Track II Project level analyses. Robert Lempert, RAND

Transcription

1 Track II Project level analyses Robert Lempert, RAND 1

2 Project Has Two Tracks Analyses consistent across basins and power pools (Track I) Water Energy 2

3 Project Has Two Tracks Analyses consistent across basins and power pools (Track I) Project level analyses (Track II) Conduct five case studies, each focusing on a specific project Goals of project-level analyses: Benefit managers of the projects studied Enrich response options analysis in basin-level studies Develop widely usable tools Bank and others can use 3

4 Initial Case Study Considers Malawi s Lower Fufu Hydro-Power Project Run of the river hydro River diversions via two catchment dams and tunnels North Rumphi (15.5 m3/s capacity) South Rukuru (15.5 m3/s capacity) 90 MW Hydropower 4

5 Analysis Builds on 1995 Prefeasibility Report Specifications from 1995 report: Hydropower capacity: 90 MW Capital cost: $118.75M (1995 USD) Annual fixed operating costs: $1.78M (~1.5% of capital cost) Estimates for Pilot Study: Firm power (historical hydrology): 479 MWhr/yr Capital cost: $290M (current dollars) Annual fixed operating costs: $1.27M/year Levelized energy cost: US$/ KWhr 5

6 Analysis Stress-Tests Project Across Wide Range of Plausible Futures Analytic Approach Use hydrological and financial model to evaluate performance metrics under many future conditions (climate and other uncertain variables) Identify cases where performance is unacceptably low Consider design choices that can reduce vulnerability in as many cases as possible Design choices include Diameter of supply tunnels Plant capacity 6

7 Overview of Pilot Analysis Examine regret of alternative designs over a wide range of plausible climate futures Regret in any future is difference between the performance of a design and the best design in that future Analysis suggests: The best design for historic climate is about 67% larger than that in the pre-feasibility report The most robust design over a wide range of futures is about 5% smaller than the best design for the historic climate 7

8 Simple Analytic Framework Used to Evaluate Performance of Lower Fufu Hydropower Project Design and Cost Tool Alternative design flows (m3/sec) Capital costs O&M costs Levelized Cost ($/KWhr) Alternative project specifications Climate projections Annual hydropower production WEAP Hydrological Model 8

9 Preliminary Scoping of Analysis Uncertainties (X) Water Management Strategies (L) Climate Temperature Precipitation Model (R) WEAP model of hydrology Design and cost tool Performance Metrics (M) Levelized cost Flow rate of North and South tunnels 9

10 Optimal Design for historical climate is larger than pre-feasibility design Levelized Cost ($/KWhr) Project Design Flow (m3/sec) While the historical optimal design has a hydro generation capacity of 148 MW, which is considerably larger than the pre-feasibility design of 90 MW, it is within the range of designs currently under discussion by AfDB and World Bank1,2. Source: 1. AfDB Malawi Country Strategy Paper project 2. World Bank (2011). Republic of Malawi, Project appraisal document: Energy sector support

11 Simulated Hydropower Production Over Time: Historical Climate Historical Optimal Design (51 m3/sec) 11

12 Simulated Hydropower Production Over Time: 56 Climate Projections Historical Optimal Design (51 m3/sec) 12

13 Wide range of production and costs over 57 climate projections Number of Projections Historical climate Average Annual Hydropower Production (GWhrs) Historical climate Historical Optimal Design (51 m3/sec) Number of Projections Levelized Cost ($/KWhr) 13

14 Hydropower production strongly related to precipitation 56 GCM Projections + Historical Average Annual Hydropower Production Range: x y GWh Historical Optimal Design (51 m3/sec) 14

15 Range of generation and regret for other designs Hydropower Generation (GWhr/yr) Levelized Cost Regret ($/KWhr) 15 Project Design Flow (m3/sec)

16 Low regret designs vary depending on expected future climate No regret Low regret Project Design Flow (m3/sec) Annual Mean Precipitation (mm) 16

17 Smaller designs slightly more robust over drier climates Levelized cost regret across precipitation Alternative Designs Annual Mean Precipitation (mm) 17

18 Smaller designs slightly more robust over drier climates More robust options: Lower regret over wider range of climate Annual Mean Precipitation (mm) 18

19 Summary A design of 51 m3/sec seems optimum for historic hydrology Designs between 39 and 61 m3/sec all have zero regret for some future climate projection Designs between 47 and 49 m3/sec seem most robust, unless decision makers believe a wetter future is likely 19

20 We Plan Four Additional Case Studies Some case studies under consideration: Criteria include: Project in one of the seven basins considered in Track I analyses Project between pre-feasibility and feasibility study phases Sufficient data in pre-feasibility study to inform the simple, excel-based model Interest among relevant decision makers 20