Sediment Transport Modeling at the Oued Fodda Watershed Level Using HEC-RAS 1D Software

GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 SEDIMENT TRANSPORT MODELING AT THE OUED FODDA WATERSHED LEVEL USING HEC-RAS 1D SOFTWARE Souhila BENKACI , Boualem REMINI Blida 1 University, Faculty of Technology, Department of Water Science and Environment, 9000 Blida, Algeria E-mail: ssbenkaci7[email protected], [email protected] ABSTRACT The objective of the current work is to determine the amount of sediments transported upstream and at the level of Oued Fodda dam. The latter is considered one of the first large dams built in Algeria. It’s exposed to a serious siltation problem that reduces its capacity every year. The simulation was executed using the HEC-RAS software, for the period varied from January 01, 2016 to April 30, 2016. The modeled section consisted of about 9999 m length, subdivided into 141 river stations distant from each other by 70 m. The observation of the studied river section bedprofile was selected as a criterion for comparing the results of the model with the real values observed. A roughness coefficient of 0.031 was used. A quantitative estimate with a determination coefficient, R2, of 0.92 was used to support the validity. The mass and concentration of sediments increased significantly in the crosssections located at the dam upstream. A total cumulative mass was estimated at approximately 712699 tons, just upstream of the dyke, and a maximum concentration of 22.35 g/l was observed, particularly for three main sections selected upstream of the Fodda wadi. However, at the reservoir level, the concentration variability is observed during flood periods, i.e., only for the most important flows. Keywords: GIS; HEC-RAS; Oued Fodda dam; Sediment transport; Siltation. 1 INTRODUCTION Due to its importance, the transport of solid materials is a major problem in the Maghreb countries [1]. In Algerian rivers, this problem has always arisen in an acute way. Due to the lack of data, its assessment remains complex. The enormous quantities of sediments transported are at the origin of the progressive reduction of the reservoirs storage capacity. In fact, around 65 million cubic meters of silt is deposited annually in Algerian dams [2]. The various bathymetric surveys effected during the period (1986–2008) by the National Agency for Dams and Transfers (ANBT) on all 59 dams in exploitation revealed that the volume lost through silting was 898 mm³, i.e. 13.4% of the total reservoirs volume [3–4]. Numerous methods for estimating solid transport exist in the literature, including models that explore physical laws such the St. Venant equation for the liquid phase, and the transport equations for the solid phase [5]. Although they faithfully represent transport phenomena, these models require the introduction of a large number of parameters. Mathematical models have been developed using three approaches: empirical models, which relate the flow of sediment to the outfall, the various climatic and biophysical explanatory variables [5]. Its major disadvantage is the difficulty of its calibration for large basins [6]. Regressive models (flow-TSS) deduced from the ratio, between the observed flow, the values of the suspended solids concentration (Lefkir, 2009), and the conceptual models, which consider erosion dynamics. These models have the advantage of estimating flows for different time steps, but they require a relatively long calibration period, and peaks in TSS concentration which are generally underestimated [7]. Researches has intensified and expanded to establish new methods: 1D models, numerous 2D models and finally 3D models, which have been developed to simulate sediment transport processes [8]. In view of this, our work is 161 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 a part of a one-dimensional solid material transport modeling using the HEC-RAS software. This code was created by the U.S. Army Corps of Engineers and has been employed by several engineering firms and governmental organizations [9]. It calculates the sediment transport capacity associated with each cross-section as a control volume for all grain sizes [10]. The Wadi Fodda sub-watershed was selected for this study. This basin is exposed to a major problem of soil degradation, which accelerates the silting phenomenon at the level of the Wadi Fodda dam. The latter is considered one of the first large dams built in Algeria. It’s intended for the irrigation of the Middle Cheliff perimeter [11]. Its storage capacity has decreased considerably over the last 18 years. The siltation rate has been estimated at about 54.29% and the bathymetric condition of the reservoir has become very worrying [12]. 2 STUDY AREA The Upper and Middle Cheliff watershed is located in the west of the Fodda Wadi basin, which is covering a surface area of 1153.5 km2. The larger Cheliff watershed, which is in the northwestern Algeria, is mostly occupied by the latter and located in its northeastern portion (Fig. 1). Figure 1. Oued Fodda watershed The Oued Fodda sub-basin has a considerable relief, an elongated shape (Fig. 2), and a highest point can reach 1950,7 m [13]. With 282 temporary wadis and 81 permanent wadis totaling roughly 1053.45 km and 897.16 km, respectively, it features a complex hydrographic network. It is mostly drained over a 92.14 km length by the Oued Rouina [11]. Our study area is exposed to a serious problem of soil degradation. Gully erosion has been identified towards the upstream of the basin [14] (Fig. 3). This is at the origin of an important production of sediment, which accelerates the silting process at the level of the Wadi Fodda dam reservoir (Figs. 4 and 5). It collects an annual volume of silt of 3.2 million m3 [15]. An annual loss of its storage capacity has been estimated to be about 4,248 km3, i.e. a silting rate of 45.43% in 2015 [13]. 162 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 Figure 2. Hypsometric Oued Fodda watershed map Figure 5. Vase deposit upstream of the Oued Fodda dam (Photo: Benkaci, August 2017) Figure 3. Gully formation at the Oued Fodda watershed Figure 4. Oued Fodda dam reservoir view with an initial capacity of 228 Mm3 (Photo: Remini, 2008) 163 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 The physiographic characteristics of the Oued Fodda sub-watershed were calculated based on ArcGisTM software [16] and summarized in Table 1 below: Table 1. Physiographic characteristics of the Oued Fodda sub-basin Characteristics Value Observation Area (km2) 1153.5 - Compactness index Kc 1.74 the basin is elongated favoring slow flow of runoff Max. Altitude (m) 1950.7 Highest point of the basin Average Altitude (m) 709.40 - Median Altitude (m) 750 Represents 50% of the total surface area of the basin Minimal Altitude (m) 142 Watershed outfall Equivalent rectangle length (km) 93.144 - Equivalent rectangle width (km) 12.384 - Mean slope index Imoy (%) 1.72 - Global slope Index Ig 0.1 Strong relief Drainage density Ds (m) 328.16 - 3 MATERIALS AND METHODS At the Wadi Fodda dam reservoir, the HEC-RAS program was primarily used to model the transport of sediment. It enables for all grain sizes and a control volume calculation of the transport capacity associated with each crosssection (Fig.6). The Exner equation, also known as the mass conservation and the sediment continuity equation, provides the foundation for its basic idea [10]: (1−𝜆𝑝)𝐵 𝜕𝜂 𝜕𝑡 =−𝜕𝑄𝑠 𝜕𝑥 (1) where: 𝜆𝑝 porosity of active layer, B channel width [m], 𝜂 channel elevation [m], 𝑄𝑠 sediment load transported [m3/s], X distance [m], T time [s]. Figure 6. Control volume used in HEC-RAS sediment calculations [17] 164 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 3.1 Model input parameters The calculation of sediment transport upstream and at the Fodda Wadi Dam requires three input data files, as shown in Fig. 7. Figure 7. Methodological diagram of the input parameters in the HEC-RAS model The Manning’s coefficient is the most important parameter to be adjusted in the calibration of the HEC-RAS model. The use of a single Manning’s coefficient may not be sufficient and adequate to represent the true roughness of a river under different flow conditions [18]. Due to assumptions in the data or the model limitations, no model will give findings that are exactly in line with the real results. But it is necessary to achieve a reasonable correlation between observed values and model outcomes [10]. For our model, the adjustment of the Manning coefficient is performed following the hydraulic simulation of the unstable flow over a period from January 01, 2016 until April 30, 2016. The water surface elevations (WSEl) associated with the measured flow rates and the output flow rates estimated by the HEC RAS software are compared as part of the calibration procedure. The values of the observed (measured) liquid flows are based on the daily liquid flows obtained at the level of the Oued Fodda dam direction during the considered period. Several roughness values were tested: 0.030, 0.031, 0.032, 0.033, 0.034 and 0.035. The Manning value retained (for which the error is minimal) is equal to 0.031 [13]. 3.1.1 Geometric Data The geometry of the studied section was initially created in ArcGisTM using a form of digital geographic vector data “TIN” (Fig. 8) extracted from the SRTM Worldwide Elevation Data (3 arc second Resolution 90 m) of the Oued Fodda watershed (Fig. 9). These data are then exported using the HEC-GeoRas software and visualized through the “Geometric Data” window (Fig. 10) with the HEC-RAS software [13]. The modeled section was approximately 9999 m in length ، subdivided into 141 river stations of approximately 70 m per part. Section 69 is located downstream (immediately upstream of the Wadi Fodda dam) at a distance of about 50 m (Fig. 11). On the other hand, section 9945 is located upstream. It is therefore, considered as the first section of the upstream Oued Fodda. Model input parameters Geometric Data - Image Shuttle Radar Topography Mission (SRTM) Triangulated irregular network (TIN) Flow Data Daily liquid flow rates Mean slope Sediment Data Initial conditions and transport parameters Bed gradation Sediment boundary conditions 165 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 Figure 8. Triangulated irregular network (Tin) of the Oued Fodda watershed Figure 9. Shuttle Radar Topography Mission image, SRTM Worldwide Elevation Data (3 arc second Resolution) of the Oued Fodda watershed 166 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 Figure 10. Geometrical data window visualized through a Google Earth image Figure 11. Characteristics of the upstream cross-section (69) of the Oued Fodda Dam dike 167 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 3.1.2 Flow data and boundary condition The simulation of sediment transport begins with the creation of a quasi-unstable flow file. Two boundary conditions are considered (Fig. 12). • The first is selected at the first cross section of the Fodda wadi (section 9945). For this condition, the data of the daily liquid flows observed at the dam upstream have been introduced with a time step of 12 days considering the duration of the simulation, which runs from January 1, 2016 to April 30, 2016. • The second boundary condition is located at the downstream (wadi Fodda dam dike). It simply corresponds to the value of the average slope (0.0242) at the level of the cross-section considered (section 69). Figure 12. HEC-RAS window of the quasi-unstable flow boundary conditions The obtained flow hydrograph (Fig. 13) illustrates the occurrence of two major floods in March 2016. The first is observed on March 12, corresponding to a flow rate of about 27 m3/s, and the second is observed on March 20 with an estimated flow rate of 20.66 m3/s. Figure 13. Flow hydrograph 168 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 3.1.3 Sediments data The sediment data file is based on three main input quantities. These are illustrated in Fig. 14. Figure 14. Methodological diagram of the sediment data file 3.1.3.1 Bed gradation The "bed gradation" data file is based on the results of the Granulometric analysis (Table 2). Samples were collected from several segments along Wadi Fodda, just at the dam upstream. The granulometric analysis sample was effected at the National Laboratory of Habitat and Construction (LNHC) of Oued Samar (Algiers). The curve is represented in Fig. 15. Figure 15. Grading curve of the sample taken along the upstream section of the Oued Fodda dam Sediment Data Initial conditions and Transport parameters Transport Functions Bed evolution Functions Sedimentary Sedimentation rate Bed Gradation Granulometric Analysis Sediment boundary conditions Rating curve Solid Flow 175 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 Figure 24. Temporal concentration variability at the upstream sections of the Fodda Wadi Dam Figure 25. Temporal concentration variability at the Wadi Fodda Dam 176 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 The spatial variation profile of the sediment concentration can be displayed for duration of simulation. A very significant variation is observed at the upstream cross-sections of the Fodda wadi and also at the sections from the downstream. An example is illustrated (Fig. 26) considering the two floods: 23 January and 2 February 2016. The maximum value of the sediment concentration was estimated around 6736.267 mg/l for the January 23rd flood and 12278.532 mg/l for the February 2nd flood. Figure 26. Spatial sediment concentration variability (Example of: 23 January and 02 February 2016 floods) 5 CONCLUSION HEC RAS software was used to model solid transport at the Fodda Wadi dam reservoir. Once all of the parameters were set to within their useful ranges, the model worked as expected. The observation of the bed profile was chosen as a standard for contrasting the output of the model with the actual values that were observed. Also the R2, which is equal to 0,92, is the coefficient of determination that supported the validation. This demonstrates strong concordance between model predictions and actual data. We can therefore accept our model as suitable for simulating sedimentary transport in one dimension. The bathymetric levees at Algerian reservoirs are rarely performed. The employed method is especially useful for watersheds missing bathymetric data or without hydrometric stations upstream of dams. ACKNOWLEDGEMENT I would like to address a great thanks to the Algiers direction of dams and transfers (ANBT), to the chief of exploitation and all the staff of the Oued Fodda dam, who facilitated my access to the Oued Fodda dam and who put at my disposal all the necessary technical and operating data, and who accompanied me particularly, for the recovery of the samples of silt at the level of the dam and at the level of the wadis located upstream. REFERENCES [1] MEDDI, M. Contribution à l’étude du transport solide en Algérie du Nord [Contribution to the study of solid transport in northern Algeria]. Larhyss Journal. 2015, no. 24, pp. 315–336. ISSN 1112-3680. Available at: https://larhyss.net/ojs/index.php/larhyss/article/view/346 [2] REMINI, B. and D. BENSAFIA. (2016). Envasement des barrages dans les régions arides [Siltation of dams in arid regions: Algerian examples]. Larhyss Journal. 2016. no. 27, pp. 63–90. ISSN 1112-3680. Available at: https://larhyss.net/ojs/index.php/larhyss/article/view/450 177 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 [3] TOUAHIR, S., A. ASRI, B. REMINI & S. HAMOUDI. Prédiction de l’érosion hydrique dans le bassin versant de l’oued Zeddine et de l’envasement du barrage Ouled Mellouk (Nord-Ouest algérien) [Prediction of water erosion in wadi Zeddine watershed and the silting of the Ouled Mellouk Dam (North-West of Algeria)]. Géomorphologie: Relief, Processus, Environnement. 2018, vol. 24(2), pp. 167–182. ISSN 1957777X. DOI: 10.4000/geomorphologie.12083 [4] MINISTRY OF WATER RESOURCES REALIZATION. Siltation of dams. MEDA program of the European Union – Directorate of hydraulic studies and development. 2010, vol. 3. [5] LEFKIR, A. Modélisation du transport solide par les modèles neuroflous [Modeling of solid transport by neuroflous models]. Alger, 2009. Doctoral thesis. Ecole Nationale Polytechnique, Département d'Hydraulique. [6] RASSI, W. Le transport solide: estimation et mesure [Solid Transport: Estimation and Measurement]. Limoges: Office international de l’eau, 2004. [7] PICOUET, C., B. HINGRAY & J.C. OLIVRY. Empirical and conceptual modelling of the suspended sediment dynamics in a large tropical African river: the Upper Niger River basin. Journal of Hydrology. 2001, vol. 250(1–4), pp. 19–39. ISSN 0022-1694. DOI: 10.1016/S0022-1694(01)00407-3 [8] GHARBI, M. Étude des inondations et du transport de sédiments associé – application au bassin versant de la Medjerda [Study of floods and associated sediment transport – application to the Medjerda watershed]. Toulouse, 2016. Doctoral thesis. Université de Toulouse, Institut National Polytechnique de Toulouse. Available at: https://oatao.univ-toulouse.fr/17382/ [9] ÉCOLE DE TECHNOLOGIE SUPÉRIEURE. HEC-RAS River Analysis System Guide de laboratoire: Exemple pour la rivière du Loup [Example laboratory guide for the Rivière du Loup]. Version 3.1.3, CTN762 Ressources hydriques. Quebec: University of Quebec, École de technologie supérieure, Département de Génie de la construction. [10] BEEBO, Q.N. & R.A. BILAL. Simulating bathymetric changes in reservoirs due to sedimentation. Application to Sakuma dam, Japan. Lund, Sweden: Lund University, Department of Building and Environmental Technology, Division of Water Resources Engineering, 2012. Available at: https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=3051998&fileOId=3052010 [11] AGENCE DU BASSIN HYDROGRAPHIQUE CHELIFF-ZAHREZ. Cadastre hydraulique du bassin hydrographie du Cheliff – Aval du barrage de Boughzoul [Hydraulic cadastre of the Cheliff hydrographic basin – Downstream of the Boughzoul dam. Chlef: ABH CZ, 2004. [12] Rapport technique du barrage oued Fodda, Wilaya De Chlef. 2004. [13] BENKACI, S. Impact du transport solide sur l’évolution des dépôts boueux dans les barrages [Impact of solid transport on the evolution of muddy deposits in dams]. Blida, 2020. Doctoral thesis. Université Saad Dahlab de Blida, Faculté de Technologie, Département de Science de l’Eau et Environnement. [14] BENKACI, S., D. ABIR, A. OUMELLAL & B. REMINI. Modélisation de l’érosion du bassin haut et moyen Cheliff par l’application Model builder sur ArcGis [Modeling of the erosion of the upper and middle Cheliff basin by the Model builder application on ArcGis]. Journal of Materials and Engineering Structures. 2018, vol.5(1), pp. 81–93. ISSN 2170-127X. Available at: https://revue.ummto.dz/index.php/JMES/article/view/1676 [15] REMINI, B. & W. HALLOUCHE. Evolution de l’envasement du barrage d’Oued El Fodda (Algérie) [Evolution of the silting up of the Oued El Fodda dam (Algeria)]. Wasser Energie Luft / Eau énergie air / Acqua energia aria. 2007, vol. 99(1), pp. 75–78. ISSN 0377-905X. Available at: https://www.eperiodica.ch/cntmng?pid=wel-004%3A2007%3A99%3A%3A414 [16] ESRI. ArcGis 10.2.1 for desktop [software]. January 29, 2019. Available at: https://support.esri.com/enus/patches-updates/2019/arcgis-10-2-1-for-desktop-engine-server-utilities-and-t-7680 [17] BRUNNER, G.W. HEC-RAS River Analysis System User’s Manual. Version 4.1, CPD-68. Davis, CA: US Army Corps of Engineers, Hydrological Engineering Center (HEC), 2010. Available at: https://www.hec.usace.army.mil/software/hec-ras/documentation/HEC-RAS_4.1_Users_Manual.pdf [18] ENVIRONMENTAL CONSULTING & TECHNOLOGY. Appendix I: HEC-RAS Modeling of Rainbow River. MFL Technical Support – Freshwater Stream. Final Report. Tampa, FL: Environmental Consulting & Technology, Inc., 2017. Available at: https://www.swfwmd.state.fl.us/sites/default/files/documents-andreports/appendix/Appendix_I.pdf [19] BRUNNER, G.W. HEC-RAS, River Analysis System Hydraulic Reference Manual. Version 4.1, CPD-69. Davis, CA: US Army Corps of Engineers, Hydrologic Engineering Center (HEC), 2010. Available at: https://www.hec.usace.army.mil/software/hec-ras/documentation/HEC-RAS_4.1_Reference_Manual.pdf [20] BRUNNER, G.W. HEC-RAS, River Analysis System Hydraulic Reference Manual. Version 5.0, CPD-69. Davis, CA: US Army Corps of Engineers, Hydrologic Engineering Center (HEC), 2016. Available at: 178 GeoScience Engineering Vol. 69 (2023), No. 2 geoscience.cz pp. 160–178, ISSN 1802-5420 DOI 10.35180/gse-2023-0097 https://www.hec.usace.army.mil/software/hec-ras/documentation/HECRAS%205.0%20Reference%20Manual.pdf [21] JIMENEZ, J.A. & O.S. MADSEN. A Simple Formula to Estimate Settling Velocity of Natural Sediments. Journal of Waterway, Port, Coastal, and Ocean Engineering. 2003, vol. 129(2), pp. 70–78. ISSN 0733-950X. DOI: 10.1061/(ASCE)0733-950X(2003)129:2(70)