Responses of hydropower generation and sustainability to changes in reservoir policy, climate and land use under uncertainty: A case study of Xinanjiang Reservoir in China

Abstract

Climate and land use changes will affect the hydrological regime, and therefore hydropower. This study which aims to develop a novel modeling framework, does not only determine the changes in hydropower generation and sustainability, but also provide robust operating rules for handling uncertainty attributed to both climate and land use changes, using Xinanjiang Reservoir in Eastern China as a case study. Specifically, projections of five bias-corrected and downscaled General Circulation Models (GCMs) and three modeled land uses representing a range of tradeoffs between ecological protection and urban development are employed to drive the Soil and Water Assessment Tool (SWAT) and to predict streamflow under 15 scenarios. We then develop a set of robust rule curves to consider the potential uncertainty in reservoir inflow and to increase hydropower generation, and a baseline rule is presented for comparison. Results show that both robust and baseline rules increase hydropower generation with increasing reservoir inflows in future, but the robust rule yields better hydropower generation, sustainability and efficiency. The streamflow under the rapid urbanization scenarios differs from that under other scenarios, but there are no significant differences in hydropower among scenarios corresponding to the non-linear relationship between streamflow and hydropower change. Our findings highlight the potential to improve water resource utilization in the future, especially based on the robust operating rule considering optimization and uncertainty, and can provide references for future hydropower planning to the other existing plants.

Introduction

Renewable hydropower plays a key role in human lives and economic development (Teotónio et al., 2017). According to the 2019 Hydropower Status Report, the world’s hydropower has been steadily developing with an increased hydropower capacity of 21.8 GW in 2018, and the total installed capacity reached 1,292 GW. To date, China’s hydropower sector has covered over a quarter of the world’s installed capacity. It is well known that China has been planning and implementing a number of hydropower plants with reservoirs in different regions, such as the Xinanjiang Reservoir (Vonk et al., 2014) in Qiantangjiang Basin, Three Georges Reservoir (Qin et al., 2019) and Danjiangkou Reservoir (Zhang et al., 2019) in Changjiang Basin, and Xiaolangdi Reservoir (Wang et al., 2016) in Yellow River Basin. These reservoir projects ensure that hydropower remains a vital and stable component of the electricity supply with inherent uncertainty in China.

Hydropower generation can be strongly impacted by hydrological regime, and thus an important issue being discussed recently is how hydropower changes in the future. Climate and land use changes are the two widely believed factors profoundly affecting hydrological processes (Alaoui et al., 2014; Ning et al., 2016; Abera et al., 2019), and thus hydropower generation. However, hydropower generation and climate change, or hydropower generation and land use change both have a coupled relationship, and they interact with each other. On one hand, climate change indirectly influences streamflow through variations in temperature, precipitation, and evaporation (Ahn and Merwade, 2014; Guo et al., 2019), and land use change alters variables such as evapotranspiration, groundwater re-charge and overland flow, resulting in changes in streamflow (Molina-Navarro et al., 2014; Zhang et al., 2017). Furthermore, the spatiotemporal variations in streamflow will change the reservoir inflow and thus affect hydropower generation. On the other hand, hydropower is beneficial to reduction in the dependence on fossil fuels and the emissions of greenhouse gas (GHG), and reservoir operation may contribute to mitigating local climate change (De Queiroz et al., 2016); meanwhile, hydropower improves the energy supply security and reliability accelerating economic development and supporting a growing global population, potentially resulting in land use re-distribution. Given the significant role of hydropower, to assess the individual and combined climate and land use changes on hydropower generation is critical for sustainable development.

Numerous studies have investigated how climate change impacts streamflow and hydropower generation, and most of them have applied General Circulation Model (GCM) projections to quantify these effects (Schaeffer et al., 2012; Boehlert et al., 2016; Kim et al., 2017; Mendes et al., 2017; Falchetta et al., 2019; Zhong et al., 2019). Turner et al. (2017a, 2017b) explored the possible impacts of climate change on global hydropower by an aggregated hydrological and hydropower plant operating model depending on GCMs; the former found that the majority of hydropower plants experienced reductions in hydropower production under all emissions scenarios, and the latter highlighted the disagreements in the direction of changes in hydropower production at the global scale. Moreover, regional studies conducted in many regions of the world have noted that hydropower generation was projected to increase in some areas and decrease in others under climate change (Chilkoti et al., 2017; Teotónio et al., 2017; Arango-Aramburo et al., 2019; De Queiroz et al., 2019). For example, Mendes et al. (2017) reported that hydropower outputs was reduced with decreasing streamflow in Iberian area by coupling GCMs and a hydrology-reservoir model, while Zhong et al. (2019) followed a similar method to predict future hydropower in the Lancangjiang hydropower base, and noted that increased reservoir inflows would cause increments in hydropower outputs for most GCMs. Additionally, the effects of land use change on hydrological processes have been widely discussed (Zhang et al., 2015; Zuo et al., 2016; Desta et al., 2019). However, to our best knowledge, there has been very limited efforts to review the impacts of land use change, and even fewer studies have focused on the combined impacts of climate and land use changes on hydropower variations.

Accordingly, the most common way to evaluate how future changes impact hydropower generation is by coupling hydrological models with climate or land use models to first simulate the streamflow response and then estimate hydropower generation based on the relevant operation policies that describe the relationship between streamflow and hydropower generation. However, the earlier mentioned studies mainly focused on the assessment of potential hydropower changes based on a power output equation with a constant hydraulic head. They did not consider the use of optimal reservoir policy as well as the ability of operating policy to mitigate the impacts attributed to future uncertain climatic and/or land use changes. The differences in hydropower plant operation policies considering climate and land use changes remain unclear. A number of methods to address uncertainty in reservoir operation have been proposed in recent years (Xu and Tung, 2008, 2009; Kasprzyk et al., 2009; Matrosov et al., 2013; Culley et al., 2016; Beh et al., 2017), one of which is robust optimization. Robust optimization has shifted from expected utility to exploratory bottom-up approaches, which can identify and secure vulnerable scenarios in advance (Giuliani et al., 2014). Managers generally refuse to use the optimization models to directly operate reservoirs, particularly when they consider realistic uncertainties (Celeste and Billib, 2009). They prefer some simpler tools instead, such as rule curves, which are different from the operating solutions informed from the robust optimization analysis (Kasprzyk et al., 2013). Moreover, the former studies generally focused on the overall changes in hydropower production and the resulting economic impacts, whereas the sustainability of the future hydropower systems, including the reliability to maintain base output, and the resiliency and vulnerability to output failure, which can be used to quantify and identify how different reservoir policies response to future changes, is still less investigated.

The Xinanjiang catchment in Eastern China is a good case study given that its power sector is highly dependent on hydropower. The Xinanjiang hydropower plant is the first nationally designed and constructed reservoir in China. Dominated by global warming, rapid urbanization and land use policies in the Xinanjiang catchment, the inflow and hydropower generation of Xinanjiang Reservoir are undergoing dramatic changes. We aim to evaluate the hydropower changes induced by different operating policies, climate and land use changes, using Xinanjiang Reservoir as a case study. The innovations of this study are as follows: (1) The overall changes in the hydropower potential and outputs as well as the sustainability of hydropower projects are assessed under the combined impacts of climate and land use changes; (2) Robust optimization curve is developed to mitigate the impacts attributed to the uncertainty, and a baseline curve is presented for comparisons. Specifically, the projections of bias-corrected GCMs and modeled land uses are employed as inputs of the Soil and Water Assessment Tool (SWAT) to predict streamflow under multiple scenarios. Then the streamflow changes and corresponding information are considered when developing robust reservoir operation rules, and the hydropower generation and sustainability in the future are finally assessed based on the baseline and robust rules.