Comparison of time series approaches for prediction of energy consumption focusing on greenhouse gases emission in Iran

Abstract

Purpose The purpose of this paper is to predict the amount of energy consumption by using a suitable statistical method in some sectors and energy carriers, which has shown a significant correlation with greenhouse gas emissions. Design/methodology/approach After studying the correlation between energy consumption rates in different sectors of energy consumption and some energy carriers with greenhouse gas distribution (CO2, SO2, NOX and SPM), the most effective factors on pollution emission will be first identified and then predicted for the next 20 years (2015 to 2004). Furthermore, to determine the appropriate method for forecasting, two approaches titled “trend analysis” and “double exponential smoothing” will be applied on data, collected from 1967 to 2014, and their capabilities in anticipating will be compared to each other contributing MSD, MAD, MAPE indices and also the actual and projected time series comparison. After predicting the energy consumption in the sectors and energy carriers, the growth rate of consumption in the next 20 years is also calculated. Findings Correlation study shows that four energy sectors (industry sector, agriculture, transportation and household-general-commercial) and two energy carriers (electricity and natural gas) have shown remarkable correlation with greenhouse gas emissions. To predict the energy consumption in mentioned sectors and carriers, it is proven that double exponential smoothing method is more capable in predicting. The study shows that among the demand sectors, the industry will account for the highest consumption rate. Electricity will experience the highest rate among the energy careers. In fact, producing this amount of electricity causes emissions of greenhouse gases. Research limitations/implications Access to the data and categorized data was one of the main limitations. Practical implications By identifying the sectors and energy carriers that have the highest consumption growth rate in the next 20 years, it can be said that greenhouse gas emissions, which show remarkable correlation with these sectors and carriers, will also increase dramatically. So, their stricter control seems to be necessary. On the other hand, to control a particular greenhouse gas, it is possible to focus on the amount of energy consumed in the sectors and carriers that have a significant correlation with this pollutant. These results will lead to more targeted policies to reduce greenhouse gas emissions. Social implications The tendency of communities toward industrialization along with population growth will doubtlessly lead to more consumption of fossil fuels. An immediate aftermath of burning fuels is greenhouse gas emission resulting in destructive effects on the environment and ecosystems. Identifying the factors affecting the pollutants resulted from consumption of fossil fuels is significant in controlling the emissions. Originality/value Such analyses help policymakers make more informed and targeted decisions to reduce greenhouse gas emissions and make safer and more appropriate policies and investment.

Evidence of Causality between Economic Growth and Electricity Consumption Expenditure in Uganda

“…Various Prediction approaches utilized by researchers in studying the nexus between electricity consumption expenditure and economic growth may be structured into two categories: causal relationship models and univariate models [1]. Doroodi et al [1], argued that, causal relationship models predict dependent variables based on one or several variables and show that there is a causal relationship between the dependent and independent variables. However, univariate models validate that a system is a function of its behaviour [13].…”

“…The importance of electricity consumption in the economic and social development through improved quality of life cannot be overemphasized [1]. Electricity as a form of energy is an essential driving force of economic growth in all economies, which directly and indirectly complement labour, capital and land as factors of production [2].…”

Application of a hybrid model based on multiple linear regression -principle component analysis (MLR-PCA) for electricity export forecasting

“…Power consumption modeling is a topic of interest for many researchers. The related literature can be divided into two parts: Univariate modeling and Multivariate modeling [4], [5]. Univariate modeling is based on time-series data such as Grey models(GM) [6,7] and AutoRegressive Integrated Moving Average (ARIMA) [8].…”