Bangzhu Zhu

Examining the links among economic growth, energy consumption, and CO2 emission with linear and nonlinear causality tests

Abstract Understanding the causal relationships among economic growth, energy consumption, and CO2 emission is important for formulating energy conservation and emission reduction policies. In this paper, we explore the causal relationships among economic growth, energy consumption, and CO2 emission in China during 1978–2012 by using both the linear and nonlinear causality tests. The obtained results show that the links examined by the linear and nonlinear causality tests are not all same. Both linear and nonlinear causality tests indicate a unidirectional causality from CO2 emission to GDP and a bi-directional causality between energy consumption and CO2 emission. Linear causality test indicates a unidirectional causality from energy consumption to GDP, while nonlinear causality test indicates a unidirectional causality from GDP to energy consumption. Finally, policy recommendations are proposed for achieving the target of coordinated, sustainable development of China.

Measuring regional energy efficiencies in China: a meta-frontier SBM-Undesirable approach

Under the framework of meta-frontier, we employ the slacks-based measurement (SBM)-Undesirable approach to explore China’s provincial energy efficiencies and meta-technology ratios (MTRs) of eight major economic regions during 2000–2014. The results obtained show that: firstly, the SBM-Undesirable model involving a undesirable output of CO2 emission is more reasonable than the SBM model for measuring China’s provincial energy efficiencies. Secondly, there are severe imbalances of energy efficiencies between regions due to their imbalanced energy technologies. Thirdly, energy efficiencies of the southern, eastern and northern coastal regions are high with advanced energy technologies. Energy technology gaps between regional and meta-technologies of southwest, eastern coastal and northern coastal regions are shrinking; however, the ones of remaining regions are widening. Fourthly, energy technology of overall China has a U-shaped trend; however, the ones of provinces in each region are characterized as a club convergence.

Self-scheduling of a power generating company

The Carbon Tax Self-Scheduling (CTSS) model for a power generating company (GENCO) is proposed in light of the deregulated electricity market environment. The model analyses the effects of GENCO profits and emissions profiles under different carbon tax scenarios, by valuing the specific part of the cost which affects the environment. The resolution method provides first a Mixed Integer Quadratic Programming (MIQP) formulation of the CTSS problem. Second, using piece-wise linearisation approximation methods, the MIQP formulation is transformed into a Mixed Integer Linear Programming (MILP) system. Simulation results of 10-100 unit systems over 24h show that the MILP formulation is efficient and precise when calculating problems of such a large scale. We conclude that the increase of carbon tax reduces carbon emissions and the reduction effect is more favorable in the case of relatively modest carbon tax. The profit of GENCO is unnecessarily negatively related to the carbon tax, while it is determined by the increased rate of electricity price. The increase of carbon tax may inhibit demand. However, the inhibiting effect may be weakened when considering increases in electricity prices combined with the carbon tax. Carbon Tax Self-Scheduling Model (CTSS) is developed.Mixed Integer Quadratic Programming (MIQP) formulation is proposed.Mixed Integer Linear Programming (MILP) formulation is proposed.With/without carbon taxes, both approaches are compared.Simulation results show that the MILP is superior to the MIQP.

Forecasting Carbon Price with Empirical Mode Decomposition and Least Squares Support Vector Regression

This chapter contains another hybrid model of carbon price forecasting that combines empirical mode decomposition and least squares support vector regression. This multiscale prediction methodology yields accurate forecasts of the carbon futures contracts, superior to ARIMA time series models.

Has foreign direct investment increased air pollution in China? A hierarchical linear model approach

Focusing on the mechanism of foreign direct investment on environment, we attempt to build a series of hierarchical linear models to explore the impact of foreign direct investment on China’s sulfur dioxide (SO2) emissions by using the panel data of industrial sector in Chinese provinces from 2002 to 2013. The findings show that: Firstly, the industrial SO2 emission shows a slow downward trend. Secondly, 27.96% of the variations of SO2 emission intensity come from the differences between the provinces. Thirdly, foreign direct investment can explain 50.50% of the different changes in provincial SO2 emission intensity due to economic scale effect, structural effect, technological effect, and environmental regulation effect. Among them, the scale effect and technical effect are negatively correlated with SO2 emissions intensity, while structural and environmental regulation effects positively. Moreover, foreign direct investment can significantly inhibit the positive correlation of structural effect and weaken the negative correlation of technology effect on SO2 emission intensity, but do not have a significant impact on SO2 emission intensity by economic scale effect and environmental regulation effect.

Allocating provincial CO2 quotas for the Chinese national carbon program

In order to improve the efficiency of climate change initiatives China launched its national carbon market in December 2017. Initial CO2 quota allocations are a matter of significant concern. How should we allocate CO emissions reduction responsibilities among Chinese provinces, assuming that provinces will not or cannot trade these responsibilities among themselves? In this paper, we allocate CO quota from the perspective of cost minimisation. First, we estimate the national CO marginal abatement cost (MAC) function and deduce the interprovincial MAC functions. Second, we build an allocation model with nonlinear programming for cost minimisation. Finally, we obtain the allocation results under the emissions reduction target by 2030. The results are as follows. (i) The national MAC was 134.3 Yuan/t (at the constant price of 1978) in 2011, with an overall upward trend from 1990 to 2011. (ii) The interprovincial MACs differ significantly and decline gradually from east to west. Hebei has the largest emissions reduction quota, and Shandong has the largest emissions quota by 2030. (iii) Compared with other criteria of per capita, gross domestic product (GDP), grandfathering and carbon intensity, the proposed approach is the most cost‐effective in achieving the reduction target, with cost savings of 37.7, 34.5, 47.9 and 33.87 per cent, respectively.

An Adaptive Multiscale Ensemble Learning Paradigm for Carbon Price Forecasting

This final chapter is devoted to an adaptive model of carbon price forecasting that makes use of artificial neural networks. Considering either ensemble empirical mode decomposition, the least squares support vector machine, or the particle swarm optimization variant, the competing models are given an extra dimension by incorporating a learning paradigm.

New Perspectives on the Econometrics of Carbon Markets

This chapter provides an accessible introduction to this book. First, we detail the importance of pricing and forecasting carbon market. Second, we review the pricing and forecasting carbon market from the perspectives of carbon price drivers, single scale forecasting and multiscale forecasting. Third, we provide the architecture of this book, and summarize the chapters.

European Carbon Futures Prices Drivers During 2006–2012

This chapter discusses the main driving factors behind carbon prices in detail. It presents key data, then proceeds with the results of cointegration test, Granger causality test, and ridge regression estimation. The chapter provides as well a comparative analysis of the equilibrium carbon price and observed carbon price, before drawing conclusions from the research.

A Multiscale Analysis for Carbon Price with Ensemble Empirical Mode Decomposition

This chapter develops ensemble empirical mode decomposition and fine-to-coarse reconstruction in order to extract carbon price signals from a multiscale viewpoint. The decomposition shows the carbon price is affected by both long-term (e.g., trend) and short-term (e.g., supply–demand fundamentals) imbalances that require appropriate forecasting strategies.