Suitability Of Sediment Flushing In Mosul Dam

M_f

mass of sediment flushed annually from the reservoir.

M_dep

mass of sediment which deposits annually in the reservoir

W_res

representative reservoir width in the reach upstream, from the dam at the flushing water surface elevation (m)

W_bot

a representative bottom width for the reservoir (m)

SS_res

a representative side slope for the reservoir

El_f

water surface elevation at the dam during flushing (m)

EL_min

the minimum bed elevation which is usually the riverbed elevation immediately upstream from the dam (m)

W_f

the width of flow at the bed of the flushing channel (m)

W_tf

the top width of the scoured valley (m)

Q_f

the flushing discharge from outlet (m³/sec)

L

the longitudinal length of the reservoir (m)

Q_s

the sediment transporting capacity passed through outlet (tones/sec)

S

the longitudinal energy gradient through reservoir

W

the channel width of the reservoir (m)

T_f

the duration of flushing (days)

M_in

the mean annual sediment inflow to reservoir

SSs

representative side slope for the deposits exposed by flushing

Wt

the reservoir top width

W_td

the value for scoured valley width at top water level if complete draw down is assumed in (m)

W_bf

the bottom width of the scoured valley at full draw down

Approximately 1% of the storage volume of the world reservoirs is lost annually due to sediment deposition (Yoon, 1992). The rate of loss of storage varies from region to region and from reservoir to reservoir, the highest rates are found in the smallest reservoirs and the lowest rates in the largest. The estimate of annual loss of storage due to sedimentation have been used in conjunction with the gross storage volume data available in the ICOLD (International Commission of Large Dams) to estimate the magnitude of sedimentation problem as shown in Figure 1 (White et al., 2000).

In most developing countries where population pressures on fragile upland, ecosystems have led to accelerated rates of soil erosion, then reservoir storage is being lost at much larger rates. While there are some options for reducing the rates at which sediment deposits in reservoirs, flushing offers the only means of recovering lost storage without incurring the expenditure of dredging or other mechanical means of removing sediment. Most sites for reservoirs have already been utilized. Interest is increasing nowadays in reducing the rate at which storage capacities are being lost. Comparing with many cost alternative methods, flushing processes can offer an attractive means of recovering and maintaining a useful storage capacity.

Flushing is the scouring out of deposited sediment from reservoirs using low-level outlets in a dam to lower the water levels, and so increase the flow velocities in the reservoir. Flushing has been proved to be highly effective at some sites in the world. For example, at the Mangahao reservoir in New Zealand 59% of the original operating storage capacity had been lost by 1958, 34 years after the reservoir was first impounded. The reservoir was flushed in 1969 when 75% of the accumulated sediment was removed using flushing process in a month of operation (Jowett, 1984).

Drawdown is the lowering of the water levels in a reservoir. Drawing down a reservoir for a few weeks or months during the flood season is also a form of flushing although the principal purpose of drawdown is to pass the high sediment loads carried out by flood flows through the reservoir. In the literature this practice is commonly termed “sluicing”. A few attempts at the sediment flushing have been proved successful (Atkinson, 1996).

There are two key requirements for effective flushing: firstly, the quantity of sediment removed during flushing should at least match the quantity of sediment that deposits in the reservoir during the periods between flushing operations, and secondly the useful storage capacity that can be maintained should be a substantial proportion (above about 50%) of the original capacity (Atkinson, 1996). The literature survey on reservoir sedimentation by Sloff (1991) discusses such criteria but they cannot be used to assess the feasibility of flushing. A tentative criteria that the ratio of reservoir capacity to mean annual runoff (C/I) must be as large as a half or more for reservoirs with a half-life shorter than 100 year is reported by Ackers and Thompson (1987). Paul and Dhillon (1988) provide plots to determine the area of low level sluice required for flushing from the initial capacity and annual sediment inflow. Pitt and Thompson (1984) report that effective flushing has generally only been observed where the draw down level is below about half the height of the dam and where the sluice capacity exceeds the mean annual flow by at least a factor of 2. Mahmood (1987) presents a number of criteria for quantifying the efficiency and effectiveness of flushing but these can only be applied after a reservoir has been flushed and thus cannot be used to predict flushing performance. Finally, to minimize turbidity impacts on fish, irrigation and tourism, winter was advised by many researchers to perform flushing from reservoirs and to eliminate the negative downstream effects of flushing (Morris and Fan, 1997).

When flushing is attempted without drawing down water levels, the high flow velocities at the outlets are very localized and the impact is insignificant. The water level in a reservoir must be drawn down to close to the bed elevation at the dam before flushing can be effective (Figure 2(a)). Moderate lowering of water levels during flushing will still significantly increase flow velocities at the upstream end of the reservoir, where bed levels will be above the water level at the dam (Figure 2(b)). Large sediment volumes will be scoured from these upstream reaches and will redeposit near the dam (Atkinson, 1996). Eventually bed levels upstream from the dam will rise to the water level during flushing and then significant sediment quantities will be transported through the low-level outlets (Figure 2(c)).

The aim of the current study is to assess the feasibility of sediment flushing from Mosul reservoir, Northern Iraq, using simple worldwide criteria and indices which require readily available data. Applying these criteria, reservoirs where flushing might be applicable can be recorded. Some of the used criteria in this study were derived and verified by Atkinson (1996) from the data existed in the successful and unsuccessful worldwide reservoirs, while other criteria were predicted and suggested by the author. Use of the assessment of such criteria will help engineers to identify reservoirs where flushing has potential. When sites suitable for flushing are identified at the design stage, the construction of low-level outlets in dam’s body with sufficient capacity for flushing is recommended. Flexibility for flushing would then be added into the project design.

A sediment balance ratio SBR is defined as the sediment mass flushed annually divided by the sediment mass deposited annually. If SBR > 1 then it is expected that a sediment balance can be achieved and so this criterion is satisfied:

The calculation of SBR is performed as follows (Atkinson, 1996):

• Derive a representative reservoir width in the reach upstream (Figure 3) from thedam at the flushing water surface elevation.

  $Wres=Wbot+2SSres(ELf−ELmin)$

  (2)

• Take the minimum of Wres and Wf as the representative width of flow for flushing conditions, W.

• Estimate the longitudinal slope during flushing,

• $S=(ELmax−ELf)/L$

  (3)

• Determine the parameter (ψ) in the Tsinghua university method for sediment load prediction:

  • (ψ) = 180 for D₅₀ >0.5 mm (ψ) = 300 for D₅₀ > 0.1 mm

  • (ψ) = 650 for D₅₀ < 0.1 mm (ψ) = 1600 for fine loess sediments.

• $Calculate the sediment load during flushing Qs=ψ Q1.6S1.2/W0.6$

  (4)
• $Determine the sediment mass flushed annually Mf=86400TfQs$

  (5)

  (86400 is the number of seconds in a day).

• Predict Trap Efficiency (TE) using Brunes curves (Brune, 1953)(Figure 4)

• Calculate the mass depositing annually which must be flushed.

  $Mdep=(MinTE)/100$

  (6)
• $Determine SBR,SBR=Mf/Mdep$

  (7)

The long-term capacity ratio LTCR is defined as the sustainable capacity to the original capacity in which the sustainable capacity is the total reservoir volume which can be calculated from the final cross sections after flushing process.

The calculation of LTCR is performed as follows (Atkinson, 1996).

• Determine scoured valley width at the top water level.

  $Wtf=W+2SSs(Elmax+Elf)$

  (8)

  W_tf is the top width of the scoured valley (m)

• Determine the reservoir top width at the elevation for the simplified geometry assumed in Figure 3.

  $Wt=Wbot+2SSres(Elmax−Elmin)$

  (9)
• If Wtf < Wt then the reservoir geometry does not constrict the width of the scoured valley and so scoured valley cross sectional area (A_f) is calculated in (m²) as:

  $Af=(Wt+W/2)(Elmax−Elf)$

  (10)
• If W_tf > W_t then the scoured valley will have constricted end:

  $hm=Wres–(W)/2(SSs–SSres)h1=Elmax–Elf–hm$

  (11)

  $hf=Elmax–Elf$

  (12)

  $Af=W∗hf+(hf−h1)hmSSs+h12SSres$

  (13)
• Estimate the reservoir cross sectional area (A_r) in (m²):

  $Ar=(Wt+Wbot/2)(Elmax–Elmin)$

  (14)
• Determine LTCR,

  $LTCR=Af/Ar$

  (15)

Wt is the reservoir top width calculated in (6.2) step (2) in (m). Wtd and hence TWR is calculated as follows:

• Determine W_tf (the minimum of W_bot and W_t which are calculated in Section 6.2).

• Calculate Wtd from the side slope SSs which is:

  $Wtd=Wtf+2SSs(Elmax–Elmin)$

  (19)
• $Determine TWR=Wtd/Wt$

  (20)

Mosul dam project is located on Tigris River in the northern part of the Republic of Iraq in the Governorate of Ninawah approximately 60 km north of Mosul city (Figure 5). The Mosul dam is a multipurpose project, its object being to provide storage of water for irrigation, hydropower generation and flood control, peak control and the power regulation capacity of the project was increased by the inclusion of pumped storage plant. The Mosul dam project is subdivided into the following schemes: The main scheme comprises of the following main structures: The dam formed by 3600 m long earthfill dam with a maximum height of 100 m. The embankment volume is 3.5 million m^3. The reservoir created by the dam will have a usable storage volume of 8.16*10⁹ m³ at (330.m.a.s.l) available for irrigation and power production. The total volume of the reservoir at the retention water level (330.m.a.s.l) is 11.1 *10⁹ m³ with a surface area of 385 km² and reservoir length of about 65 km. The diversion of the river was made by means of 2 diversion tunnels which converted into bottom outlets 10 m Dia. 650 m long, each having a capacity of 1300 m³/sec at the minimum operating reservoir level of (300.m.a.s.l.) (Ministry of Irrigation, 1985). The following parameters for Mosul dam and reservoir are used as input for the application of the feasibility criteria of Mosul reservoir (Table 1).

SBR = M_f/M_dep

M_dep = (M_in TE)/100 = 40* 10⁶ *95/100 = 38 *10⁶

M_f = 86400 * T_f* Q_s

Q_s = 650 * 2500^1.6 * 0.0011^1.2/(650)^0.6 = 1027 ton/sec

Q_s will be divided by 3 because the rates of erosion in Mosul reservoir is about the third as in China according to the universal aerial iso-erosion maps (Morris and Fan, 1997).

Q_s = 342 ton/sec

M_f = 118195200 ton/year SBR = M_f/M_{dep =} 1199232000/38*10⁶ = 31.16

According to (Atkinson, 1996)

For successful flushing SBR must be more than 7.

H_f = 330 -290 – 40

h₁ = 330 -290 –h_m

h_{m =} W_res – W_f/2(SS_s- SS_res) W_res = W_bot + 2 SS_res (El_f –El_min)

W_res = 2000 +2* 20(290 -245) = 3800

h_m = (3200 -650)/2(5-20) = -105

h₁ = 40 – (-85) = 125

A_f = 650 *40+ (40+h1) * hm* SSs+ h²*SS_res = 186250 m²

A_r = W+ (W_bot/2) (El_max –El_min) W_t = W_bot +2 SS_res (El_max –El_min)

Wt = 2000 + 2*20(330 -245) = 5400

Ar = 5400 + (2000/2) (330 -245) = 90400 m²

LTCR = Af/Ar = 186250/90400 = 2 > 0.8 (where 0.7 is the minimum ratio for the LTCR), (Atkinson, 1996) O.k.

DDR = 1 – (El_f –El_min)/(El_max –El_min) =

1 – (290 -245)/(330 -245) = 0.47 < 0.8

FWR = Wf/Wbot = 650/2000 = 0.32 < 1

For successful flushing SBR > 7, LTCR > 0.7, DDR< 0.8, FWR <1 (Atkinson, 1996).

1. Storage capacity of reservoir

   If capacity inflow ratio < 0.5 (successful flushing) For Mosul reservoir C = 8* 10⁹ m³, I = 15.5 * 10⁹ m³ C/I = 0.6

2. Sediment potential

   sediment potential > (0.5 -1)% original capacity 40 *10⁶ m³/8 *10⁹ = 0.005 * 100% = 0.5%

3. Shape of reservoir basin

   Effective flushing will be for narrow steep sided reservoir valleys with steep longitudinal slope and W/L should be less than 0.1.

   For Mosul reservoir the width length ratio W/L = 4/65 = 0.062

4. Hydraulic condition required for efficient flushing.

5. Flushing discharges must be at least twice of the mean annual flow (Q_f > 2Q_inflow)

   For Mosul reservoir Q_f/Q_inflow = 2500/500 = 5 O.K

6. Flushing volume must be at least 10% of the mean annual runoff:

   For Mosul reservoir flushing volume = 2500*24*3600*40 = 8.3 * 10⁹ m³ Flushing

   volume/Mean annual runoff = 8.3*10⁹/15.5*10⁹ = 55.7%

According to the capacity inflow ratio value, Mosul reservoir is classified as seasonal storage reservoir. The trap efficiency of Mosul reservoir is estimated according to Brune curve (Brune, 1953). The retention period (residence time) of Mosul reservoir is calculated as the capacity of reservoir divided by daily inflow, while the sedimentation index of the reservoir is calculated as the retention time divided by the mean transit velocity of the flow in the reservoir. Finally, the specific storage of reservoir is calculated as the capacity of reservoir divided by the river basin area controlled by the reservoir (Unesco, 1985). These predicted reservoirs parameters shown in Table 2 were used by the author as a supplementary criterion in addition to those mentioned by Atkinson (1996) for predicting the feasibility of sediment flushing from Mosul reservoir.

Reservoir where there is regular annual cycle of flows and defined flood season can be considered as a suitable hydrological condition for sediment flushing to substitute and restore the released water amount during flushing. This hydrologic condition may be represented by the existence of annual flood season due to snowmelt during spring and summer months such as the present case study (Mosul dam) in which snow melt flow comes from the upstream highland of Tigris River catchment area.

To predict the flushing feasibility of Mosul reservoir, the sediment balance ratio SBR and the long-term capacity ratio LTCR were assessed. Some factors which determine the values of the sediment balance ratio, and the long-term capacity ratio are inherent characteristics of the site which include: the shape and size of the reservoir and the imposed hydrological conditions and the imposed sediment input. Some factors are controllable. These include the operation of the reservoir between flushing operations, the design of the flushing system including elevation, capacity and the operation of the flushing system including discharges and duration.

Flushing criteria obtained from the attempts of flushing sediment from a number of reservoirs (Atkinson, 1996) were used to test the feasibility of sediment removal from Mosul reservoir. Table 3 shows the results of the application of those criteria. The calculation of these criteria shows that the SBR is more than 7 which agree with successful flushing. The LTCR gave a value higher than 0.8 which also agree with the successful flushing. While the draw down ratio DDr is less than 0.8. and the FWR gave a value less than 1.0. Moreover, Mosul reservoir is hydraulically large in which the capacity of this reservoir about 60% of the mean annual inflow. The sediment deposition potential of Mosul reservoir is about 0.5% which is close to the criteria of successful flushing for sediment potential of reservoir which is (0.5-1)%. The hydraulic conditions required for efficient flushing agree with the available condition in Mosul reservoir in which the flushing discharges must be at least twice the mean annual flow in which: Q_f/Q_in = 2500/500 = 5.

Another hydraulic condition for Mosul reservoir which agree with the criteria for successful flushing is the flushing volume which must be at least 10% of mean annual runoff. For Mosul reservoir flushing volume is 55.7% of the annual runoff.

The shape of Mosul reservoir agrees with the shape needed for successful flushing which is a narrow steep sided reservoir valley. Mosul reservoir width length ratio is 0.062 which agree with the required reservoir width depth <1.

Finally, a hydrological condition needed for successful sediment flushing is represented in reservoir when there is a regular annual cycle of flows and defined flood season. This may be represented by existing annual cycle of snow melt during spring and summer months which exists in the upstream watershed area of Mosul reservoir.

From the application of the worldwide universal criteria on sediment flushing in reservoir predicted by many scientists such as^[4], these criteria were applied on Mosul reservoir in north Iraq. From the analysis of the available field data concerning flushing sediment in many reservoirs in the world, the author in general concluded that the efficiency of sediment flushing can be increased with a smaller depth of stored water, the greater discharge of the flushing stream, the greater the dimensions of the flushing outlet, the lower the location of the outlet, the more favourable the location of the outlet, the longer the flushing lasts, the narrower the reservoir (steep banks), the steeper the original stream gradient through the reservoir, the shorter the reservoir, the straighter the reservoir, the more advanced the silting (close to dam as possible), the finer the particles in the sediment, the rounder the particles of the sediment, the younger and the less consolidated the sediment. On Mosul dam, it was concluded that the main requirements in the applied criteria for flushing sediment can be verified in Mosul reservoir like the SBR, the LTCR, the topography (shape of reservoirs) (i.e., the ratio W/L), the hydraulic conditions (Qf/Qin) and (Vf/Vin) except some the hydraulic condition (draw down ratio) which did not fit completely with the criteria. In general, Mosul reservoir may be flushed, and sediment can be reduced successfully which means that the flushing process may be considered feasible. It is recommended to apply flushing procedure on Mosul reservoir and studying its effectiveness through making reservoir surveying after the flushing process, if possible, to check the degree of the success of this process. The flushing process must be evaluated briefly with the environmental effects of this process downstream of the dam and must go forward smoothly with other objectives of the dam.