Divundu Hydro Power Project

1 Introduction

1.1 General
The Okavango or Kavango River (known as Cubango in Angola) originates on the Benguela Plateau in Angola. The rainfall at the plateau is between 1 200 and 2 000 mm pr. year. From here the river flows in a south eastern direction as a boundary river between Angola and Namibia for approximately 400 km. In Caprivi, the Okavango River flows through Namibia for a distance of approximately 50 km before it enters Botswana and the Okawango Swamps. The rainfall in this region is only 400-600 mm pr. year.

The mean flow is 365 m3/s at Popa Falls in Caprivi. The natural drop at the Popa Falls is approximately 2,5 m in the wet season and 4 m in the dry season. The falls is more similar to rapids with a width of approximately 1000m.

1.2 Existing situation in the Okavango River
Powerplants, irrigation, water supply A small hydro power plant (50 kW) is in operation between Divundu and the Angola border, approximately 20 km upstream the Popa Falls. 

Roads etc.The major route, B 8, crosses the Okavango River at Divundu. Secundary roads are also located at  the river banks.

Transmission lines There is a 33 kV transmission line from Rundu to Divundu.

Others The actual area for the hydropower developement is located in Divundu with such facilities as  shops, electricity supply, water supply, hospital, airport (small aircrafts) and restcamps (in connection with the attraction Popa Falls). 

2 Available Data / Assumptions

General: The planning is based on data from the studies in the sixties, some later drafts, new maps and a site visit.

Maps: 1: 50.000 (1998), 1: 25.000 (1968), partly 1: 4500 (1968)

Hydrology: For all alternatives an average annual runoff of 11 500 mill.m3 is used. This figure is presented in the studies from the sixties. This is based on a period of 21 year (1947-68) at Mukwe gauging station close to Andara near Popa Falls. In fact, the gauging station was opened in 1965, but based on earlier gauge post observations, the record has been extended to earlier years by S.W.A. Water Affairs Branch. The lowest discharge in the low flow period is 160 m3/s. The discharge is higher than 200 m3/s in 10 months, and average annual discharge is 365 m3/s.

Sediment transport: No data on sediment transport is found, but has to be taken into consideration. Therefore flushing-arrangements is assumed.

Geology: NA

Seismology: No seismic investigations are done at possible plant sites.

Simulations: No EDB-based simulation is done. The calculation of the energyproduction is based on a flow-duration curve. This is made on the basis of a flow-duration curve for Okavango at Rundu and information from the low flow period at Mukwe.

Cost estimate: The cost curves from the Norwegian Water Resourses and Energy Directorate (NVE) are used. The level of cost is primo 1999.

Uncertain factors/ Need for further investigations: The flow-duration curve based on long term observations at Mukwe should be constructed based on updated data from Rundu and Andara gaging stations. Profiles of the actual damsites should be taken and also seismic refraction surveys. The actual masl should be chequed both upstream and downstream the Popa Falls during the year.

Budget prices for the electro-mechanical equipement and local prices for concrete, construction etc. should be checked in the local market. Possible places for aggregates for production of the concrete ought to be investigated.

Environmental and socio-economic studies have to be carried out. The tourist attraction Popa Falls and resettlement of people are keywords here. This is important before deciding on the site and size for the weir and the power house building.

Multipurpose aspects ought to be investigated.

3 References
Okavango River Prosject, Popa Falls Hydro-Power Scheme, Preliminary Feasibility Investigation by Hydroconsultants in collaboration with and at the direction of Director of Water Affairs, S.W.A Branch, August 1969. 

4 Divundu Hydro Power Plant
4.1 The Main Data

Divundu hydro power plant

Units

Alt. A Alt. B
Total capacity at mean head MW 19 20
Total energy production  GWh 150 160
Total cost primo 1999  mill. N$ 297 345
Unit Investment cost 1) Ncent/kWh 23 25

1) Annually distributed investment cost plus cost of operation and maintenance divided by the average annual production. 10% discount rate, 40 year lifetime, annual maintenance/operating cost 0.03 N$/kWh. 

4.2 Project Description
4.2.1A Description of Alternative A
Appendix 1: Alt A; Map 1:50 000

The weir will be located approximately 2 000 m upstream of the "Popa Falls" in Okavango River close to Divundu village in Caprivi. A gross head of 9 m in the wet season and 10 m in the dry season will be obtained by construction of the 6 m high weir (from surface riverbed to to the highest regulated water level, HRWL). The weir will be 930 m long. The reservoir inundates 1.4 km2 of land. The water surface will normally be kept close to HRWL with option of 0.5 m regulation. The scheme will not affect Angola (inundation of land), and only marginally affect Botswana due to approximately 1.5 mill m3 annually increased evaporation. The power plant will be located at the outlet in the river downstream of the "Popa Falls". The total length of the waterway is 2 500 m. The access to the plant will be from the existing road, B 8. The plant will be connected to the local grid by a short transmission line. 

4. 2.1B Description of the Alternative B
Appendix 2: Alt B; Map 1:50 000

The weir is located on the Popa Falls. The power station is located downstream of the rapids similar to alt. A. The gross head is approximately 9 m. The weir is 6 m high and the length is 2 150 m. The reservoir will inundate 2.1 km2 of land, and evaporation will be approximately 2 mill m3 annually. 

4.2.2 Catchment Area - Runoff

Part of
Catchment
Area
Identification Area
km2
Specific 
runoff 
l/s/km2

Mean inflow

m3/s
mill m3/year
1 Okavango at 
Divundu
Approx 203 000 1.8 365 11500

 

4.2.3 Reservoirs

Alternative Name of
Reservoir

Before regulation

After regulation

Active 
Reservoir
    Area at
NWL
km2
NWL
masl
Area at 
HRWL
km2
HRWL
masl
LRWL
masl
mill. m3
A   1.5 1001 2.9 1006 1005.5 1.1
B   2.4 1001 4.5 1006 1005.5 1.7

The inundated area is measured to 1.4 km2 in alt A, and 2.1 km2 in alt B. The transport of sediments in the river is unknown, but arrangement for flushing is assumed. A gravity-concrete weir is assumed (rockfill can be used as an alternative at both sides).

Alt. A:
The weir will be 8 m high (from cleaned damfoot to HRWL) and the crest 930 m long. The total volume will be 6 000 m3.

Alt B:
The weir will be 8 m high (from cleaned damfoot to HRWL) and the crest 2 150 m long. The total volume will be 35 650 m3.

Rock is exposed at the damsites. Some excavation and grouting are necassary to obtain a watertight and stable damfoot. 

Diversion during construction will be obtained by means of cofferdams, culvert(s) and gates. It is assumed that some overflow during construction will be accepted.

The spillway is calculated as a fixed concrete weir. The alternative could be fixed labyrinth weir or with  inflatable rubber gates. The Q1000 flood is estimated to be 8 000 m3/s with a max flood level up to  HRWL+2,5 m in alternative A, and HRWL + 2 m in alternative B. The alternative with inflatable rubber  gates will reduce the increase of water level during floods, and give an option for increased head during  low flow periods.

4.2.4 Waterways
Section (from – to)
Type
Length, m
Cross-section 
(m2 )
Diameter (m)
Alternative
Inntake – cone
Headrace tunnel / canal
2 000
150 m2
A
Cone-turbine
Steel
10
(2 x 5) m
A
Turbine – outlet
Tailrace canal
500
150 m2
A





Intake-turbine
Steel
10
(2 x 5) m
B
Turbine – outlet
Tailrace canal
500
150 m2
B


4.2.5 Power Station
The station will be located downstream of the "Popa Falls" on the left river bank.


Alt. A
Alt. B


Dry 
season
Wet 
season
Dry 
season
Wet 
season
Underground / open powerhouse

Open
Open
Access tunnel
m
None
None
Capasity at max head
MW
20
21
Maximum flow 
m3/s
250
250
Type of turbine

Kaplan
Kaplan
Number of units

2
2
Tailwater (Dry/wet season)
masl
996 / 998
996 / 998
Voltage
kV
33
33


4.2.6 Peak Power
The future value of peak power will probably give a higher optimum capacity. The plant can be operated to meet daily peak power demand.


4.2.7 Data on the powerplant (with restrictions; 20 m3/s water release during 5 months)

Divundu Hydro Power Plant

Dry / wet 
season1)
Dry / wet 
season1)

Alt. A
Alt. B
1
Hydrological data


Catchment area 
km2
?200 000
?200 000

Mean annual inflow 
mill m3 
11 500
11 500

Reservoir volume as % of annual 
inflow
mill m3 / %
1.1/0
1.7/0
2
Power Station data


Mean gross head 
m
10/9 
10/9

Mean energy equivalent 
kWh/m3
0.023/0.020 2)
0.024 / 0.022 2)

Max flow at mean head 
m3/s
250
250

Max capacity at mean head
MW
19/16 3)
20/18 3)

Utilization time 
hours
~6 500
~6 500
3
Energy, average 


Energy production, firm 
GWh
126
135

Energy production, surplus
GWh
24
25

Energy production annual 
GWh
150
160
4
Cost estimate/Economy


Construction period 
years
3
3

Cost estimate primo 1999 
mill N$
297
345

Specific cost 
N$ /kWh
1.98
2.16

Unit investmentcost 4)
Ncent/kWh
23
25
1) 60 % of the year is estimated as dry season and 40 % as wet season.
2) The energy equivalent is based on the net head, which is lower in Alt A than B due to energy loss in a longer waterway.
3) The output during the dry season will be 15 - 19 MW and 16 - 20 MW in the wet season.
4) 10% discount rate, 40 years lifetime, annual maintenance / operating cost 0.03 N$/kWh.


4.2.8 Access roads

The power plant will be located close to the existing infrastructure, which are: roads, grid, restcamps, shops, hospital, airport (small aircrafts) etc. The need is only to build short access roads to the powerstation and the weir.
Section (from – to)
Type
Length, km
Temporary/Permanent
Exist. road-station
Upgrad./new tarred road
2
Permanent
Exist. road- weir
Gravel surfaced
2
Permanent


4.2.9 Grid
The powerplant is located close to the local 33 kV-grid, which also connects Divundu to Rundu through a 200 km 33 kV-line The connection to the local 33 kV-grid will be permanent. The power from Divundu will mainly be used in the Caprivi area, but will also be a supplement to national demand. The need for upgrading the transmission line to Rundu has not been evaluated. However, the Divundu scheme will enhance connection of the Eastern Caprivi (Catima Mulilo) to the Namibian transmission network.
Section (from-to)
Lenght, km
Voltage, kV
Capacity existing transmission line 
(assumed)
Powerstation - grid
2
33
30 MW

4.2.10 Tele Communication
Local grid will be used.

4.2.11 Location of (Tunnel) Masses
The tunnel/canal masses will be placed near the power station. As an alternative the masses can be used for construction of roads or for other purposes.

4.2.12 Borrow pit - Mass – Quarry
It is assumed that the aggregates for the concrete work can be taken from suitable places in the neighbour-hood.


4.2.13 Costs
Costs (mill N$) pr 01.01.99 (10% pa interest during construction period)


Divundu Alt. A
(Dam wall above the rapids) 
Divundu Alt. B
(Dam wall on the rapids) 
1
Reservoir, weir cost
39
117
2
Diversion 1)
0
0
3
Waterway (Penstock included)
61
17
4
Power station (civil works)
14
14
5a
Power station (mechanical)
39
39
5b
Power station (electrical)
41
41
6
Access roads, Temporary lines, Transportation
4
4
6b
Permanent transmission lines
1
1
7
Housing. Workshop. Admin. block. Storage, etc.
5
5
8
Landscape design, etc
5
5
9
Contingency 
17
20
10
National taxes 
0
0
11
Planning. Administration)*
23
26
12
Compensation. etc.**
8
9
13
Financing costs (I of 1-12)***
40
47

Total estimate cost
297
345
1) Included in reservoir costs.

Specific cost: Alt A: 1.98 N$/kWh, Alt. B: 2.16 N$/kWh, 
Based on mean annual production. 

* Planning/administration are calculated with 10% for civil work, and 5% for electrical and mechanical work.
** 4% of 1 – 6 is used, which is normal according to experiences from other hydropower scehemes. 
The mitigation for each scheme will decide the actual level.
*** Calculation formula:
I = (1.05(1.10 t - 1))/(0.10 t) - 1, "t" is construction period in years
"I" is to be multiplied with the sum of 1 – 12

Unit investment cost with different discount rate (Ncent/kWh)
Alt
7 %
10 %
15 %
A
17.2
23.2
37.9
B
18.5
25.1
35.0


4.3 Hydrology, Consequenses

4.3.1 Use of reservoir
The level of the reservoir will be held close to HRWL most of the year. (If peak power is needed, a small regulation (0.5 m) can be used. The length of the reservoir at HRWL will be approximately 4 km in alt. A, the surface area is 2.9 km2, and the inundated land area is 1.4 km2. The length of the reservoir at HRWL is 6 km in alt. B, the surface area is 4.5 km2, and the inundated land area is 2.1 km2.

4.3.2. Discharge
The capasity of the plant is 68 % of Q mean. This means that in approximately 70 % of the year there will be water over the weir and so in Popa Falls. The rest of the year at least 20 m3/s compensation water is included in the calculations. This is released through the weir.

4.3.3 Flood
Implementing the reservoir will not make any influence on the floods or waterlevel in the river downstream of the plant. Upstream of the weir a certain backwater will occur, but in this case the effect will be small and will not affect Angola.

4.4 Compensated Initiatives
4.4.1 Planned Compensation
Resettlement of directly affected people is included in the costs. This cost will depend on the local situation. In this case we have used 4% of 1 – 6 in the cost to come up with a general figure. The 4% is normal according to experiences from other hydropower schemes. The mitigation for each scheme will decide the actual level. In the cost estimate 8 mill N$ for alt A, and 9 mill N$ for alt B is assumed to be used for compensation.

The minimum amount of water release is assumed to be 20 m3 / s in approximately 5 months.


4.4.2 Possible Compensation
Peak power operation is possible during the day, and should be discussed in later studies in order to highlight possible negative environmental impacts and actions to reduce them. 


5 Socio-economic and Environmental Considerations

The plant will have a long utilization time, and be in operation nearly all the year. The reservoir created by the weir is not big enough to manipulate the river flow, therefore it will always be water in the river downstream of the scheme. 

The inundated area is approximately 1.4 km2 in alt A, and 2.1 km2 in alt B. The scheme will not affect Angola. Approximately 75 huts, 15 houses and 4 campsites around Divundu will be affected by the increased water level. A socio-economic programme must therefore be established to secure new land and improved life for the people directly affected. Due to the increased water surface, the annual evaporation will increase by approximately 1.5 mill m3 (alt A) and 2 mill m3 (alt B), and thereby have a minor impact on the flow into the Okavango Swamps. However this flow is 11 500 mill m3 annually at Divundu.

The rapids inhabit a small-sized rare fish species, which could be negatively affected during construction, and later operation of the plant. The biological life in the river between the dam and the outlet from the powerstation must be thoroughly studied and measures taken to prevent severe impacts. 

The scheme will enhance the connection of the Caprivi to the main grid in Namibia. Secured electricity supply is a prerequisite to develop the agriculture potential in Caprivi. This added value to the local society should be investigated.

6 Other alternatives

The plant has not been optimised. In addition to the two described alternatives there are several alternative heads, capasities, damsites, spillways and combinations that should be investigated at later stages of the planning.
1

2

01.09.00

ABBREVIATIONS:

NWL Normal Water Level
HRWL Highest Regulated Water Level
LRWL Lowest Regulated Water Level
masl meter above sea-level

UNITS

Output 1 MW = 1000 kW
Energy 1 GWh = 1 mill kWh
Voltage 1 kV = 1000 V