Study of the relationship between the siltation rate of Algerian dams and the runoff coefficient

HEC–HMS

Hydrologic Engineering Center–Hydrologic Modeling System

NC

number curve

SMA

soil moisture accounting

Nowadays, the protection and preservation of water resources have become an inescapable necessity in the face of the increasing different water needs (domestic, industrial, agricultural, etc.) (Olmstead 2010). Among the countries suffering from this problem is Algeria. It will be one of the poorest countries in terms of water potential. Today the water potential of Algeria is estimated at 20.5 billion m³, 12.3 billion m³ of surface water, and 2.7 billion m³ underground for the northern regions of the country. Then, 0.4 billion m³ of surface water and 5 billion m³/year are exploitable in the Saharan regions (Benchettouh et al. 2017). The surface water potential of Algeria is estimated at 10.9 billion m³, but the volumes currently mobilized through the 67 dams in operation hardly exceed 2.7 billion (Mohammed et al. 2014). One of the main concerns is siltation in reservoir dams (Serbah 2011). This natural phenomenon is only the consequence of the most characterized degradation of watersheds in the form of water erosion of the soil (Mohammed et al. 2014). On the other hand, the per capita water consumption in Algeria is considered below standard. From 1,500 m³ per inhabitant per year in 1962, the availability is no more than 500 m³ currently and will be 450 m³ in 2020 (Bouzid 2010).

The Algerian state has built almost 115 dams, among which 70 are large dams. With a capacity of 5,608 billion m³ and regularizing a volume of 2.8 billion m³/year on average, reservoirs and small dams ensure a volume of 13.7 million m³ and the big dams 227 million m³/year (ANBT 2012); despite all that, the dams are still threatened by the siltation phenomenon which reduces on average by 20–30 million/year.

The intensity of siltation is highlighted by the importance of the solid inputs of watercourses which vary from a few hundred thousand to a few million m³/year. The highest solid inputs are observed at the following dams: Beni-Amrane of 4.10 hm³/year; Grib of 3.2 hm³/year; Oued-Fodda: 2.66 hm³/year; Fergoug: 1.5 hm³/year; Djorf-Torba: 1.3 hm³/year; SMBA: 1.0 hm³/year (Touaibia 2019). Erosion is the natural process that can be accelerated by human activities disrupting the soil, caused by water involving tears from the soil mass, and this tearing of the soil is the result of the impact of rain and forces runoff shear (Hadidi et al. 2016). The rain that rips loose soil particles and the surface runoff that transports them to the bottom of the slopes provide the energy for this erosion and this transport. The movement toward the foot of the two slopes is in the form of a thin flow sheet or channeled flow (gully or gullies) (Hadidi et al. 2018).

In semi-arid regions, characterized by erratic and often large rainfall, climatic factors greatly influence the loss of soil. In this context, the Negev developed a model based on the fact that the amount of soil removed from the ‘splash’ effect is indicative of the high amount of rainfall (Poeppl et al. 2019). Williams also developed a solution of an immediate turbine unit for unpowered units, where the suspended load linearly varied with the volume being discharged away (Williams 1978). In the same field, Bergaoui et al. studied the use of hydrofluoric measurement data that have an unstable relationship with the basin of a layer in the Tebaga region in central Tunisia (Bergaoui et al. 1998). It was concluded that 84% of the solid linear transport model is explained by the maximum flow. This work focuses on the use of information about the intensity of rain and other factors leading to the maximum flow of floods to assess suspended solids.

Some of the factors affecting sediments are rain, the presence of slopes, the geological side of the Earth, and the nature of the climate. In Algeria, the geomorphological and hydrological factors are considered very favorable in accelerating the phenomenon of erosion and sediment transport, where the average amount of pond sediments that the Mediterranean tributaries spend each year is estimated at 120 million tons (Samir 2013). One of the consequences of this phenomenon is the deposition of the Algerian dams due to silt.

The different regions of Algeria are characterized by a highly irregular climate. This anomaly comes as sporadic showers, which are sometimes very important and can be responsible for abrasion and exceptional sediment transport. The spatial variation of erosion and sediment transport phenomena in North Africa is very high, either because of the severity of the phenomenon or its consistency. The main factor of these phenomena is water. Langbein & Schumm showed how erosion varies according to annual precipitation (Langbein & Schumm 1958). Corrosion is a growing function of annual high precipitation. The aim of this article is to study the relationship between the silt dam rate of Algerian dams and the runoff factor.

Finally, it should be recalled that this work aims to create a relationship that allows the assessment of major contributions to be made based on one or more hydrometeorological parameters. After that, the results of this study can be used by dam managers to take the necessary precautions and avoid problems in the water dams. The results of this work are also considered an important database for experts and study offices during the study of the construction of new dams in Algeria.

The remainder of this article is organized as follows. Section 2 presents the history of dams in Algeria. Section 3 explains the materials and methods. Section 4 finds the relationship between the runoff coefficient CR and the siltation rate. Finally, Section 5 concludes this article.

The first works date back to the 19th century when some small dams were built in the western region of the country. The first dams built in this region were built in the middle of the 20th century: Foum El Gueiss (Khenchela) in 1939, followed by K'sob (M'sila) in 1940, and Zardezas (Skikda) in 1945 (Salah-Eddine 2018). However, the construction techniques of the time not being perfected, some of them were quickly swept away by the Wadis, and others were rebuilt several times (Cheurfas, Tlelat, and Fergoug). At least three reasons can explain this:

• –

  Dam technology was not yet mastered.

• –

  The regime of watercourses was not well known.

• –

  The technical means implemented for the construction of these dams were often insufficient, which caused the work to drag on: the construction of the Djidiouia dam (0.7 hm³), for example, spanned 20 years (1857–1877), that of Hamiz (14 hm³) over 25 years (Table 1) (Salah-Eddine 2018).

The dams, therefore, began to be erected in Algeria from the 19th century, and this is the case of the first dam built in Meurad in 1859 (Tipaza). Due to its very low capacity (0.8 hm³), it is no longer used for irrigation. Then, the second dam (1860) made of the earth at Tlelat had a capacity of 0.7 hm³ and a height of 27 m, and it broke in 1862 after it was put into operation for sealing reasons. It was rebuilt in 1870 and reinforced in 1904. It is still in operation and used by the Zahana cement plant (Sidi Bel Abbès). There are projects characterized by a very long period to completion, for example, the Hamiz dam was a long-term project that required 40 years (Salah-Eddine 2018). It was designed in 1854, but work was only started in 1869 and ended in 1894. Other dams were built toward the end of the century (Cheurfas, Fergoug), but experienced design problems due to the fact that production techniques were not yet mastered. It was then that the construction of large structures was stopped in the 1890s. It did not resume until 1920, when the construction of large dams began, designed on the most favorable sites. Between 1932 and 1948, nine dams were built (Table 2). These were works of large and medium capacity (except that of Foum El Gueiss of small size): Oued-Fodda (1932), Boughzoul (1934), Bakhadda (1936), Ghrib and Foum El Gueiss (1939), K'sob (1940), Zardezas (1945), Beni Bahbel (1946), and Bou-Hanifia (1948). During the same period, two first-generation dams were raised: Hamiz and Cheurfas.

The volume that they regulated was barely 910 million m³ and concentrated mainly in the west of the country, and 15 dams with 93% of the total capacity of the identified were located west of the meridian of Algiers.

It was, therefore, necessary to wait for the second generation of the dam, comprising several hydraulic buildings, for eastern Algeria to benefit from a few works (Ksob, Zardezas, Foum El Gueiss, Foum El Gherza, and Cheffia, with the latter not being completed until after independence in 1965) of relatively modest size, except for the Cheffia whose capacity is rather large (168 hm³). The effort was sustained in the west of the country, which saw the construction of the dams of Sarno, Bakhadda, and Meffrouch. During this period, hydroelectric dams were also part of colonial policy. Two major dams in Little Kabylia were launched: Ighil Emda, which was built in 1954, and Erraguene, which was not completed until 1963.

The inherited situation prevailed until the early 1980s, and independent Algeria thereafter, in a certain way, renewed colonial policy, and during the first 20 years, from 1962 to 1980, only three new dams were built (La Cheffia in 1965, Djorf-Torba in 1969, and Sidi Mohamed Ben Aouda in 1970), but this increased the storage capacity from 910 to 1,660 hm³. However, during this period, the legacy of the colonial era continued to deteriorate. Indeed, siltation alone made these dams lose 2–3% of their useful capacity annually. To preserve this potential, it was, therefore, necessary to either elevate the structure (Zardezas in 1974, Ksob in 1977) or proceed to their silt removal, an excessively expensive operation and immobilize the dam during the works, which is not without prejudice to users, or finally rebuild them entirely (Fergoug and Cheurfas). It was during this same period that Morocco embarked on a far-reaching hydraulic policy. No fewer than 19 dams were put in water between 1961 and 1980, including 6 for 1978 alone. Several are very large: El Massira, Mohamed V, Oued el Makhazine. The theoretically regulatable capacity of Moroccan dams is 9.5 km³ which represents 59% of the mobilizable potential (16 km³).

This dramatic delay accumulated by Algeria in terms of hydraulics between 1962 and 1980 resulted in a regression of the irrigated areas, therefore a drop in agricultural production and a sharp deterioration in the satisfaction of urban demand for drinking water.

From the 1980s, and following a long period of drought, the country opted for a bolder policy, which manifested itself in an unprecedented dynamic in the hydraulic sector. In an attempt to catch up, the studies were systematized, substantial investments were made, and unprecedented mobilization of material and human resources was made. The result of this policy is a spectacular revival of dam projects by the restarting of construction sites that have been on standby for a long time. It followed the construction of 19 dams in just 10 years (1980–1990), and the prolific period is between 1985 and 1989, where no less than 15 dams entered service, or 3 dams per year, carrying the total at 37 structures and a storage volume equal to 3.9 billion m³. Eleven were in the west, 9 in the Cheliff, 7 in the center, and 10 in the east.

The decade 1990–2000 experienced a significant decline for two main reasons: (1) security, making work too risky for workers who are both foreigners and nationals, and (2) financial, i.e., the country was experiencing enormous difficulties due to capital injection. Hence, only seven dams were put into service.

The second ‘hydraulic’ awakening or the new awareness of the real danger that threatens the whole country did not take effect until after the 2002 crisis, which had awakened the spirits and increased the pace, thanks to the upturn. The financial situation resulting from the major surge in oil prices increased between 2000 and 2006 since no fewer than 13 dams were filled (Bouzid 2010). Thus, each year, one or two dams were put into service, and in 2009, 60 dams were in operation, 58 of which with a capacity of more than 10 million m³ each and an overall regularized volume of more than 7 billion m³. These are works that rarely reach 300 million m³. Algerian dams are therefore of medium capacity, and the largest of them have a capacity of 450 hm³ for the Gargar dam (Relizane), 640 hm³ for Koudiat Acerdoune (Bouira), and 795 hm³ for the Beni Haroun dam (Mila). Experts confirm that the natural conditions in Algeria do not allow for dams of greater capacity as is the case in Morocco (with five dams exceeding 1 billion m³ each) of Syria and its Tabqa dam (12 km³) or Egypt with the Aswan dam (162 km³) (Table 3).

This distribution highlights that the evolution of the number of dams is much more in favor of the east region that has known no fewer than 10 projects since 1980, followed by those in the west and the center with five structures and finally the Chlef region with only three projects (Bouzid 2010).

This change of situation, which was rather favorable in the West, is also noticeable through the capacity of the dams. If in the aftermath of independence, the theoretical capacity was much greater in the west, it is now in the domain of the east region, which is around 36%, while the west is content with 30% of the estimated capacity. This advantage can also be seen in the occupancy rate, which is, for an average of the 9 years, more than 62% for the east region and only 29% for the west region (K). Other dams are under construction or planned. However, these considerable efforts, even if they have already produced appreciable results, improving the satisfaction of needs, do not make it possible to make up for the dramatic backwardness that the country has accumulated because the increase in needs – notably agricultural and urban – is so dizzying that one can hardly be optimistic. This dilemma prompted the public authorities, helped by the financial upturn, to resort to new technologies such as transfers, regeneration of wastewater, the interconnection of dams, and above all desalination of seawater. Finally, we can identify at least two reasons to explain this revival or this reversal of hydraulic policy:

• –

  The too high level of food dependence and a critical level of water scarcity in Algerian cities.

• –

  Excessive stagnation in the agricultural sector is a limit to industrial development.

In the eastern region of northern Algeria, the two chains Atlas Tellien and Atlas Saharien (Abdelkader 2008), which border toward the sea and toward the Algerian part of the Atlas, behave very differently from each side of a line from Algiers to Biskra, the west; the two folded areas remain largely independent unlike in the East, where branches from the two chains put them in contact from the start, towards the Tunisian border, while those from the South approach near the shore so that they almost relay the coastal reliefs.

As a rule, the Northern Algeria Region belongs to the Mediterranean climate characterized by dry summers and mild winters with accentuated precipitation in the period from October to March. The predominant influence of the sea and the orientation mountain chains intervened in the formation of the climatic regions more or less parallel to the coast, as one goes from the sea toward the Sahara, the regions follow one another in the form of coastal strips, i.e., Tellian Atlas and the Tellian High Plains. Rainfall varies from 400 to 500 mm, further south, in the region of the high Stépiennes, where it is less than 300 mm. Minimum rainfall is observed in the Sahara region, less than 100 mm/year. Average annual temperatures are between 14.0 °C in Batna and 21.7 °C in Biskra (Hadidi et al. 2016).

In the coastal region, the hottest month is August, while in the sublittoral region (Atlas Tellien, Hautes plaines stepiennes, Atlas Saharien), it is the month of July. The annual temperature range in the sublittoral region, although it is important according to its value, varies imperceptibly on the territory from 20.4 °C in Batna to 42.9 °C in Touggourt (Hadidi et al. 2016).

The great values of the amplitude, higher than 20 °C, show that the varieties of the continental climate dominate in the sublittoral region and in the littoral region where the temperature ranges from 13 to 18 °C dominates the varieties of Mediterranean climate (Hadidi et al. 2016).

Since the climate has an important influence on the intensity of the erosion process, it is studied in more detail in volume 1 of this dossier from the parameters.

Soils of Algeria were classified into two main categories, zonal soils and azonal soils (Hadidi et al. 2016). Among the zonal soils, four types of soils have been identified: unsaturated soils, limestone soils, balanced floors, and wind accumulation soil. Then, the soils were regrouped based on alluvial limestones: limestone alluvial soils, dune soils, and marsh soils.

The vegetation of northern Algeria is hunting, while interpreting the data obtained by satellite remote sensing.

The principle of this technique is raising funds along the profile, which was previously identified. The choice of profiles depends in particular on the length of the restraint (Kassoul et al. 2005). For a better estimation of the trapped sediments, the profiles must be parallel and materialized on the ground by visible landmarks (canvas) from the boat (topographic apparatus) whatever the height of the water body. In the case of a meandering reservoir, the profiles must be blamed enough to quantify the volume between two sections.

This coefficient is strongly influenced by the ground cover and the slope as shown in Table 4 in which the few values of this coefficient are obtained from the standards. These values reflect the ability of soils to run off based solely on soil cover.

We note in particular the very high runoff rate given for roads and roofs. As we have seen, this is explained by the fact that these surfaces are practically impermeable.

• –

  Height of the rains: In case runoff arises from waterlogging, or the intensity of rain in 30 min is 25 mm/h (Wischmeier & Smith 1978). This value has been questioned by Europeans who have shown that runoff can appear for much lower threshold values (2–10 mm/h) in a Mediterranean climate. Daily precipitation can reach 100 mm. The instantaneous intensity of 5 mm/min thunderstorm rain is accompanied by the high intensity in the north of France (an intensity of 220 mm/h was noted near the Marne valley in September 1987).

• –

  Soil moisture before the downpour is the second explanation for the volume runoff: this parameter is expressed either by the deficit of saturation of the soil before the downpour (porosity not waterlogged) or by the time in hours that passed before the rain or Köhler's clue (Wischmeier & Smith 1978). Then, soak rain is generally higher for the soil than when the soil is wet. There is an interaction between the state of the soil and the previous humidity of the ground. Research has shown that simulating rain on dry loamy soil can cause rapid soil degradation only if the same rain falls on already moistened plots.

• –

  The surface of the basin and the state of the soil surface: The surface permeability and the water retention capacity of the soil favor infiltration and therefore prevent runoff. The infiltration flow depends on the surface condition and the porosity, cracking system, and the biological activity, and the roughness of the soil surface mainly influences the soaking rain, but this influence diminishes when the slope increases because the volume stored in puddles decreases on steep slopes.

• –

  The influence of the slope: The inclination of the slope decreases the volume runoff because a steep slope gives better internal drainage and a slower formation of dandruff is gradually destroyed by the runoff energy. The length factor of the slope also affects the volume runoff, but theoretically this volume in percentage remains constant along the slope. It appears in many cases, when the soils are bare, that the runoff coefficient decreases if the slope increases.

• –

  Fluid speed: This speed depends on the thickness of the dripping blade, the slope, and roughness.

• –

  Plant cover: Runoff is increased by the low vegetation cover.

Hydrologic Engineering Center–Hydrologic Modeling System software calculates the flow volumes by subtracting from the precipitation the quantities of water that are stored, infiltrated, or evaporated during their journey on the watershed. In addition, the areas of a watershed are classified into two categories:

• –

  Directly connected and waterproof surfaces where the flow is direct and is done without losses.

• –

  The permeable surfaces subjected to losses are described by the following different models:

• • ▪

    01 – initial loss model at constant rate;

  • ▪

    02 – constant deficit and loss rate model;

  • ▪

    03 – model based on the number curve;

  • ▪

    04 – Creen and Ampt model (Aouiche et al. 2023).

For all of these models, the losses are calculated for each time interval and subtracted from 1.2.3.4. The constant rate initial loss model is presented in the study by Kim & Shin (2018).

In these models, the assumptions are as follows:

• – The maximum potential loss rate, noted F_c, is constant.

• – There is an initial loss I_a which represents the interception and the storage in the depressions of the catchment area; as long as I_a is not reached there is no runoff;

• This operation can be summarized as follows:

• – We denote Pet: The surface precipitation average or time T and Pet runoff at time T.

  

  (7)

  

  (8)

  

  (9)

The difficulty of these methods is as follows:

• – Determination of losses from initials. They depend on the conditions of the plural event to be studied (e.g., if the soil was already saturated with water by previous rains, the initial losses will be almost zero). These losses also depend on the development of the soil type. It is estimated that these losses are equal to 10–20% of the rain for the forest, while in urban areas, it is between 2 and 5 mm of the water height.

• – The determination of the constant loss rate corresponds to the absorbing power of the soil expressed in mm/h. The values are presented in Table 5.

A variant of this model is the almost continuous model, which takes into account periods without rain during the event and therefore includes regeneration (with a rate to be fixed) of the initial losses. This is the ‘deficit’ model. In general, one will not directly determine the initial losses and the rate of loss, but one will proceed rather to a calibration of the model starting from real data.

Soil moisture accounting (SMA) is a model that can study long periods with alternating rain and dry weather (Kim & Shin 2018). SMA simulates the movement of water through the deflected elements of a catchment area. From precipitation and evapotranspiration data, surface runoff, infiltration, evaporation, and deep percolation are calculated. The watershed is represented by a series of storage layers:

• –

  Storage by vegetation interception: This layer represents the water retained by the vegetation (trees, grasses) and therefore does not reach the ground. Evaporation is the only way to empty this layer.

• –

  Storage by interception of the surface (i.e., in small soil depressions): This layer is filled when the maximum infiltration is reached. When full, it overflows to scream surface runoff.

• –

  Subsurface storage (soil profile storage): It represents the water retained in the soil at a shallow depth and is likely to be subjected to evapotranspiration.

• –

  Underground storage: This water is obtained by percolation (the rate of which is to be defined) and considered as lost to the system.

The runoff blade parameter plays a very important role in erosion and therefore the quantities deposited in reservoirs. Several authors introduce the trickle blade into the equations for evaluating the erosion rate, e.g., Fournier's relationship. On the basis of this hypothesis, we consider it interesting to introduce this parameter into our correlation.

On the other hand, for the two regions east and center, there is no correlation between the two parameters. The effect is least justified by the influence of vegetation and the nature of the soil coming from the latter estimate.

Preventive methods against the siltation of Algerian dams are as follows:

The first idea that comes to mind to reduce this phenomenon is to prevent the formation of sediments – products of erosion by effective treatment of the catchment areas. We can cite a few methods used in Algeria:

• –

  Restoration and protection of soils.

• –

  The formation of benches and torrential correction.

• –

  The creation of small gabion dams in the small talwegs.

• –

  The creation of a spreading beach.

• –

  Crop planning following contour lines.

• –

  The planting of long-stemmed vegetation in the wadis: it should be noted that the tamarisk trees that have grown upstream of the Bou-Hanifia, Frgoug, and Chenrfas dams provide real sediment traps.

For the General Directorate of Forests it will be a question of treating an area of 1.5 million hectares by the year 2017, that is to say, a rate of realization of 6,700 ha/year.

• •

  However, the impact of the preventive treatment of watersheds cannot be measured immediately.

• •

  Healing methods

• •

  Hydraulic flushes

The elimination part of the sediments includes the following:

• •

  The so-called Spanish flushes used during the first autumn floods (more loaded with solid matter) are valid especially for dams of lesser importance and with annual regulation (dam of Hamiz, Beni-Amrane), but the disadvantage lies in the release of large quantities of water, and this voluntary loss is hardly acceptable in Algeria.

• •

  Sludge is flushed through the bottom outlets when the dam is full and when excess water enters the reservoir. This method is used in most Algerian dams. Their effectiveness, however, is limited to the creation of a necessary cutout around the bottom outlet to prevent the valves from blocking.

• •

  Withdrawal by ‘density current’ or ‘underflow’ due to the silting valves of such devices exists at the level of the dams of Eghil Emda, Sidi Mohamed Ben Aouda, K'Sob, Zardezas, Deurdeur, Guenitra, Ain-Dalia, and Cheurfas. The interest in these gates is mainly due to the evacuation in the immediate vicinity of the dam of the rather fluid muds, which occurs during the first floods of autumn to avoid clogging the bottom draining organs.

The different reservoir desilting techniques in this area are as follows:

• • Hydraulic dredging (stationary suction dredge, self-propelled dredge)

• • Mechanical dredging (dredging with buckets, clamshell bucket, and earthmoving equipment). This technique requires first the resolution of the following problems:

• • – The extraction and transport of mud using a minimum volume of water

  • – Definition of rejection zones

  • – Exploitation of the content of the water reservoir at the same time as the desilting

In Algeria, hydraulic desilting is always more advantageous than mechanical desilting:

• • Execution times

• • The cost per m3 to be desilted

• • Nonbulky equipment

• • The uninterrupted exploitation of the water resource

The phenomenon of siltation affects the Maghreb countries, more particularly our country, Algeria, and has increased from the 2000s (precisely from 2018), and it has a very serious effect on the national economy and the water resources of our country. To combat this phenomenon:

• –

  We must first find the most effective method to really estimate the siltation rate before building a dam to avoid wasting money randomly and to protect the national economy.

• –

  Given the importance of the surface condition factor of cultivated plots on the risks of runoff, erosion, and solid transport, the role of farmers is paramount. This condition depends on the use of the soil, the cropping system, tillage, and the choice of tools.

The preventive and curative fight against these phenomena will, therefore, be with the farmers. However, as a last resort, when runoff cannot be completely avoided, the creation of reservoirs to protect the village remains necessary and complementary to the agronomic and hydraulic actions staged in the watershed.

During this work, we tried to estimate a typical relationship between the siltation rate and the runoff coefficient and find the correlation between the two at the level of each region: center, Chelif, east, and west (Table 6).

The relation gives a good correlation between the two parameters in the west, Chélif, and center regions with an exponential relation of type y (Equation 13), where a and b are constants relative to the region. On the other hand, for the two regions, east and center, there is no correlation between the two parameters. Regarding the other regions (east and center), there are no good results and this is due to the following reasons:

• –

  There are developed and undeveloped watersheds.

• –

  There are wooded basins and other degraded areas.

During our work, we faced many problems in finding the data and filling the existing gaps, and after this work, we will try to find the correlation between the siltation rate and other factors that can influence siltation to have a final empirical relation that estimates the rate of siltation by these factors. This work may help in building dams in Algeria.

I thank the National Dam Agency (ANBT) and more particularly the dam maintenance department for helping teachers and students in terms of information and data. I would also like to thank the entire staff of National Hydraulic Resources Agency (ANRAH) who helped me to do this modest job.

• •

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.