Creating Markets for Wind Electricity in China: Hybrid Approaches to Energy Modeling and Regulation

China’s rapid economic growth – largely industrial, energy-intensive and reliant on coal – has generated a range of environmental, public health and governance challenges. Renewable energy development, spurred by sustained government intervention, has a large potential to address some of these impacts. However, while China now leads the world in deployment, it faces large performance gaps and inefficiencies in operation. Notably, curtailment (or waste) of wind and solar as a result of grid restrictions is high and increasing, creating uncertainty in China’s ability to meet longer-term environmental objectives including international climate commitments.

Curtailment in China, Texas Figure - Wind curtailment in major wind provinces of China (wind share of generation > 5% in 2016) and Electric Reliability Council of Texas. Source: NEA, NBS, ERCOT

Concurrently, China is in its third decade of a protracted and incomplete transition toward competitive electricity markets. China joins most OECD and over 70 non-OECD countries that have reformed their historically vertically-integrated utilities to incorporate market competition. While a “standard liberalization prescription” exists, differences in institutional histories, resource endowments and broader economic context affect design and execution. Previous work has demonstrated that these diverse pathways make attributing outcomes to particular institutional reforms difficult.

To understand incomplete liberalization efforts in authoritarian China, much literature describing governmental decision-making places strong emphasis on the relationship between the center and localities. While formally subject to the ultimate authority of the central government, local governments have significant autonomy over many aspects of governance, including annual electricity generation and consumption plans. China’s market reforms since the 1980s have thus been analyzed through the lens of entrepreneurial subnational entities experimenting with different modifications of legacy planning institutions. Nevertheless, the electricity sector points to several potential gaps in central-local theories: the grid company as a third party with distinct interests, as well as the technical limitations on electricity system operation and trading.

Establishing causal relationships for electricity system outcomes such as renewable energy integration is difficult, reflecting complex interactions of physical constraints, economic incentives and political conflicts. Supply and demand must be balanced instantaneously and physical restrictions affect production and transport possibilities and requirements. Wind energy, in particular, has high variability, which can lead to operational difficulties with conventional power plant inflexibilities and network conditions. Institutions for planning, operation and regulation of the entire electricity system can also lead to political economies that may strongly support incumbents, inhibit flexible trading of electricity and electricity services, and disincentivize efficient coordination across many related functions. Decision-making for planning and operating the electricity system may not follow de jure specifications, complicating quantitative analysis.

These twin challenges of institutional and technical complexity motivate a multi-method approach combining qualitative interviews and quantitative modeling techniques to assess the impact of institutional design and implementation on meeting overarching system goals of efficiency and wind energy utilization in four case regions of China. I construct narratives of design and management of grid operations through semi-structured interviews of key stakeholders from grid companies, policy and regulatory bodies, and generation companies (conducted 2013-2016). Concurrently, I have developed a highly detailed grid operation model used on a daily basis by most power system operators known as a unit commitment and economic dispatch optimization, which solves for the optimal scheduling of generators on the time frame of hours to days. From the narratives, I identify institutional conflicts and, for a select group, incorporate them into the grid model as additional constraints.

Four cases of diverse local electricity market experiments at the regional or provincial level with significant wind development were chosen to understand the range of political processes surrounding system and market operations: a peaking balancing market in the Northeast; excess wind electricity exchanges in Western Inner Mongolia; bilateral contracts in the Northwest; and hydropower exchanges from Yunnan to Guangdong in Southern Grid. These four are located in three different grid companies, and have diverse geographies and institutional histories. The market experiments range from the very common (bilateral contracts) to the highly experimental (peaking balancing market).

Qualitatively, a picture emerges of a power system in transition. Grid operations are still heavily planned, but with various inputs for out-of-plan transactions. The grid company makes most decisions on short-time scales relevant for wind integration, but is heavily constrained by a constellation of long-term decisions (monthly and longer) made by government agencies. Quantitative results show novel insights into wind integration and market reforms: wind curtailment increases dramatically when provincial barriers to trade resulting from annual planning and inability to share (short-term) reserves are considered. The (long-term) quota on its own increases costs but does not substantially affect wind utilization. Other administratively-set parameters, such as minimum generation outputs, have a strong impact on outcomes and demonstrate the potential for improvements if direct incentives for efficiency were instituted such as through markets.

Methodologically, engaging the model-building process and case interviews concurrently over the course of the study has demonstrated benefits of the multi-method approach: between field visits new features were added to the quantitative model, and model results helped improve depth of question guides in subsequent interviews. I will particularly note this with regard to processes underlying reserve sharing. The integrated lens using the quantitative model to test various counterfactuals is helpful in elucidating the effects of interacting institutions identified in the interviews, such as barriers to trade. Finally, cross-case comparisons and modeling lead to more precise scoping conditions. The optimization approach is potentially extensible to other sectors where technological parameters strongly affect institutional design.

These findings raise serious questions about the ability of current reforms in China’s electricity sector to increase efficiency and achieve renewable energy integration goals: in particular, plans dominated by intra-provincial markets, rather than seeking to capture benefits from inter-provincial trading. Establishing trading arrangements among relatively autonomous jurisdictions is notoriously difficult. China’s provinces in terms of political autonomy share similarities with EU member states, which have only recently made great strides toward integration, and Indian states – indicating potential fruitful future cross-national comparisons of drivers and outcomes in energy and environmental politics. As renewable energy shares increase around the world, this dissertation can help illuminate a range of conflicts to be expected.


J. Ignacio Pérez-Arriaga (Chair), Professor & Director of the BP Chair on Energy & Sustainability, Instituto de Investigación Tecnológica (IIT), Universidad Pontificia Comillas, and Visiting Professor at MIT Sloan School of Management

Valerie J. Karplus, Assistant Professor of Global Economics and Management, MIT Sloan School of Management

Margaret Pearson, Professor of Government and Politics, University of Maryland