment was fed by a vibrating unit in 7 m upstream of the weir. In the new weir experiments the scanned bed surface area was limited by the length of the traverse. The scan resolution was Δx = 0.5 mm in longitudinal and Δy = 1 cm in lateral direction. The accuracy of the resulting height was ± 0.1 mm and the measurement range was 10 cm. The scan started 108 cm and ended 0.8 cm upstream of the weir, the distance to the flume walls was 6 cm. In case of bed surface depths deeper than the range of the laser scanner, measurements were complemented with a point gauge. Measurement of flow velocities and feeding sediment were conducted in these experiments as well as it was done in the existing weir experiments. The experimental discharges were chosen in relation to the critical discharge of the Sediment (Q crit ). Q crit was determined under uniform flow conditions with 33 l/s in flume 1 and 29 l/s in flume 2. The discharges used in the experiments were chosen with 0.5, 1, 1.2 and 1.5 Q crit. Additional experiments were carried out for the highest discharge (1.5 Q crit ). The tests contained sediment feeding (10 g/s), a doubled wheel width and a changed HPM position, using the standard as well as the wide HPM. The experimental parameters are summarized in Table 1. Table 1: Experimental parameter. No. Theme Q/Q crit [-] Comment 1 Existing weir 0.5 HPM position on the left side Sediment feeding: 10 g/s; HPM left HPM position in the middle of the weir Doubled HPM width; HPM left Doubled HPM width; position in the middle of the weir 9 New weir 1.5 Sediment feeding: 10 g/s; HPM left It has to be mentioned that the experimental setup closely follows the 1:1 model which was built in the Iskar River (see above). The Iskar River is a mountain river system with a wide range of occurring discharges. The design discharge of the HPM (Q d =5 m 3 /s) is less than the available discharge and especially the sediment transporting discharge which was calculated for uniform flow conditions with 49.1 m 3 /s [Koll & Möws, 2011]. Accordingly a percentage of the water always flows over the weir. In the laboratory experiments only a part of the discharge (6 to 7 l/s) flew through the HPM and most of the water passed the weir. II.3 Results II.3.1 Existing weir The area upstream of the HPM was nearly unaffected in the experiments using discharges up to 1.2 Q crit. The remaining weir height of only 1 cm above the bed surface was high enough to cause a backwater which reduced the shear stress to a value smaller than the critical shear stress of the bed material. Thus the sediment was eroded only very locally in the vicinity upstream of the HPM and no sediment was transported in the area further upstream. The reason for this erosion was the instability of the resulting free sediment layer upstream of the HPM inlet (see experimental setup). In the experiment (No. 4) using a discharge of 1.5 Q crit erosion occurred upstream of the HPM. Directly in front of the HPM (2 cm upstream) the resulting scour reached a maximum depth of 7 cm directly in front of the HPM (2 cm upstream), i.e. 1 cm below the top of the HPM bottom. The scour was narrowed by the flume wall on the left side and on the right side of the HPM the width of the scour was larger than the clearance of 967

4 the HPM construction. The experiment no. 4 was repeated twice to check sensitivity and reproducibility of the results. The results are very well reproducible with variation of the scour depth and width of less than 6 and 7 %, respectively. The scour length was 14 cm without variation. The experiment was repeated with sediment feeding (No. 5) and the generated scour was smaller and less deep. In experiment No. 4 the scour was 20 cm long and 7 cm deep and in experiment No.5 the scour size decreased to a length of 8 cm and a depth of 5 cm. In both experiments the width of the scour was almost the same (30 cm (No. 4) and 29 cm (No. 5)). Due to sediment feeding the transport rate was higher in experiment No. 5 and thus sediment which was eroded in front of the HPM was replaced by incoming material. The scour equilibrates at a certain size, depending on the amount of incoming and eroded sediment. However, in both experiments sediment was transported into the erosion zone, but remobilized and transported to the right side and over the weir. No sediment was transported through the HPM. Figure 3: Bed topography upstream of the weir with and without sediment feeding at 1.5 Q crit. The origin of the vertical coordinate is at weir crest. The higher transport rate in experiment No. 5 had an influence on the upstream bed topography as well. A shallow gravel bar with a height of 1 cm developed upstream of the scour and a narrow channel was eroded in the middle of the flume which widened towards the weir (Figure 3). A change of the position of the HPM from the left side to the middle of the weir (experiment No. 6 compared to No. 4) did not cause remarkable changes in size and depth of the resulting scours. However, the scour width was limited by the left flume wall in experiment No. 4 (and No. 7) but it developed symmetrically in experiment No. 6 (and 8). For a comparison of the experiments only the extent of the erosion zone on the right side of the HPM is given in Table 2, not the total width. Moreover the scour depth of the according experiments is summarized in Table 2. The lengths of the erosion zones were almost constant throughout the experiments. In experiment No. 4 the length of the erosion zone was extended to 20 cm by a smooth slope ratio (Figure 4) at the side of the flume wall. The experiments using the doubled wheel width (clearance of 25 cm) and the experiments where the HPM is located on the left side caused slightly deeper scours. The extent of the scours increased in experiment No. 7 and No. 8 due to the wider opening of the inlet structure (see Table 2). However, no sediment was transported through the HPM. Table 2: Parameters of the erosion zone, depending on width and position of the HPM construction. Experiment HPM width / position Depth [cm] Extend [cm] Length [cm] No. 4 small / left 7,0 16, small / middle 6, wide / left 7,8 28,5 16,1 8 wide / middle 7,4 20,5 17,4 968

5 Figure 4: Bed topography upstream of the weir depending on HPM width and position. The reason why no sediment was transported through the HPM can exemplarily be shown in the visualization of the stream lines for experiment No. 4 (Figure 5). The flow was decelerated in front of the wheel and directed towards the weir side. A horizontal eddy developed which was almost symmetrically when the HPM was positioned in the middle of the flume. Sediment which entered this area or was mobilized in front of the wheel was transported to the right side of the HPM (both sides in case of No. 6 and No. 8) and passed the weir. Additionally the straight blades of the HPM induced turbulences which pushed transported material backwards [Hecht et al., 2011]. This effect of pushing sediment away from the HPM in combination with the backwater effect and a continuous sediment transport into the scour upstream of the HPM leads to sediment guiding beside the HPM over the weir. As aforementioned the experimental setup is geared to the 1:1 Iskar model. At that site the design discharge of the HPM is less than the available discharge and even the sediment transporting discharge, thus a percentage of water is always flowing over the weir. Figure 5: Visualization of the stream lines upstream of the inlet to the HPM in experiment No. 4. The flow direction is from top to down. The flow deceleration caused by the water wheel is illustrated in Figure 6. Flow velocities measured in two cross-sections, 5 cm and 50 cm upstream of the construction, are shown for the four experiments summarized in Table 2. In the experiments with the HPM located at the left flume wall the flow velocities in 50 cm upstream of the construction are only slightly smaller in the left half of the flume than in the right half. In case of the HPM positioned in the middle of the flume its influence on the flow field is recognizable in a distance 50 cm upstream, but still small. However, close to the construction the velocities dramatically decreased in front of the wheel, depending on the width of the HPM. In the experiments with the smaller HPM the decrease of the flow velocity was much more distinct than in these with the wide HPM. The lateral influence on the flow was sharply limited in the experiments No. 4 and No. 6. An influence over the whole flume width is illustrated in experiments No. 7 and No. 8 (Figure 6). The limited influence on the flow velocities across the flume explains the smaller width of the scour (No. 4 & 6) compared to the experiments with the large HPM. 969

6 However, in general the flow velocities were lower in the experiments with the large HPM which was confirmed by measurements of the water level. In 5 cm upstream of the HPM the water levels reached 8.1 cm in the experiments No. 7 & 8, compared to 7.1 cm with the small HPM (No. 4) and 6.5 cm under uniform flow conditions. Accordingly the HPM has a blockage effect, resulting in higher water levels at higher ratios of the HPM width to weir width. Figure 6: Flow velocities measured in 5 and 50 cm upstream of the weir. The dashed lines show the side walls of the HPM construction. II.3.2 New weir At the beginning of experiment No. 9 the backwater effect of the weir was crucial. The shear stress was too low to transport the fed sediment and a sediment body developed at the feeding point up to a certain height. After the sediment body grew up to 6 cm above the sediment surface a dune like body developed which grew towards the weir. In the vicinity of the weir (about 1 m upstream of the weir) the sediment front grew faster on the weir side than on the side with the HPM and a scouring process started upstream of the inlet structure. The scour reached a maximum depth of about 12 cm (5 cm below the bottom of the inlet structure) and a maximum length of 22 cm during the ongoing experiment. Sediment was deposited around the border of the scour, but transported into the scour as well (Figure 7). It has to be noticed that the scour depth, shown in Figure 7 was running out of the range of the laser scanner. The absolute scour depth of 12 cm was measured with a point gauge. 970

7 (a) (b) (c) Figure 7: Bed topography upstream of the weir at certain stages of sedimentation in the backwater area and at the end of experiment No. 9. The origin of the vertical coordinate is at weir crest. When the front of the sediment body reached the scour the amount of incoming material was higher than the eroded material and the scour was filled up. The resulting scour had a width of 28 cm, a length of 10 cm and a depth of 5 cm, so the dimensions were comparable to these from experiment No. 5. Upstream of the scour raised a gravel bar with a height of 1 cm over the weir height and on the right side of the flume an erosion channel was developed. The resulting topography is comparable to the sediment feeding experiment (No. 5), but the topography is more distinct due to the influence of the HPM construction from the beginning of the experiment. Figure 8 shows the measurement of the cross section 2 cm upstream of the weir corresponding to Figure 7b. It is obvious that the depth of the scour could destabilize the construction in nature. Thus, scour protection is required, especially, if a HPM is constructed together with a new weir or if the backwater area is not silted. However, transport of sediment into the HPM was not observed. Figure 8: Bed topography of the cross section 2 cm upstream of the weir during the sedimentation (Figure 7b). The scattered line shows the borders of the inlet structure. The origin of the vertical coordinate is at weir crest. III CONCLUSIONS Hydraulic model tests were performed to investigate the impact of installing a hydrostatic pressure machine (HPM) as a run-of-river power plant on the upstream bed topography. Two different general cases of an existing weir and a new weir were investigated. Furthermore, the influence of the discharge, sediment feeding and the size and position of the HPM has been investigated. Laboratory experiments were carried out in tilting flumes at the Leichtweiß-Institute for Hydraulic Engineering and Water Resources using movable bed consisting of fine gravel. Changes in the bed topography were determined by scans after a constant discharge over 24 hours ( existing weir ) or over 12 hours from the point where the sediment front reached the weir structure ( new weir ). The flow velocity was 971

8 measured at the end of each experiment in two cross sections upstream of the weir. It has to be noticed, that most of the discharge was flowing over the weir and only a percentage was flowing through the HPM. In the experiments using the existing weir situation only negligible changes in bed topography up to a discharge of 1.2 Q crit (critical discharge of the bed material under uniform flow conditions) were observed. Due to the experimental setup there was just a small erosion zone which occurred upstream of the inlet to the HPM. At a discharge of 1.5 Q crit a scour developed upstream of the HPM. Sediment was continuously transported into the scour, but mobilized again and transported to the side and over the weir. The scour was wider than the HPM construction and extended, depending on the width of the HPM. Also the maximum scour depth increased with an increasing width of the HPM. However, the scour length was almost constant, except for the experiment with sediment feeding. A higher transport rate in the experiment with sediment feeding was resulting in a smaller scour length and depth. At a certain scour size the scour dimensions were stable and all incoming material was eroded and transported beside the HPM over the weir, too. The sediment feeding in the new weir experiments resulted in the development of a sediment body, growing towards the weir. In the vicinity of the weir the front grew non-uniform and scouring started upstream of the inlet to the HPM. During further sedimentation of the backwater area the scour depth increased up to 12 cm. This big scour depth may endanger the weir stability or the HPM construction in nature. When the sediment front reached the scour, it was filled up to a level comparable to the level of the existing weir experiments with sediment feeding. However, all sediment that was transported into the scour was not transported through the HPM, but redirected and transported over the weir. Flow visualization as well as velocity measurements illustrated that the flow was decelerated in front of the HPM. The development of a horizontal eddy was observed during the experiments. Sediment particles transported towards the HPM were pushed backwards by turbulences induced by the straight blades of the HPM and directed towards the side(s) of the HPM where they passed the weir. Correspondingly no particles passed the HPM during the experiments. IV ACKNOWLEGMENTS The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/ ) under grant agreement No V REFERENCES AND CITATIONS Koll, Ka., Möws, R (2011): Report on morphodynamics of the LSM site. EC services Deliverable Report 2a.1 of the EU project Hylow. Müller, G. (2009) - Recent developments in hydropower with very low head differences. 33rd IAHR Congress: Water Engineering for a sustainable Environment, Hecht, V., Schneider, S., Linton, N., Müller, G. (2011): Sediment and debris passage. Internal task report 2.7 at EU project Hylow. 972