Financial risks associated with hydropower development

Numerous studies have indicated that hydropower economics are sensitive to changes in precipitation and runoff (Alavian et al. 2009; Gjermundsen and Jenssen 2001; Mimikou and Baltas 1997; Harrison and Whitington 2001, 2003). Uncertainty about future hydrology presents a great challenge for infrastructure planning and engineering. Most hydropower projects are designed on the basis of recent climate history (typically a 30-50 year historic time series of flow data) and the assumption that future hydrological patterns (average annual flows and their variability) will follow historic patterns. This notion that hydrological patterns will remain "stationary" (unchanged) in the future, however, is no longer valid (Milly et al. 2008). Under future climate scenarios, a hydropower station designed and operated based on the past century's record of flows is unlikely to deliver the expected services over its lifetime. It may be over-designed relative to expected future water balances and droughts, as well as under-designed relative to the probability of extreme inflow events in the future.

Over-designed projects, resulting from reduced and more variable inflows relative to the historical time series, incur financial risk by generating lower levels of power production than forecast, leading to reduced electricity sales and revenue, including failure to meet firm energy commitments. Capital costs for hydropower are high compared with alternative energy options, and the financial risk of over-design is significant (World Bank 2010). Development of the hydropower sector according to the generation plan of the Southern African Power Pool (NEXANT 2007), for example, will require an investment of $10.7 billion over an estimated 15-year period. A comparable investment in energy efficiency and renewable technologies including biomass, solar, wind, and small-scale hydro, would aggressively expand decentralized (on- and off-grid), clean energy access and markets in Africa (Hankins 2009).

Financial and technical analyses to assess the feasibility of hydropower projects typically evaluate the financial impacts of a range of factors on the ability to generate a positive cash flow; these analyses apply traditional engineering cost/financial analysis to characterize construction and operational costs (e.g., size and location of the project) and future trends that could affect project revenues (e.g., changing demand, new supply, and economic drivers affecting the price of electricity). These assessments rarely evaluate potential power generation and associated revenue changes associated with climate change. When climate considerations are incorporated, the financial risks may significantly undermine the feasibility of existing and future hydropower projects.

The regional economic impacts of reduced hydropower generation from Kariba Dam during the 1991-92 drought, for example, included an estimated $102 million reduction in GDP, a $36 million reduction in export earnings, and the loss of 3,000 jobs (Magadza 2006). Droughts of this magnitude (or worse) will occur more frequently with regional climate change.

Harrison and Whittington (2002) examined the susceptibility of the proposed Batoka Gorge hydroelectric scheme to climate change, with an emphasis on financial risks associated with the project. They reconstructed a flow series for inflows to Batoka Gorge, using the U.S. Army Corps of Engineers HEC-5 reservoir routing program. Inflows to the model were generated using rainfall-runoff models based on precipitation according to three different climate change scenarios (IPCC 2001). Their simulations suggest a strong sensitivity of the Batoka Gorge project to changes in climate. The models indicate significant reductions in river flows (mean monthly flows fell between 10-35%, and both wet season and dry season flows declined), declining power production (mean monthly production fell between 6-22%), reductions in electricity sales and revenue, and consequently an adverse impact on a range of investment measures. Harrison et al. (2006) note that climate change scenarios alter not only the financial performa nce of hydropower schemes such as Batoka, but also the financial risks they face. Changes in climate lead to significant variability in economic performance – reducing not only the mean values for energy production, but also the reliability of electricity sales income.

By building the "wrong" infrastructure in future, we may actually limit our future options for climate adaptation. An alternative path, focused on climate-smart investments that factor in financial risk and the ecological functions of river systems, is urgently needed.

In the face of hydropower blackouts caused by low water levels, governments are often forced to buy expensive emergency power, which is not included in the risk analysis for large-dam hydropower. For example, after the 2009 drought in Kenya brought reservoirs to their lowest levels in 60 years, the government brought in Aggreko PLC, a U.K. firm that supplies temporary diesel generators. For an extended period, reports the New York Times.1 Aggreko was delivering roughly 140 MW at a cost of $30 million per year, not including fuel purchases. Meeting future needs through diesel generation could cost the Kenyan government more than $780 million a year – a key reason Kenya is now building wind farms and geothermal plants to bring its hydro-dependency down from 60% to 35%.

Hartman (2008) notes that hydropower planners have been aware of climate change for years, but until recently it was assumed that climate trends were too uncertain, and the range of natural variability too high, to make reliable predictions. From a financial point of view, it was argued that changes beyond 20-30 years from present would have little impact on the financial return of hydropower investments– introducing a mismatch between financial time horizons and water resource management implications, as the physical lifespan of hydropower assets is much longer than the pay-back period. Large and financially powerful hydropower operators from the temperate regions, who might be expected to lead the way in terms of new policy and research, are also the ones expected to be less affected by climate change. Environmental impact assessments and other planning guidelines still do not usually include guidance on hydrological variability and climate change, beyond the impact of extreme flood events on dam safety. To gether, these factors have led to a neglect of climate change risks in hydropower planning – in an approach that might be called either "wait-and-see" or "head-in-the-sand" (Hartman 2008).

Among the major hydropower projects in operation or planning for the Zambezi River, the financial risks of climate change were considered by hydropower developers only for the Kafue Gorge Lower project (Stenek and Boysen 2011). This analysis, for the International Finance Corporation (IFC), combined three GCM models and two SRES emission scenarios to project a set temperatures and precipitation levels over four time periods (base, early-, mid-, and late-century) for the Kafue River Basin. The outputs from each GCM/emission scenario combination were used as inputs for the hydrologic flow modeling of the Kafue River Basin, which provided climate-modified flow rates across four time horizons for each of the GCM/emission scenarios. The flow series was routed through a reservoir model to assess energy production, and a financial risk model. IFC results indicate that future emission projections have a significant impact on the operations, and therefore the financial viability, of Kafue Gorge Lower project. None of the scenarios exceeded the average annual generation of about 2,450 GWh needed to satisfy investor requirements, and most of the scenarios considered did not yield acceptable returns to investors. The study notes that, "given the significance of water flow on the financial viability of hydropower projects, adaptation planning should include considerations such as climate change, conservation, and development that introduce variability into available water flow to the project." The study concluded that climate change will significantly impact the financial performance of ZESCO's hydropower plants, with the financial viability of hydropower investment dependent on the relative severity of climate change on the basin. These impacts highlight the importance of considering changes in water supply due to climate change when implementing financial analyses for hydropower projects. Governments and investors must become better informed about climate change risks to future hydropower projects by analyzing projects for projected changes in available water flow and power generation, rather than assuming constant flows and power generation rates.

The financial risks of climate change are not under serious consideration for other proposed hydropower projects in the Zambezi River Basin. The design and operation of Mphanda Nkuwa Dam in Mozambique, for example, assumes the continued validity (stationarity) of the mean and variability of the historic flow series, despite climate change forecasts to the contrary. The project has not been evaluated for the risks associated with reduced mean annual flows and more extreme flood and drought cycles, which include the risk of structural failure if the design flood is underestimated, financial risk associated with overestimated firm power generation, reduced revenue from total energy production, and other uncertainties.

Under-design of hydropower projects also poses significant financial risk with respect to future climate change scenarios. The occurrence of extreme flooding events on a more frequent basis (Boko et al. 2007) may threaten the stability of large dams and/or force more frequent spillage, which exacerbates downstream flood damage. The design flood rule curve that governs the risk associated with over-topping Kariba and Cahora Bassa Dams, for example, is based on the historical hydrological record and may not result in adequate reservoir storage capacity for large flood events. The financial and social impact of a major dam failure in the Zambezi River Basin would be nothing short of catastrophic.