Introduction

The world is facing serious energy efficiency and energy security issues. Continuously rising temperatures globally have been reported, thus necessitating efforts to improve energy security and environmental sustainability to move the world towards greater energy independence. In light of the extreme energy and environmental issues, a low-carbon economy is urgently needed in many countries (Anser et al. 2020i; Anser et al. 2020g; Anser et al. 2020b; Anser et al. 2020j; Anser et al. 2020a). Low-carbon solutions involve technological, structural, and governance approaches. For today’s global economy, energy and the environment are one of the most critical concerns. One of the fundamental reasons for the widespread depletion of the environment has been the worldwide production and utilization of oil (Tiep et al. 2021; Singh et al. 2018). Many countries have found ambitious energy usage and diversification planning to improve energy production to facilitate industrial economic development while reducing energy utilization (Zhang et al. 2018; Ahmad et al. 2020a; Petrović-Ranđelović et al. 2020).

It is believed that improving energy efficiency is a meaningful way to reduce energy consumption and minimize environmental emissions (Anser et al. 2020f; Anser et al. 2020e; Anser et al. 2020h; Anser et al. 2020c; Anser et al. 2020d). Consequently, the corresponding technical advancements in energy production and utilization must be encouraged (Asif et al. 2020; Sarker et al. 2020; Iram et al. 2020; Tehreem et al. 2020). The development of a fair and detailed climate policy framework for increasing energy efficiency is also important (Ouariachi et al. 2019). Many researchers and policymakers are consistently trying to improve the energy efficiency level with different optimal methodologies (Liu et al. 2020; Lin et al. 2020; Jun et al. 2020; Geng et al (2019). According to the International Energy Agency (IEA), the study explains that energy efficiency is aimed at controlling global energy consumption Sağlam (2017). The most efficient use of energy is to generate more sustainable development (Wasif Rasheed and Anser 2017; Xu et al. 2020; Ahmad et al. 2020b). The use of fossil fuel resources has been declining for the past few decades. Therefore, policymakers and researchers continuously look for novel sources for upgrading energy efficiency. The primary focus of previous studies was on the importance of clean and cheap electricity for sustainable development and the effective use of energy in specific areas to assess energy efficiency (Yousaf et al. 2020; Tehreem et al. 2020; Wasif Rasheed and Anser 2017; Xu et al. 2020).

The super-efficient model has become popular in recent years for assessing efficient and inefficient decision-making units (DMUs) because the super-efficient model measures the performance in both beneficial and harmful situations. The advantage of this new study is that it uses two data envelopment analysis (DEA) methods for measuring the energy efficiency of the Asia-Pacific regions (Mohsin et al. 2020b; Mohsin et al. 2018; Mohsin et al. 2021). However, DEA models have favorable theoretical properties, and so the environmental efficiency performance scores of many comparable entities are 1. Hence, it becomes challenging to make a meaningful comparison. According to the DEA, if a DMU obtains an efficiency score of 1, it is fully efficient and below 1 it is not efficient. Therefore, it is meaningful to include output shortfalls and input excesses in the DEA-based models while measuring energy efficiency performance (Yang et al. 2021; He et al. 2020; Mohsin et al. 2020a, b). Few studies have conducted energy efficiency analysis using a two-stage DEA model, particularly in Asia-Pacific countries.

Consequently, the purpose of this study is to fill the literature gap and to use advanced DEA methodology to calculate the optimum energy performance of the Asia-Pacific regions, which will be helpful for future sustainable development and policymaking (Hosseinzadeh Lotfi et al. 2011). Various research on energy performance, such as economic growth and environmental quality, have been performed; for example Bampatsou and Halkos (2019) explored the energy efficiency of energy-generating equipment, and Blum (2015) studied environmental efficiency in relation to economic growth. Song et al. (2018) assessed energy and CO2 emission efficiency, and Iftikhar et al. (2018) analyzed both economic and environmental efficiency (Mohsin et al. 2019; Mohsin et al. 2020a; Mohsin et al. 2021).

Even though numerous factors affect energy conservation implementation, the economic benefits are profound, according to the studies described above. Economic incentive policies, among other policies, were found to play a significant role in encouraging the adoption of energy-efficient goods, mostly when the goals and subsidy levels were acceptable (Sun et al. 2020b). As a consequence, GDP was introduced as an output element. The bulk of energy efficiency study is done at the national and regional levels, with no literature on Asia-Pacific economic energy efficiency (Sun et al. 2020c; Sun et al. 2020a; Sun et al. 2020b). On the other hand, the standard DEA model avoids input-output, resulting in errors in estimation performance. The super-performance DEA model, which essentially addresses the deviation issue created by the standard DEA model’s exclusion of the slack component, is commonly utilized in energy efficiency measurement. The Asia-Pacific energy market did not reduce energy usage during this period. More significant measures are needed to fulfill the European Commission’s energy efficiency targets. Consequently, a more in-depth analysis of the Latvian industrial sector is expected to recognize the potential for improved energy production and have specific recommendations for national policymakers.

Our contribution seeks to examine the energy efficiency mechanisms of the Asia-Pacific region by using DEA. Consequently, our primary aim is to determine the driving mechanism of maximization of energy efficiency without reducing economic growth and progress. The study uses DEA to assess the degree of interaction between energy use, environmental pollution, and economic development for this reason. Our paper also seeks to establish a series of indicators that present multiple facets of energy efficiency comparably and systematically. The aim is to go beyond reducing energy waste by raising energy production per unit of energy input to improve energy security. Unlike others, we measure energy efficiency using capital, labor, and energy consumption as input variables. GDP is utilized as a desirable output, while CO2 emission is an undesirable result.

Consequently, a data envelopment method incorporating computer simulation is created to resolve energy efficiency and energy security and the study’s target. The optimized approach will minimize the number of slacks (visualization of resources) through a simulation framework to determine the most effective point of energy efficiency. DEA is used to rate the applicable scenarios within the current range and find the best retrofit approach for enhancing energy efficiency.

The remainder of the paper is formulated as follows: The “Literature review” section defines the energy status of Asia-Pacific countries. The “Data and methodology” section presents the background of the selected model. The “Results and discussion” section elaborates on the results and discussion. The final section sums up the conclusion and policy suggestions.

Literature review

Energy initiatives and policies in Asia are needed to meet the world’s infinite demand for energy, ensure energy stability, and reduce the impact of energy on the environment. Massive power plants are required. Hydroelectric power and coal electricity are no longer capable of serving developing nations’ needs in energy generation. These countries are participating in the quest for alternative renewable energy sources. We note that collecting raw materials and underground facilities could be managed under a single roof if an agreement were reached between Asian countries (Alemzero et al. 2020b; Sun et al. 2020a; Alemzero et al. 2020a; Ríos and Olaya 2018; Iqbal et al. 2019; Al Asbahi et al. 2019; Sun et al. 2019); Chander (2017); Kamran (2018). Cooper’s model is based on Farrell’s method of mathematical programming, also called linear programming. Banker and Chang et al. (2006) later developed a DEA model with a variable return-to-scale (VRS) mechanism. Different DEA models have various advantages Ouenniche and Tone (2017). The SMB model with the ability to measure super-efficiency provides the best way to deal with undesirable outputs of CO2 emission. It can also compare the efficient DMUs, including country, city, and organization. CCR and BCC SEA models are two different DEA models in which efficient DMUs are given an efficiency score of 0 and an inefficient score below 1.

BCC and CCR models also recognize when DMUs are efficient and inefficient. The various techniques were developed as part of the DEA method. The super-efficiency process has both beneficial and harmful situations. It remains to be seen whether the super-efficiency model works better than both inefficient and effective DMUs. The super-efficiency strategy can also provide an essential method for decision-makers to explain the actual position of different forms of efficient DMUs.

The same level of inputs maximizes the outputs in the CCR-DEA model; however, as a by-product of valuable outputs, the manufacturing mechanism generates undesirable outputs, such as CO2 pollution and contaminants. Environmental sustainability aims to create more good outcomes and fewer wrong outputs by utilizing the most periodic inputs (i.e., natural resources). In such instances, standard DEA models struggle to measure environmental efficiency because wrong outputs necessitate complicated dealings to achieve a more detailed estimate. Several DEA models have measured environmental performance by using low outputs in the conventional DEA system. This input-oriented model outperforms the DEA super-efficiency and SMB models, among others.

This imperative is compounded by the unequal allocation of resources in the Asia-Pacific region and the complexities of the global energy markets. This is expressed in the Committee on Energy’s first-session report (E / ESCAP/73/30) and shows the relationship among the increasing demand for energy, dependency on fossil fuels, low energy usage, restricted use of renewable energy, lack of access to sustainable and clean energy, and the need to move to a low-carbon and environmentally sustainable energy future. While countries in the region accept the need for the transformation to a renewable energy infrastructure, the variety of circumstances they face suggests no concrete path or timetable for achieving this (Hanif et al. 2019). With large differences in geography and technological skills, Asia-Pacific countries have varying levels of development and uneven resources. There are several policy choices to calibrate each nation’s response to driving progress towards a sustainable energy future. Many countries are well on the path to a sustainable future for energy, while some have just started. Regional collaboration for the utilization of capacity, expertise, technologies, investment, and local markets development is an integral component of the transition to renewable energy. However, the Asia-Pacific region needs to use available energy resources efficiently.

The ESCAP Member States invited the Secretariat to draw up this “Regional Renewable Energy Cooperation in Asia and the Pacific” theme study for 2017 in response to these opportunities and to add to the knowledge base and catalyze action. This publication takes note of the challenges and successes made in the region-wide transition to renewable energy. It outlines some of the strategies to be discussed by policymakers in the region to balance the economic, social, and environmental facets of life and provides a series of recommendations for implementing a regional partnership mechanism for renewable energy. The Asia-Pacific region’s transition to renewable energy has also begun. Investment in renewables has overtaken investments in fossil energy region-wide. In several nations, more sophisticated energy efficiency plays a part in decoupling economic development from energy consumption. Cross-border communication connections for electricity are being built, and several others are proposed. Innovations of energy are proceeding quickly and are unleashing opportunities to challenge existing models of energy supply. At the household, district, and industrial levels, renewable energies, such as solar, generate low-cost energy. To change the power generation and transport markets, storage systems, smart grids, and hybrid cars are poised. Innovative strategies, system thinking, alternative financing sources, and advanced business models must respond to this new reality and leverage these developments to help the energy transition.

Energy security of Asia-Pacific countries

According to the IEA, the overall demand for power generation increased by 4.9% from 600 million kilotonnes of oil equivalent (ktoe) in Asia-Pacific countries to 18.7138 million ktoe. Asia-Pacific countries have achieved this significant increase. Global primary energy demand in the Pacific increased from 30.6% in 2000 to 39.1% in 2016, with the region comprising 71.8% of worldwide increases in primary energy demand (Zhang and Song 2021; Ma et al. 2019; Guo and Yuan 2020). All South Asian countries are projected to develop oil consumption by 6%. India has oil reserves of 187,138 million barrels (Table 1) of the most significant crude oil in Southeast Asia. Sri Lanka and Pakistan, in contrast, have 34,568 million and 4608 million, respectively. Unless India continues to consume its crude oil at the current pace, stocks will be depleted for 35 years (Yarovaya et al. 2021). As far as natural gas is concerned, India, with a combined capacity of 390,944 million cubic meters, followed by 116475 million cubic meters and 48,378 million cubic meters, respectively, for Japan and Indonesia. India, Pakistan, and Bangladesh only produce natural gas with a proportion of production reserves of 34, 20, and 17. For biomass, India has the highest share of 7479 million tonnes.

Table 1 The total energy supply of South Asian countries
Full size table

Table 1 shows that the per capita energy consumption in Pakistan and Sri Lanka ranges from 119.7 kWh in Bangladesh to more than 635 kWh. The overall regional electrification rate is 73%, meaning that about 418 million people live without sufficient electricity. This figure reflects over one-third of the 1.3 billion people worldwide without access to electricity. Therefore, the average non-solid fuel access in the area is 38%.

The electricity supply in South Asian countries (Fig. 1) shows that most countries depend on a single source for more than 50% of their power. The range of options to satisfy different energy needs is constrained by a single source, which raises energy security issues. In addition, despite large reserves of indigenous coal, Japan imported 28 million tons of coal in 2015, primarily due to low-level domestic coal and technical limits.

Fig. 1
figure 1

The total energy supply of Asia-Pacific countries

Full size image

Data and methodology

In this section, we discuss the proposed methodology. In this study, we use two DEA models, as suggested by Andersen and Petersen. One super slack-based measure model was created to examine the relative efficiency of Asia-Pacific countries using the constant returns scale (CRS). There are vital steps in place for different input and output DMUs. Using DEA model techniques, the principal objective is solved very easily for multi-complex problems, and multiple inputs and outputs also address unfamiliar issues. Cooper and his team proposed a DEA in 1979, called the Cooper–Carlos–Redous (CCR) method, using CRS techniques to calculate the technical efficiency (TE). Banker and his team subsequently introduced a new model called the BCC model. Efficiency results are found through the CCR model, and those DMUs have less input of the highest output levels evaluated efficiently. Unlike the CCR method, VRS techniques are used to measure efficiency (Conlon and McGee 2020). Furthermore, both models are further divided into two types: the first type is input-oriented, which is the main aim to minimize the inputs but simultaneously increase the output level, and the second is output-oriented, with a primary focus on increasing the output level without decreasing the input level.

The CCR model of DEA was introduced by Charnes et al. (1978) to measure decision package size and the technology’s overall efficacy. The CCR model assumes that there are no decision packages and has m types of input indicators and t types of output indicators, resulting in the following performance measurement CCR model (Chege and Wang 2020; Blanchard 2019). In the DEA-CCR model, a DMU score of 1 indicates efficient, and below 1 indicates inefficient. There is no other logic tool to differentiate between efficient and inefficient DMUs (Emrouznejad and Liang 2018; Matsuyama 2019), their position or numerical rating of successful DMUs without influencing the rating non-efficiency DMUs, which is a drawback of the CCR-DEA model. A modern slacks-based measure of super-efficiency in DEA incorporates the super-efficiency model created by Liu et al. (2013). The key benefit of this model is the relative value of each productive DMU. Having a high DMU score indicates DMU super-efficiency, which is assigned a position or a rating of all the original efficient DMUs. Anderson and coworkers developed a new super-efficiency model based on a radial technique called the CCR-I input-oriented model to rank all efficient DMUs. This model data removes the DMUs to evaluate the solution set, an efficiency score, a super-efficiency model. The CCR-I model results in super-efficiency DMUS, and these values and consequences apply to rank the DMUs. To maximize radial environmental and energy efficiency, non-radial steps must be taken. Here are the model’s optimal solutions: When DMU = 1, DMU applies a restriction to the CCR model with DEA performance. When the DEA data of a framework enables is efficient, both system and measure are efficient; when the DEA data of a decision package is efficient, both the system and scale will be efficient. Since the data envelope review can produce multiple effective decision packages that cannot be further correlated and synthesized, Fleishman and Anderson (1980) suggested the super-efficiency data model, which can evaluate the efficiency of different decision elements.

$$ {\displaystyle \begin{array}{c}E\_1=\mathit{\min}\ \theta \\ {}s.t\end{array}} $$
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\ \uplambda \_\mathrm{j}\kern0.5em \mathrm{X}\_\mathrm{i}\mathrm{j}+\mathrm{S}\_\mathrm{i}\hat{\mkern6mu} \left(\mathrm{x}-\right)=\mathrm{X}\_\mathrm{i}\mathrm{j}0,\kern1.5em \mathrm{i}=1,\dots, \mathrm{m}, $$
(1)
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\uplambda \_\mathrm{j}\kern0.5em \mathrm{e}\_\mathrm{l}\mathrm{j}+\mathrm{S}\_\mathrm{l}\hat{\mkern6mu} \left(\mathrm{e}-\right)=\uptheta\ \mathrm{e}\_\mathrm{l}\mathrm{j}0,\kern1.5em \mathrm{l}=1,\dots, \mathrm{L}, $$
(2)
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\uplambda \_\mathrm{j}\kern0.5em \mathrm{y}\_\mathrm{r}\mathrm{j}+\mathrm{S}\_\mathrm{r}\hat{\mkern6mu} \left(\mathrm{y}+\right)=\mathrm{y}\_\mathrm{r}\mathrm{j}0,\kern1.5em \mathrm{r}=1,\dots, \mathrm{S}, $$
(3)
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\ \uplambda \_\mathrm{j}\kern0.5em \mathrm{b}\_\mathrm{kj}=\uptheta\ \mathrm{b}\_\mathrm{kj}0,\kern1.5em \mathrm{k}=1,\dots, \mathrm{K}, $$
(4)
$$ \uplambda \_\mathrm{j},\mathrm{S}\_\mathrm{i}\hat{\mkern6mu} \left(\mathrm{x}-\right),\mathrm{S}\_\mathrm{l}\hat{\mkern6mu} \left(\mathrm{e}-\right),\mathrm{S}\_\mathrm{r}\hat{\mkern6mu} \left(\mathrm{y}+\right)\ge 0,\kern1.25em \mathrm{for}\ \mathrm{all}\ \mathrm{j},\mathrm{I},\mathrm{l}, $$
(5)

The CCR-DEA model is used for maximizing inputs through outputs. However, at the same time, the production process produces the worst outputs, including CO2 emissions and pollutants, as spinoffs of valuable results (Fig. 2). With minimal inputs, environmental sustainability aims to produce more good outputs and fewer negative outputs. In such situations, conventional DEA models do not calculate environmental performance because undesirable results require complex dealings to obtain a more detailed performance calculation. To date, numerous DEA models have been used to assess environmental efficiency by incorporating undesirable results into the traditional DEA framework. Among other models, this input-oriented model is the best in DEA super-efficiency and SMB models.

$$ {\displaystyle \begin{array}{c}\mathrm{E}\_2=\min 1/2\kern0.5em \left(1/\mathrm{L}\sum \_\left(\mathrm{l}=1\right)\hat{\mkern6mu} \mathrm{L}\uptheta \_\mathrm{l}\hat{\mkern6mu} \mathrm{e}\kern0.5em +1/\mathrm{K}\sum \_\left(\mathrm{k}=1\right)\hat{\mkern6mu} \mathrm{K}\uptheta \_\mathrm{k}\hat{\mkern6mu} \mathrm{b}\ \right)\\ {}\mathrm{s}.\mathrm{t}\end{array}} $$
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\uplambda \_\mathrm{j}\kern0.5em \mathrm{X}\_\mathrm{i}\mathrm{j}+\mathrm{S}\_\mathrm{i}\hat{\mkern6mu} \left(\mathrm{x}-\right)=\mathrm{X}\_\mathrm{i}\mathrm{j}0,\kern0.5em \mathrm{i}=1,\dots, \mathrm{m} $$
(6)
$$ \sum \_\left(j=1\right)\hat{\mkern6mu} n\lambda \_j\kern0.5em e\_ lj+S\_l\hat{\mkern6mu} \left(e-\right)=\theta \_l\hat{\mkern6mu} e\ e\_ lj0,\kern1em l=1\dots \dots L, $$
(7)
$$ \sum \_\left(j=1\right)\hat{\mkern6mu} n\lambda \_j\kern0.5em y\_ rj-S\_r\hat{\mkern6mu} \left(y+\right)=y\_ rj0,\kern1.5em r=1,\dots, S $$
(8)
$$ \sum \_\left(\mathrm{j}=1\right)\hat{\mkern6mu} \mathrm{n}\uplambda \_\mathrm{j}\kern0.5em \mathrm{b}\_\mathrm{k}\mathrm{j}=\uptheta \_\mathrm{k}\hat{\mkern6mu} \mathrm{b}\ \mathrm{b}\_\mathrm{k}\mathrm{j}0,\kern1.5em \mathrm{k}=1\dots ..\dots \mathrm{K}, $$
(9)
$$ \uplambda \_\mathrm{j},\mathrm{S}\_\mathrm{i}\hat{\mkern6mu} \left(\mathrm{x}-\right),\mathrm{S}\_\mathrm{l}\hat{\mkern6mu} \left(\mathrm{e}-\right),\mathrm{S}\_\mathrm{r}\hat{\mkern6mu} \left(\mathrm{y}+\right)\ge 0,\kern1.25em \mathrm{for}\ \mathrm{all}\ \mathrm{j},\mathrm{I},\mathrm{l}, $$
(10)
Fig. 2
figure 2

Energy efficiency score

Full size image

The non-parametric DEA method efficiently provides accurate effective new DMUs and achieves the best solution on the construction frontier. There are many inputs converted into outputs for DMUs. The DEA model can make intelligent the pure or overall technical, scale, and super-efficiency. There are two DEA models, the CCR model and the BCC DEA model; both are helpful benchmarking methods and handle multiple inputs and outputs. Additionally, Cooper et al. suggested selecting the best DMUs. Firstly, all DMUs use the inputs and outputs to allocate accurate data, and secondly, to evaluate the relative efficiency of all DMUs. In standard, the desired inputs and outputs are different measurement units, and the efficiency scores smaller input amounts and more significant amounts.

This study uses three inputs and one output: capital, labor force, and CO2 emission. The result has only GDP (gross domestic product) with the input-oriented approach. In addition, measuring CO2 emission with solid fuels and natural gas products and measuring GDP adds to a country's total gross capital. In this study, we use the input model analysis with the super-efficiency CRS-I model and SBM CRS method to assess energy efficiency and CO2 emission using the DEA input-oriented method for Asia-Pacific countries in 2015-2018.

Meanwhile, the input model has infeasible results. This study did not recognize undesirable models, CO2 emissions and select the input and output with the related energy efficiency analysis. Here we discuss the energy efficiency inputs and outputs of old studies. In this view, Suzuki Motors indicated that different inputs such as energy, primary energy consumption, non-energy consumption energy, and input such as population and output such as CO2 and GDP are used to calculate the efficiency of Europe, APEC, and Asian countries. George et al. used labor and capital as input and GDP as output.

Results and discussion

Descriptive statistics

In this research work, we collected data for 15 Asia-Pacific regions for the period 2015–2018, using the World Bank and IEA as data sources. Unfortunately, data for some countries are missing, as data were not available for those countries. Four variables were used in the study: three input variables, i.e. labor force, capital, and CO2 emissions, and one desirable output, gross domestic product (GDP). Table 1 highlights the input and output indicators.

Table 2 shows the descriptive statistics analysis of all four input and output variables. Table 3 shows the average input variable of the labor force reported as 119,956,207 thousand from 2015 to 2018. China (CHN) has a maximum labor force in 2017 reported as 787,399,317 thousand, and New Zealand (NZL) has a minimum labor force reported as 2513,18 thousand. The capital average is about 549,629,222,900 (constant 2010 US$) persons from 2015 to 2018. The maximum capital country is China (CHN), with about 4,794,552,264,789 (constant 2010 US$) persons in the year 2017, while the minimum capital average reported for Sri Lanka (LKA) of about 26,590,573,979 in 2015. CO2 emissions input increased between 2015 to 2018 on average at 1009 K tons of carbon. The highest CO2 emission input is reported in China (CHN) at 94,29 K tons of carbon in 2018, while the lowest CO2 emission reported about 18 K tons of carbon from Sri Lanka in the year 2015. To calculate the GDP average growth from 2015 to 2108 at current US dollars was reported about 18,671 billion. The maximum GDP was recorded in Singapore (SGP) at about 64,582 billion current US dollars in 2018, while Bangladesh (BGD) has minimum GDP average growth recorded about 1248 billion current US dollars in 2105.

Table 2 Input and output variables list
Full size table
Table 3 The input and output descriptive statistics results
Full size table

Energy efficiency analysis

Generally, the DMU has an efficiency score that varies between 0 and 1. This paper uses the DEA model to overcome the following issues to solutions. We studied energy security and environmental sustainability and propose pathways to increase energy supply and mitigate climate change. There is no problem with a density between two or more DMU scores if it is 1 or more than 1, although Andersen, Petersen, and Tone established a super-efficiency model and SBM model used for efficient DMUs. DEA basic models do not correctly assess the DMU’s efficiency score, but when a score >1 for each DMU is assigned, they can compare it with other DMUs. We used a super-efficiency score to make the liner programming formula given in Europe and the super SBM score used for the linear programming formula, which is shown in the equation. Table 4 represents the ranking of countries proposed in this study based on model results. The results of both models indicated that Singapore, China, New Zealand, Pakistan, Philippines, Japan, India, Indonesia, and Bangladesh are efficient DMUs during all the periods in both models. It was also found that variations in DMU efficiency scores result in rankings. In 2018, Hong Kong showed efficiency during that year, but results indicate inefficiency to the active region in succeeding years. Results further showed that Thailand and Vietnam are inefficient in 2015 and 2016 because they gained both DMUs and got the highest efficiency score, and they also improved their ranking. Pakistan was found to have an efficiency score and changed in ranking accordingly. In this background, both the optimal model measurement of DMUs and the relation with the super SBM model are more detailed and efficient for the rating and efficiency score. As discussed already, only successful DMUs achieve these rankings. There has been no improvement in results scores less than 1 (Ali and Erenstein 2017; Aghion et al. 2008; Abdullahi 2019).

Table 4 DEA model results in efficiency scores and ranking for the study countries
Full size table

Table 5 presents the efficient and inefficient DMU results of both models. During all studied years from 2015 to 2018, countries with efficiency scores of 1 or above were Bangladesh, Pakistan, China, Singapore, New Zealand, Philippines, Japan, India, and Indonesia, while Malaysia, Thailand, Vietnam, and Hong Kong had efficiency scores of 1 or above in a few years. However, Australia and Sri Lanka obtained a low average in all study years.

Table 5 Efficient and inefficient sets of DMUs
Full size table

In this study, we selected 15 Asia-Pacific countries and further categorized them into four regions, namely: East Asia (China, Japan, Hong Kong), Southeast Asia (Thailand, Philippines, Vietnam, Malaysia, Singapore, Indonesia), South Asia (Pakistan, India, Bangladesh, Sri Lanka) Australasia (Australia, New Zealand). In Table 6, in both super and SMB models, regions that obtained low-efficiency scores included Australasia. East Asia and Southwest Asia received increasing efficiency scores between 2015 and 2018 (Figs. 2 and 3).

Table 6 DMU efficiency average results based on DEA models
Full size table
Fig. 3
figure 3

Super SBM average score

Full size image

Figure 4 shows the super-efficiency score. In 2050, the United States, China, Canada, Japan, South Korea, and Russia are expected to lead building energy production, accounting for 80.04% of the overall demand of Asia-Pacific economies. Building demand in Asia-Pacific economies will grow from 1387.4 Mtoe in 2016 to 2456.8 Mtoe in 2050 over the projected era, owing mainly to increasing energy demand in emerging economies. Building demand could be held below 1770.46 Mtoe in 2035 to economic growth by 45% by 2035 relative to 2005, and then increase to no more than 2000 Mtoe in 2050. In this case, the construction sector’s energy consumption by 2050 would be 1922.6 Mtoe, a 25% decrease from the existing energy security model. Sustainable resources accounted for just 22 Mtoe and 1.69% of overall building energy production in 2010. Thus, this proportion would barely change, while construction of zero-energy buildings will generate 11%, 27%, and 54% onsite by 2050, respectively.

Fig. 4
figure 4

Super-efficiency average score

Full size image

Robustness analysis

Table 7 shows sensitivity analysis of the influence of the efficiency score of each DMU with different DEA model efficiency results. We developed seven more cases to examine an impact on efficient and inefficient DMUs. We also found south Asia’s efficiency scores (Sri Lanka) and Australasia (Australia). Fewer efficiency scores recorded for East Asia (Hong Kong) and Southeast Asia (Malaysia, Vietnam, Thailand), while East Asia countries (China, Japan), South Asia countries (Pakistan, India), South Asia (Singapore, Indonesia, Philippines) and Australasia (New Zealand) were found to be efficient for the whole period.

Table 7 Different model results
Full size table

This study measures the fundamental concepts of the current DEA models. It examines the validity of this quantitative method quantitatively to the energy output of 15 countries in the Asia-Pacific region, further divided into four areas. Performance ratings for the DMUs are taken into account in the present research based on input-oriented super-efficiency analysis and input-oriented super SBM approaches. It also evaluated the ranking and efficiency score of all DMUs based on the available data. There are six vital points to shed light on obtained results and three recommendations about this context. For the years 2016 to 2020, China has set 10 million m2 of nearly zero-energy buildings. By 2025, all newly built buildings in Korea will be zero-energy buildings, according to the 2nd Energy Master Plan. In comparison to government priorities, pioneering non-governmental organizations (NGOs) have formulated mid-to-long-term objectives that are much more aggressive than government objectives. In the excess energy efficiency case, where building energy code standards are not upgraded, Asia-Pacific energy demand would begin to grow before it exceeds 2456.8 Mtoe in 2050, with no plateau before then.

Overcome energy use, supplement with higher-quality dryers, and put a premium on seeking a gas consumption alternative that is affordable and environmentally sustainable. Renewable technology has already played a vital part in low-carbon energy systems, both now and for the future. Many countries are transitioning to clean energy sources, including solar, wind, and hydro, to reduce their natural gas use (Alkire and Foster 2011). Consequently, governments have announced strong policies to tackle electricity scarcity to provide a safe atmosphere for their people. Solar PV, wind, and biogas are examples of non-fossil fuel energy sources that can offset fossil fuel energy. Many countries also use conservation energy in drying machines of milling factories, which has negative economic and environmental consequences; thus, energy-efficient tools need the rice milling factories (Afonso and Furceri 2010; Iqbal et al. 2020). This thesis proposes the model based on several previous prospect tests.

Energy security and energy efficiency

Energy imports control Asia-Pacific economies, and energy consumption continues to grow, accounting for roughly 60% of global energy demand. Asia-Pacific economy members decided in 2011 to reduce average energy intensity by 45% by 2035 relative to 2005 levels and rationalize and phase out unsustainable fossil fuel subsidies that promote excessive use while still delivering critical energy services. The Paris Agreement was drafted and accepted by consensus at the 2015 United Nations Climate Summit. The key aim is limiting the world surface temperature growth this century to below 2 °C and restricting global temperature rise to 1.5 °C over preindustrial levels. The planet must meet the building sector’s greenhouse gas (GHG) emissions reduction capacity to limit global warming. According to reports, we need to reduce CO2 emissions by 50–65% from today’s levels by 2030 and entirely phase out fossil fuel CO2 emissions by 2040 to have a 67% probability of reaching the 1.5 °C goals.

Figure 5 shows natural gas consumption (billion cubic meters). This research explores the fundamental principles and models of the super DEA methodology available. Energy security and environmental sustainability and propose pathways to increase energy supply and mitigate climate change. It empirically evaluates the applicability of this method of analysis to the energy efficiencies of 15 Asia-Pacific countries.

Fig. 5
figure 5

Natural gas consumption

Full size image

Figure 6 shows primary energy consumption (billion cubic meters). First of all, traditional efficiency analyses do not thoroughly equate to efficient DMUs, because in both super performance and super SBM analyses, we have identified multiple successful DMUs. We observed in both DEA models that no changes are reported in inefficient countries during all these years. However, we also realized that the DEA models provided precise results in DMU rankings and ratings. For example, Hong Kong was ranked as 5, 3, 3, 3, and 1 in the super-efficiency model for the years under examination.

Fig. 6
figure 6

Primary energy consumption

Full size image

Figure 7 shows the unit price of natural gas imports (USD/cubic meters). Secondly, this is to evaluate and rank the fossil model based on 2015–2018 results. Bangladesh, Pakistan, China, and Singapore are all productive and above average countries, and Bangladesh, Pakistan, China, Singapore, New Zealand, Philippines, Japan, India, and Indonesia are below average. Both DEA models were performed for all the years. Thermoset found that the average efficiency scores for 10 regions of Asia-Pacific were higher than the scores for five regions of Asia-Pacific.

Fig. 7
figure 7

Unit price of natural gas

Full size image

On the other hand, in the super SBM model, Bangladesh, Pakistan, and China have been included in fewer comparison sets than Hong Kong compared to the super-efficiency model. Lastly, relative to the remaining efficient nations, Bangladesh, Pakistan, which were lower in terms of efficiency score and rating, inefficient countries took Hong Kong more often as the reference country. This result may be because Hong Kong is far closer to the efficiency mark than other efficient countries.

Conclusion and policy implications

In the first point, we compare the efficient DMUs using advanced efficiency analysis methodologies. Policies are needed that strengthen cross-border energy trade and increase trans-regional renewable energy investments for long-term energy security and environmental sustainability in the region. Meanwhile, we observed during the investigation that different DMUs are efficient in the results of both model during all study years, although we have to understand that the super-efficiency model and SMB have additional efficiency and ranking score. For instance, using the super-efficiency model during all years from 2015 to 2018, China recorded ranking scores 15, 15, 15, and 15, whereas the same country using SMB model results recorded different ranking scores as 15, 14, 14, and 14. Thus, the outcome depends on the distribution of scales, which is developing a super model compared to the SMB model, giving high efficient and accurate results.

In point two, after examining the DEA model ranking based on data from 2015 to 2018, we observe that New Zealand, Singapore, China, Pakistan, Philippines, Bangladesh, India, Japan, and Indonesia are efficient for above the average. Still, Sri Lanka, Malaysia, Vietnam, Thailand, Australia, and Hong Kong are under the average DEA models, and all studied. In point three, to assess the countries’ efficiency scores and rank them, in both models for all the years, results show that Australasia and Southeast Asia have below-average efficiency scores. At the same time, East Asia and South Asia have total average efficiency near 1. These results show that South Asia and East Asia achieve their energy efficiency target, but Australia and Southeast Asia need to apply multiple strategies to reach their target result.

In comparison, there was a decrease in this ratio in the super SBM model. In the super SBM model, Singapore and New Zealand are taken more often than the super-efficiency model as a place point. In that regard, it is reasonable to assert that East Asia took a more frequent role in the super SBM model than South Asia, Australia, and Southeast Asia in all DEA models and all the years reviewed. In point five, we assumed that Asia-Pacific countries would become more efficient when they reduce the inputs. Asia-Pacific countries are worried about controlling these variables since GDP and GCF rely on multiple inputs and outputs in the economy. Paying for workers will create a decline in society’s well-being and proceed to instability, and there would be no improvement in this feedback. As a result, Asia-Pacific countries suggest a practical and reasonable estimation of how to achieve a significant boundary under the defined targets of decreasing fossil fuels (solid fuels, natural gas, or total petroleum products). Observing the type of fuel used will be limited. Next, initiatives aimed at growing energy conservation can emphasize helping a single province within a country, enhancing the total factor energy quality as a consequence of the province’s policy assistance. The spillover impact of the entire factor energy performance of local region and also boost.

Second, lower energy quality means eliminating backward manufacturing potential and pursuing more renewable growth routes. There is a significant geographic disparity in Asia-Pacific energy efficiency. The strategy should concentrate on directing funds away from overcapacity sectors like steel, nonferrous metals, chemicals, construction materials, and thermal power in low productivity areas.

Simultaneously, promoting the growth of emerging industries such as green environmental protection, high-tech, modern service industries, green agriculture, green ecology, green tourism, and rural economic revitalization in provinces with lower total energy efficiency values will increase planned production and boost energy quality by creating a green economy. To increase families’ opportunities and income, access to electricity, piped water systems, and other facilities are fundamental. However, access to electricity is the first step in a household’s usage of renewable electricity compared to biomass or kerosene for heating and lighting. Policymakers also need to establish suitable policies to ensure that households are connected as smoothly as possible to the energy grid. Although the construction of the power grid and distribution would remain slow due to investment trends, other initiatives would provide easy access to remote areas, such as rooftop solar photovoltaics, solar farms, and small stand-alone generators. To encourage investments in this area, policies to support these distributed energy systems are needed.