Reservoir Operating Policies for Environmental Flows of Large Dams in the Zambezi River Basin

RESERVOIR OPERATING POLICIES FOR ENVIRONMENTAL FLOWS OF LARGE DAMS IN THE ZAMBEZI RIVER BASIN

F.F. NYATSANZA, S. GRAAS and P. VAN DER ZAAG UNESCO-IHE, Institute of Water Management, The Netherlands [email protected]

PRESENTATION OUTLINE

 Introduction  Background and study area  Problem Statement  Research questions  Methodology  Results  Conclusion  Recommendations and further work INTRODUCTION

 The Zambezi basin is important to its riparian countries for food energy and ecological diversity (World Bank, 2010), with the Marromeu Complex and the Kafue Flats recognised as wetlands of world importance (Ramsar sites), the Mana Pools as a UNESCO World Heritage Site.

 The interference of large scale hydropower schemes in the Zambezi basin has disrupted the environment, especially in the lower Zambezi (Beilfuss, 1999)

 The alteration of flow to the lower Zambezi by the dams has resulted in ecosystem degradation in the delta and affected the fisheries (Hoguane, 2000)

BACKGROUND AND STUDY AREA

•Zambezi Basin is 390 000 km² and 2750km long

•4 major reservoirs- Kariba, Cahora Bassa, Kafue and Itezhi Tezhi

Parameter Unit Kariba Itezhi Tezhi Kafue Cahora Bassa Construction 1955-1959 period Inflow(net) m3/s 1272 302 383 2376 Mm3/s 40114 9524 12078 74930 Live Storage Mm3 64744 5000 1178 51750 Reservoir Index Year 1.61 0.52 0.10 0.69 (residence time) Installed power MW 1350 Storage 900 2075 Source: Klassen, undated PROBLEM STATEMENT

Illustration of the changes in the hydrograph at Tete, downstream of Cahora Bassa

Source: Ronco et. al, 2010

 Despite environmental flow studies in the basin, e-flows have not been implemented due to fear of opportunity costs (i.e. hydropower foregone) (Chen, 2010)  Problem? Floods of 1978, though having a toll on human life and property, had unplanned benefits for the environment. And so were the emergency flood releases of1997 (Beilfuss, 1999)

RESEARCH QUESTIONS This study analyses current reservoir operating policies and investigates the possible effects of reoperation of the reservoirs by forcing Cahora Bassa and Kariba to release artificial floods in February and March on the hydropower production.

1. What are the current reservoir operating policies for Kariba, Itezhi tezhi, Kafue and Cahora Bassa reservoirs? 2. What is the impact of re-operating the reservoirs for environmental flows on hydropower production?

METHODOLOGY

 Data for the study was obtained through field visits in the basin in 2011  Analysis of current reservoir operating policies using historical data (1980-2010)  Reservoir simulation on a monthly timestep using Waflex (Savenije, 1995)  Scenario analysis  Reductions in hydropower produced for Kariba are measured against historical production, and not actual demand. For HCB, they are measured against modeled production over the modeled period. This is because Cahora Bassa has only been fully operational since the late 90s.

SCENARIO DESCRIPTION Scenario Description BAU Current operating conditions with hydropower driving dam operation (BAU_FC for Cahora Bassa) 1 Recover the natural flows in Feb and Mar by 50% at Cahora Bassa

2 Recover the natural flows in Feb and Mar by 50% at Cahora Bassa & Kariba 3 Recover the natural flows in Feb and Mar at Cahora Bassa

4 Recover the natural flows in Feb and Mar by at Cahora Bassa & Kariba Kariba Cahora Bassa 6000 3000 2500 5000

2000 4000 Illustration!

1500 3000

m^3/s m^3/s 1000 2000 500 1000 0 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 100% Pre-Dam BAU, Post-Dam 100% Pre-Dam Post_CB 50% recovery 50% recovery MODEL CALIBRATION (KARIBA)

HP Production Water Level 490,00 1 400 1 200 485,00 1 000 800 480,00 600 475,00 400 200 470,00 0

465,00

13 25 37 49 61 73 85 97

109 121 133 145 157 169 181 193 205 217 229

73 19 37 55 91

109 127 145 163 181 199 217 235 253 271 289 307 325 343 HP_KAR computed WaterLevel computed

Outflows R2 NSC 12000 10000 8000 HP 1 1 6000 4000 Water 0.87 0.79 2000 level

17 33 49 65 81 97

129 145 161 177 193 209 225 241 257 273 289 305 321 113 Outflows 0.49 0.48 Outflows computed RESULTS

 The reservoirs operate based on rule curves: - Cahora Bassa and Kariba have maximum rule curves - ITT and Kafue have minimum rule curves

 The main operation objectives in the Zambezi basin are:  Maximizing hydropower production  Ensuring dam safety  Ensuring sufficient storage for flood water for flood protection (Kariba and Cahora Bassa) RESULTS Comparison of historical water levels and the reservoir rule curves

Rule Curve Rule Curve (2005+) Rule Curve vs. Lake Levels: Cahora Bassa 1981/82 Kariba Lake Levels 1981/82 335 1982/83 1983/84 490 1982/83 1984/85 1983/84 1985/86 330 1986/87 1984/85 1987/88 1985/86 1988/89 485 1986/87

325 1989/90

1990/91 1987/88

1991/92 1988/89 320 1992/93 480 1989/90 1993/94 1994/95 1990/91 1995/96 315 1991/92 1996/97 1997/98 475 1992/93 Water Water Level(m) 310 1998/99 1993/94

1999/00 Water Level(m) 1994/95 2000/01 2001/02 1995/96 305 2002/03 470 1996/97 2003/04 2004/05 1997/98 300 2005/06 1998/99 2006/07 465 1999/00 2007/08

2008/09 2000/01

16-Jul

01-Oct 19-Oct

08-Sep 26-Sep

12-Dec 30-Dec 17-Jan 04-Feb 22-Feb

17-Apr

10-Jun 28-Jun

06-Nov 24-Nov

03-Aug 21-Aug

12-Mar 30-Mar 23-May 05-May 2009/10 2001/02 2010/11 RESULTS

1981/82 1981/82 1982/83 Kafue Gorge Lake Levels ITT Lake Levels 1982/83 1983/84 1983/84 978 1984/85 1035 1984/85 1985/86 1985/86

1986/87 1986/87 1987/88 977 1987/88 1030 1988/89 1988/89 1989/90 1989/90 976 1025 1990/91

1990/91 1991/92

1991/92 1992/93 1992/93 1993/94

1020 975 1993/94 1994/95 1995/96

1994/95

1015 1996/97 1995/96 1997/98 974 1996/97

level(m) 1998/99 1997/98 1010 1999/00

Water Water Level(m) 1998/99 2000/01 973 1999/00 2001/02 1005 2000/01 2002/03

2001/02 2003/04 2004/05 972 2002/03 1000 2005/06 2003/04 2006/07 2004/05 995 2007/08 2005/06 971 2008/09 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 2006/07 2009/10 2007/08 Rule Curve RESULTS Kariba

• No hydropower Kariba Hydropower production 10000 shortages in normal 9000 years for scenario 2 8000

7000

6000 • Hydropower 5000 4000 shortages of up to GWh/year 3000 36% in dry years for 2000 both scenarios (2 & 1000 0

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11

1990/91 1991/92 1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 1992/93 1993/94 1994/95 1995/96 • Hydropower BAU Scenario2 (50%) Scenario4 (100%)

shortages of up to Max Units BAU Scen 2 Scen 4 24% in normal years potential 6250 6066 HP GWh/yr 11232 6556 in scenario 4 KAR (4.7%) (7.5%)

RESULTS Cahora Bassa

Cahora Bassa HP production (full capacity demand) • Cahora Bassa would 20000 be able to produce full 18000 16000

power in normal years 14000 but not in dry years 12000 10000

8000 GWh/year • Shortages of up to 6000 36% 4000 2000 • Reduced to 27% when 0

Kariba releases for e-

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11

1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 flows too (scenario 2), BAU_FC Scenario 1 (50%) scenario 2 (50%) scenario 3 (100%) scenario 4 (100%)

•When Kariba fails, Max Units potentia BAU_FC Scen 1 Scen 2 Scen3 Scen 4 then the shortages are l GWh/y 12312 11933 11865 more (2005/06 & 17928 12334 12334 CB r (0.2%) (3.2 %) (3.4%) 2006/07) RESULTS

Example of flow results : EFR_Feb vs. BAU

Kariba Scenario 4. 20000

15000

10000 •Shows the average 5000

natural flows for Feb and 10^6m^3/month March can be met in

normal years. 0

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11

1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 •However the frequency BAU Scenario4 100% EFR_Feb and magnitude of bigger floods is reduced. EFR_Mar vs. BAU 20000

•Although it may contribute to flood 15000 management, it is not 10000 known whether which 5000 flooding regime would be 10^6m^3/month more beneficial (every 0

year, or every few years)

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11

1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 BAU Scenario4 100% EFR_March CONCLUSION

 The rule curves for Kafue and ITT are not strictly adhered to  Cahora Bassa can release e-flows to recover the hydrograph by 50% or 100% in normal to wet years, without large negative impact on HP production. However, fewer shortages are experienced when Kariba also releases for EFR  Therefore flood releases in February and March in the Zambezi are possible in normal to wet years without significantly affecting hydropower production (depends on the magnitude of EFR).  However, the frequency and magnitude of very large floods is reduced by imposing these regular EFR flood releases.

RECOMMENDATIONS AND FURTHER WORK

 Further studies are needed to determine the environmental flow requirements and dynamics, especially inter-annual variations  Compare environmental flow requirements at more locations in the basin, and find synergies between the operation of reservoirs  Analyse low flow requirements and its trade-offs  Study the economics of environmental flows in the Zambezi for tradeoff analysis and cost/benefit sharing policies between hydropower and the environment in the basin ACKNOWLEDGEMENTS

 This work was part of the POWER2FLOW project financed by the UNESCO‐IHE Partnership Research Fund (Netherlands). In addition, the authors would like to thank HCB, Fisheries Research-Songo, ARA Zambeze, ZRA, ZESCO Limited, DWA (Zambia), Meteorological Office (Zambia), WWF and WaterNet. Without these organisations and their staff, this piece of work would not have been possible.