How can net petroleum importers achieve risk aversion in a globalized world: a multi-regional input–output perspective

Net embodied petroleum importers face greater petroleum security risks in international trade due to their high foreign dependence. Therefore, paying better attention to the national flows and drivers of embodied petroleum consumption can effectively serve net embodied petroleum importers to avoid petroleum security risks and formulate environmental protection policy. This study utilized the multi-regional input–output model to determine the top five embodied petroleum net importers (i.e., the USA, Japan, Germany, the UK, and France) under the economic globalization, and traced the flows of embodied petroleum consumption through international trade. Combined with the logarithmic mean Divisia index method, this study also deeply investigated the drivers of changes embodied petroleum in net imports. The results show that from the perspective of international trade, the USA was not only the largest importer but also the largest exporter of embodied petroleum. Among the trade flows of embodied petroleum, the largest trade flow was from China to the USA, which indicated that China was the world factory and the USA was the consumer power. The embodied petroleum trade markets of Germany, the UK, and France was mainly distributed in Europe & Eurasia resulted from European economic integration. The largest contributor to the decrease of embodied petroleum imports in the top 4 net importers was the petroleum intensity effect. Meanwhile, the import dependence effect was the largest contributor to the increase of France’s embodied petroleum net imports. In order to further avoid the risk of petroleum security, it is an effective path worth exploring to construct the diversified petroleum product import strategy and improve energy efficiency.


Introduction
As the blood of industry, petroleum is an indispensable mineral resource to support national economic and social development. Petroleum supply security is an important guarantee for economic growth, and economic strength is an important indicator to measure the political power of a country. Due to the strategic and non-renewable nature of petroleum resources, obtaining and controlling sufficient petroleum resources have become one of the important goals of national security strategies, which has caused many petroleum-related social and political problems. There is a significant imbalance in the regional distribution of oil resources supply and demand, which makes the transnational and cross-regional flow of petroleum inevitable. In the context of economic globalization, petroleum trade is extremely active in international economic activities, and countries around the world meet their domestic petroleum demand through international trade.
However, when countries formulate national petroleum security strategies in the complex international petroleum trade pattern, they usually only consider the consumption of petroleum directly used in any specific production and final demand sector, while ignoring the indirect petroleum trade which embodied in the use of products and services that require petroleum inputs in the production process. Petroleum embodied in traded products and services is called embodied petroleum, which can substitute for direct petroleum trade Responsible Editor: Roula Inglesi-Lotz * Feng Wang lwjy1263@163.com ZaiXu Zhang mfmc1275@163.com and consumption. In the context of global trade, countries are increasingly participating in international trade, and the phenomenon of importing of embodied petroleum products or services from foreign countries to meet domestic petroleum demand has become widespread. In order to properly investigate the petroleum consumption and the impartiality of petroleum resources distribution in various countries, it is imperative to comprehensively review inter-regional petroleum flows and linkages in global supply chains. The embodied petroleum has been introduced to further identify the flows of petroleum under international trade, which can also be used to quantify the total petroleum demand that embodied in products and services consumed by individuals, towns, cities, or countries. Under international trade, embodied petroleum across global supply chains can be used to explore petroleum flows and the impartiality of petroleum resources distribution in various countries. For countries with net imports of embodied petroleum, petroleum resources have far more restrictive effects on the development of their national economic strength and national status than the net exporters of embodied petroleum. Because the petroleum products produced by net importers cannot satisfy domestic demand, their dependence on foreign countries is greater than that of net embodied petroleum exporters. Once the trading partners providing embodied petroleum products and services cannot guarantee supply, this will cause risks to the petroleum security of net importers. Therefore, the embodied petroleum consumption in the international trade of net embodied petroleum importers is more worth investigating.
Given that most of the previous studies on petroleum safety risk aversion have focused on direct petroleum consumption, ignoring the importance of embodied petroleum, coupled with the higher petroleum security risks of implied net petroleum importers, the contributions of this paper lie in the following points: (1) holistically examine the embodied petroleum trade flows in the globalized world economy to analyze the international trade pattern of global embodied petroleum, and further identified the top 5 net importers of embodied petroleum; (2) deeply determine the embodied petroleum trade flows and their foremost embodied petroleum trading partners of the top 5 net importers; (3) decompose the change of net embodied petroleum imports of top 5 net importers to explore the key factors that influencing the change of net embodied petroleum imports.
The rest of the paper is organized as follows: the "Literature review" section presents the related literature review. The "Methodology" section describes the MRIO model and decomposition methodology. The "Results and discussion" section describes the global trade volume of embodied petroleum, the regional and sectoral trading patterns, and the decomposition results of top 5 net importers. The "Conclusions and policy implications" section concludes.

Research on embodied petroleum
As the blood of industry, petroleum provides energy and material support to a wide range of industries, and there is many indirect petroleum hidden in these commercial products and services, which was called embodied petroleum. Embodied petroleum was the extension of embodied energy which was a considerable indicator used to estimate total energy demand and was conceptualized by Chapman in the 1970s (Chapman 2007). In 1980, an early definition of embodied energy was the total energy input embedded in the products or services which included direct energy input and indirect energy input (Costanza 1980). So far, the embodied energy method has been widely used in order to study the consumption and transfer of energy more systematically and accurately Limmeechokchai and Suksuntornsiri 2007;Lixiao et al. 2014;Su and Zhang 2016;Wu et al. 2016b). Xu et al. examined the embodied energy in imports of the UK and pointed out that if the embodied energy in net imports are considered, the gap between energy consumption and production in UK is far greater than the gap that is generally believed, which will have an impact on UK's energy security (Tang et al. 2013). Yang et al. established an EIO-LCA (Environmental Input-Output Life Cycle Assessment) model to evaluate the impact of Sino-USA trade on embodied energy consumption, and found that the USA was the net importer of embodied energy in the trade between the two countries (Yang et al. 2014). Li et al. evaluated the energy embodied in Macao's foreign trade during 2000-2011, and the results showed that the embodied energy of Macao was more than twice that of direct energy use (Li et al. 2014). The indirect effects emphasized by embodied energy have been recognized by various fields, such as carbon emissions (Arce et al. 2016;Davis and Ken 2010;Limmeechokchai and Suksuntornsiri 2007;Minx et al. 2009;Wiebe et al. 2012), coal use Wu et al. 2016b), land use (Weinzettel et al. 2013Yu et al. 2013), and water footprint (Ali et al. 2018;Chang et al. 2016;Chen and Chen 2013b).
Here, embodied petroleum was expanded from embodied energy, which indicates total petroleum consumption hidden in commercial products and services. For example, as a big manufacturing country, China produces shoes, sports equipment, tires, and other items that are exported to foreign countries. Synthetic rubber can be found in these products, and petroleum is the main raw material for making synthetic rubber. Petroleum is therefore indirectly used by these products made in China and flows between countries through international trade. Based on embodied petroleum accounting, the indirect petroleum flows can be used to better measure human use of petroleum resources and further quantify total petroleum demand that embodied in products and services consumed in international trade. In fact, many high-income countries are always importing from low-and middle-income areas to meet the domestic demand for petroleum-intensive products and services in order to reduce domestic energy consumption and protect the domestic environment. Therefore, in a highly globalized world, some high-income countries have shifted resources and environmental pressures to low-and middle-income countries through international trade (Arto et al. 2016;Chen et al. 2017a, b;Kanemoto et al. 2011;Peters and Hertwich 2006;Robert Alan et al. 2015). Tang et al. applied the Input-Output approach to quantify the net petroleum exports embodied in international trade, and found that China's dependence on foreign petroleum was underestimated (Tang et al. 2012).  revealed the overall evolution of the global oil supply chain from the perspective of both physical and virtual oil use, and provides measures to ensure the security of global oil supply.
However, at present, there is relatively little research on embodied petroleum, and most of the research is focused on embodied energy and embodied carbon emissions. Previous studies on embodied energy and embodied carbon emissions were usually in one country (Cellura et al. 2014;Limmeechokchai and Suksuntornsiri 2007;Tang et al. 2013), two countries (Dong et al. 2010;Wu et al. 2016a;Yang et al. 2014), an industry (Hong et al. 2016;Thormark 2002;Zhu et al. 2012), or a global perspective Chen and Chen 2013a;Zhang et al. 2017), but not enough for net importers. However, for countries with net imports of embodied petroleum, their actual petroleum requirement and petroleum import dependence are always underestimated. In order to accurately measure the total petroleum demand and regional and sectoral trading patterns of net importers, it is important to explore a comprehensive review for embodied petroleum and petroleum footprint between regions of net importers.

Research on factor decomposition technique
An increase in embodied net petroleum imports can exacerbate petroleum security risks in international trade throughout the country. Therefore, exploring the driving factors that trigger changes in net imports of embodied petroleum can provide a basis for effectively avoiding petroleum security risks. When performing factor decomposition analysis, there are frequently two main decomposition techniques: structural decomposition analysis (SDA) and exponential decomposition analysis (IDA). Since SDA is dependent on the input-output model that is also adapted to the accounting of embodied energy and carbon emissions in international trade, many previous studies have utilized the SDA method to quantify the driving forces behind the embodied energy and embodied emissions in trade (Das and Paul 2014;Su and Ang 2012b;Supasa et al. 2016). Su et al. investigated the key factors of Singapore's emission changes by applying SDA method and found that fuel switching and energy efficiency can contribute to curbing emissions growth ). Zhong used SDA method to decompose the change of embodied energy into six factors and found that the total final demand is the key driving force for the increase of embodied energy consumption (Zhong 2018).
However, the application of the SDA method needs to be based on input-output tables, so the impact factors explored are limited. The IDA method is a decomposition method based on time series data, and has received extensive attention because of its simple calculation and various decomposition factors Ang 2012a, 2016). The IDA method consists of the Laspeyres index method and the Divisia index method: the former application is hindered by the residual problem; the latter can be further developed into the arithmetic mean Divisia index (AMDI) and logarithmic mean Divisia index (LMDI) methods (Jianbo et al. 2016;Li and Wang 2016;Lin and Long 2016). LMDI is more popular than AMDI because of its simple calculation, easy to understand, and accurate decomposition results; in addition, LMDI has become the most popular method of factor decomposition research (Ang 2005;Ang and Choi 1997;Ang and Liu 2001;Ang et al. 2003Ang et al. , 2014Cansino et al. 2015;Zhang and Da 2015).When investigating the driving factors behind the embodied energy and emissions in international trade, the LMDI method was also utilized by many scholars (Dong et al. 2010;Liu et al. 2015;Su et al. 2013). Chen et al. employed the LMDI method to explore the drivers of embodied emissions during 2000-2001, and the results showed that trade structure and embodied emission intensity were the significant contributors to Macao as a net embodied emissions importer (Chen et al. 2017a, b). Tang et al. employed the Environmental Input-Output method and LMDI method to investigate the main factors that influencing the embodied energy use changes in trade of Guangdong's each sector (Tang et al. 2018).
In order to make the factors of decomposition not limited by the input-output tables, this paper used LMDI method to factorize the change of net embodied petroleum imports.

Multi-regional input-output model
Input-output models are basically used when accounting for energy flows embodied in international trade, including single-region input-output (SRIO) model and multi-region input-output (MRIO) model (Huang et al. 2020). Most single-country studies used the SRIO model due to low data requirements, but the results were biased due to the SRIO model's "technical identity hypothesis" (Wiedmann and Barrett 2013). The MRIO model considers that the energy consumption intensity of different regions varies with the level of production technology, which can make the measurement results more accurate (Zhang et al. 2021). Furthermore, the MRIO analysis can trace the consumption products of a country to its producing country and then derive the amount of energy transfer, which is beneficial to analyze the industrial linkages and trade links between countries and departments Qi et al. 2014). Therefore, this paper utilized the MRIO model to investigate petroleum consumption embodied in international trade flows.
The multi-regional input-output tables based on the world economy of m-region, n-sector provide a unified framework for MRIO analysis to explore the flow of environmental resources on a global scale, whose fundamental structure is shown in Table 1 (see Appendix Table 1). The row balance of tabular scheme which represents that each country's outputs are used to satisfy intermediate use and final use can be used as the starting point for MRIO analysis, which could be expressed as follows: where x r is the total economic output of the r country; A rr denotes the demand of each country's domestic production departments, that is, the direct consumption coefficient matrix; A rs denotes the demand of different country's production departments, and the element of the input coefficient matrix is a rs ij = z rs ij ∕x r j , where z rs ij denotes the intermediate use by the sector i of country r from the sector j of country s. y rr is the final demand of r country that from domestic production, and y rs is the final demand of country s that produced from country r. Equation (1) can be rearranged as follows: In simple notation, Eq. (2) can be rewritten as follows: (1) where L rs = I − A rs is the inverse Leontief matrix and represents the quantity of total output of country r resulting from an increase in the final demand of country s by one unit. Given the final demand, the total output that meets this final demand can be obtained by multiplying the inverse Leontief matrix and the final demand. Therefore, with properly given the inverse Leontief matrix and final demand, the total economic output of country r can be obtained as follows: The embodied petroleum intensity of sector i in country r satisfies f r i = p r i ∕x r i , where p r i denotes the total embodied petroleum use of sector i in country r. In the context of a globalized world economy, energy resources flow internationally through international trade. By expanding the inverse Leontief matrix, the petroleum inverse Leontief matrix can be obtained by multiplying the inverse Leontief matrix and a diagonal matrix F which contains the embodied petroleum intensity. Therefore, the embodied petroleum (EP) can be calculated by multiplying the petroleum inverse Leontief matrix and the final demand, which is shown as follows: In detailed matrix notation, Eq. (5) can be rewritten as follows: where EP rs represents the embodied petroleum consumption of country r for the final demand of country s, and can describe the embodied petroleum flows from country r to country s.
The total embodied petroleum in imports of country r, EPM r , can be expressed as follows: The total embodied petroleum in exports of country r, EPE r , can be expressed as follows: The total embodied petroleum in net imports of country r, EPN r , can be expressed as follows:

Decomposition methodology: the LMDI method
In order to analyze the net importers of embodied petroleum more comprehensively in global world economy, in this paper, the total embodied petroleum in net imports is decomposed into many elements to analyze the driving factors affecting the changes of total embodied petroleum in net imports. From a methodological point of view, in order to diversify the factors of decomposition and not be restricted by the input-output table, this paper adopts the LMDI method to decompose the changes in net implied oil imports.
Based on the Kaya identity (Kaya 1989), the total embodied petroleum in net imports can be expressed as the following identity: where EPC r = ∑ m r EP sr represents the total embodied petroleum consumption of country r; GDP r and P r represent the gross domestic product and total population of country r, respectively. D r = EPN r EPC r represents the share of the total embodied petroleum in the net imports to the total embodied petroleum consumption of country r, which is defined as embodied energy import dependence; I r = EPC r GDP r represents the embodied petroleum consumption per unit of gross domestic product of country r, which is defined as petroleum intensity; G r = GDP r P r represents gross domestic product per person of country r, which is defined as per capital GDP. The LMDI method includes two forms: additive decomposition form and multiplicative decomposition form (Ang 2004(Ang , 2015Jiang et al. 2018). In this paper, the additive decomposition form is used to decompose the change of the embodied petroleum in net imports into four driving factors, which is shown as follows: where t represents the target year and 0 represents the based year; ΔN r Dt represents the import dependence effect; ΔN r It represents the petroleum intensity effect; ΔN r Gt represents the economic effect;ΔN r Pt represents the population effect. These four driving factors can be calculated by the following formula:

Data source
The data on input-output calculations including MRIO table and petroleum use are from the World Input-Output Database (Timmer et al. 2015(Timmer et al. , 2016. The World Input-Output Database has been developed to analyze the effects of globalization on trade patterns, environmental pressures, and socio-economic development across a wide set of countries. The database covers 27 EU countries and 13 other major countries in the world for the period from 1995 to 2009 (see Appendix Table 2). The petroleum use includes Crude oil, NGL and feedstock, Diesel oil for road transport, Motor gasoline, Jet fuel (kerosene and gasoline), Light Fuel oil, Heavy fuel oil, Naphtha and Other petroleum products according to the list of Energy commodities in the WIOD system of satellite accounts. The data about GDP (constant 2010 US$) and population are from Word Bank. Entering the twentyfirst century, the world economy is growing rapidly, but it is difficult for the world petroleum production to grow synchronously, and the issue of petroleum import and export has aroused greater concern. Therefore, the time range of the study period in this paper is focused on 2001-2009.

International trade pattern of global embodied petroleum
In order to clearly show the changing trend of the international embodied petroleum trade pattern in different periods, this paper calculates the global petroleum consumption embodied in import and export trade according to the MRIO model, as shown in Fig. 1 and Fig. 2, respectively. The USA is the largest embodied petroleum importer, leading the other 39 regions with absolute advantage, which indicated the USA imports massive petroleum-intensive products and services. The status of the USA as the world's largest importer of implied oil consumption has also been verified by other scholars, such as Wu and Chen (2019), and Wang and Jiang (2019). However, embodied petroleum , with an average annual growth rate of 9.70%. As China becomes the world's new manufacturing hub, its rapidly growing industrial output is increasingly demanding implied petroleum products . China continued to import petroleum-intensive products and services from foreign countries, making its embodied petroleum in imports increased rapidly, which resulted in that EPM of China has surpassed Germany since 2007 and surpassed Japan in 2009. In addition, From the perspective of the global embodied petroleum import pattern, countries with considerable EPM are basically developed countries, except China. This reflects that developed countries tend to transfer high-pollution and highenergy industries to developing countries, and then obtain energy-intensive products and services through international trade to reduce domestic energy consumption and environmental pollution (Suri and Chapman 1998).
Regarding the embodied petroleum in exports (see Fig. 2), the USA was also the leading exporter with 8.59 million TJ of average value of embodied petroleum, but no longer had the absolute advantage over other countries during 2001-2009. The USA was not only the largest embodied petroleum importer but also the largest embodied petroleum exporter, which indicated that the USA was committed to promoting global trade liberalization and was the largest participating country in international trade between 2001 and 2009. Although the USA has advocated new trade protectionism in recent years, hoping to revive the domestic economy by destroying the global free trade system, it is undeniable that the USA has made huge profits in the process of trade liberalization (Li and Whalley 2021). China, as the second largest exporter of embodied petroleum, its growth trend of EPE was similar to that of the USA, with growth from 2001 (3.91 million TJ) to peak in 2008 (8.88 million TJ) and a downward trend in 2009 (7.71 million TJ). Russia and South Korea were the third and fourth embodied petroleum exporters followed by Germany and the Netherlands, with 5.44 million TJ and 4.80 million TJ of average values of embodied petroleum in exports, respectively. However, the embodied petroleum in exports of top 6 exporters all showed a down tendency during 2008-2009.  Fig. 4 The whole world economy is divided into 20 major economies, and the division criteria are shown in Appendix Table 2. International trade causes trade flows of embodied petroleum between regions, since the ROW under the WIOD division contains more complex countries, the trade flow associated with ROW is relatively larger than other countries. Among the trade flows of embodied petroleum (regardless of the trade flows of ROW), the largest trade flow was from China to the USA with 2.18 million TJ of embodied petroleum exported from China to the USA, which indicated that China was the world factory and the USA was the consumer power. China, as the world factory, not only meets domestic demand, but also provides petroleum-intensive products and services to foreign countries. ROW is the largest recipient of Chinese exports of embodied petroleum, followed by the USA and Japan. The second trade flow was related to the export from Canada to the USA. On average, 19 million TJ of embodied petroleum was imported into the USA each year during 2001-2009, of which 22.80% was from China and Canada. The third and fourth trade flows of embodied petroleum were exports from EU21, 1.89 million TJ of embodied petroleum was exported to Germany and 1.06 million TJ of embodied petroleum was exported to the USA. The trade flow which also exceeded 1.00 million TJ was from Russia to EU21 with 1.02 million TJ of embodied petroleum. The embodied petroleum trade flows in the USA are the most significant except for ROW; therefore, the USA was regarded as one of the world's major trading centers during 2001-2009.

Regional and sectoral trading patterns of embodied petroleum net importers
Based on the average value of EPN during 2001-2009, the top 5 net importers of embodied petroleum were the USA, Japan, Germany, the UK, and France. In order to explore more deeply the regional and sectoral trading patterns of embodied petroleum net importers, the import and export transactions of the five leading net importers have been analyzed in more detail, which were shown in Fig. 5. The 41 regions divided by WIOD were aggregated into five major regions, and 35 sectors divided by WIOD were aggregated into seven major sectors, which are shown in Appendix Table 2 and Appendix Table 3.
In the USA, as the largest embodied petroleum importer, 41.70% of its average import volume was from ROW during [2001][2002][2003][2004][2005][2006][2007][2008][2009]20.93% from Asia Pacific, 20.65% from Europe & Eurasia, 14.90% from North America, and 1.82% from South America. Most imports from other regions were all required by their heavy manufacturing sector. The average value of embodied petroleum in imports of the USA during 2001-2009 was 19 million TJ, and 15.4 million TJ of embodied petroleum imports came from the heavy manufacturing sector. Therefore, from the perspective of sectoral trading patterns, the heavy manufacturing sector was the largest supplier of imported embodied petroleum products for the USA. The second largest supplier for the USA was the transport sector, and imports of embodied petroleum products related to the transport sector accounted for 8.22% of total embodied petroleum imports, followed by light manufacturing sector (3.25%) and electricity sector (2.97%). Similarly, the USA was also the largest exporter of embodied petroleum products. Of the exports of the USA, ROW was the significant receiver followed by Europe & Eurasia and North America. 76.85% of the exported embodied petroleum products was from the heavy manufacturing sector in the USA, and 14.50% was from the transport sector.
Japan's import markets of embodied petroleum products were mainly distributed in ROW (49.62%), which mainly provided by their heavy manufacturing sectors. In addition, the imports from Asia Pacific and Europe & Eurasia constituted 29.90% and 10.15% of Japan's total embodied petroleum imports, respectively. On the sectoral level, the embodied petroleum imports from the heavy manufacturing sector make up 80.17% of the total embodied petroleum imports in Japan, which was much higher than other sectors. Similar to the USA, the second supplier for Japan was the transport sector, and imports of embodied petroleum products related to the transport sector accounted for 10.75% of total embodied petroleum imports, followed by electricity sector (2.57%). For the embodied petroleum exports from Japan, the largest trade flow was exported to ROW, accounting for 35.08% of total exports, followed by Asia Pacific (23.56%). It is worth noting that Japan's embodied petroleum imports from North America account for a small share of total imports, but 22.23% of its petroleum products were exported to North America. From the perspective of sectoral trading patterns, similar to imports, the heavy manufacturing sector was the Japan's largest export recipient, but the proportion decreased to 66.49%. Among Japan's total embodied petroleum products exports, 25.47% of that was exported to foreign transport sector, and 3.67% was to service sector.
The average value of embodied petroleum in imports of the Germany during 2001-2009 was 6.34 million TJ, and  Table 2 for details of the 20 major economies) Europe & Eurasia were the largest markets for Germany which provided 3.43 million TJ of embodied petroleum. Among the Germany's imports from Europe & Eurasia, 85.28% were associated with the heavy manufacturing sector. In addition, of the total embodied petroleum imports in Germany, 24.60% was from ROW and 12.20% was from Asia Pacific. On the sectoral level, the heavy manufacturing sector made the biggest contribution to Germany's imports which accounted for 81.82% of total embodied petroleum imports. Meanwhile, the mining sector played a weak role on total imports, accounting for only 0.92%. For the embodied petroleum exports from Germany, the embodied petroleum products provided by the heavy manufacturing sector constitute 84.37% of Germany's total exports, which were mainly exported to three markets of Europe & Eurasia, ROW and North America. It is worth noting that in the embodied petroleum trade between Germany and North America, Germany's imports from North America accounted for 7.93% of total imports, and North America is the second smallest import market that is only higher than South America. However, Germany's exports to North America accounted for 14.43% of total exports, and North America is the third largest export market higher than Asia Pacific.
The UK was the fourth largest net importer of embodied petroleum according to the average value in net imports during 2001-2009. The embodied petroleum import markets in the UK were mainly distributed in ROW, while the export markets were mainly distributed in Europe & Eurasia. Among the United Kingdom's total imports, 43.38% was from ROW, 35.90% was from Europe & Eurasia, 10.57% was from Asia Pacific and 9.53% was from North America. Of the United Kingdom's exports of embodied petroleum product, 52.59% are exported to Europe and 20.41% are exported to North America. However, as the largest import market, ROW was only the third largest export market. The embodied petroleum from the UK exported to ROW accounted for 20.34% of total exports. From the perspective of sectoral trading patterns, both import and export trades are mainly driven by the heavy manufacturing and the transport sector.
For France's embodied petroleum international trade, whether it is import or export, the largest trade markets were mainly distributed in Europe & Eurasia. 42.63% of total embodied petroleum imports was imported from Europe & Eurasia, and 51.24% of total embodied petroleum exports was exported to Europe & Eurasia in France. ROW was the second largest importing market, and 39.06% of embodied petroleum products were imported from there. Similarly, ROW was the second largest exporting market, and 25.49% of embodied petroleum products were exported to there. For France, the embodied petroleum products imported from abroad are basically provided by their heavy manufacturing sector. Similarly, the embodied petroleum products exported to foreign countries were also required by their heavy manufacturing sector of France. The transport sector also played an important role in the process of embodied petroleum international trade. The embodied petroleum imports associated with the transport sector made up 8.01% of total imports in France, and the exports associated with the transport sector accounted for 13.42% of total exports.

Decomposition of embodied petroleum in net imports
In order to effectively realize the risk aversion of embodied net petroleum importers in terms of petroleum security, this paper conducts a factor decomposition analysis of the changes of embodied petroleum in net imports, and the decomposition results of top 5 net importers are shown in Fig. 6 and Fig. 7. The petroleum intensity effect and the import dependence effect were the main factors in the decline of the United States' EPN, which led to the decline of 297 million TJ and 172 million TJ of the United States' EPN, respectively. Moreover, the economic effect and the population effect played the negative effect on the decrease of United States' EPN, accounting for about − 24.25% and − 26.58% of total imports decrease, respectively. As shown in Fig. 6, from the impact of the annual effect, the petroleum intensity effect always played an active role on the decrease of United States' EPN and the population effect always played a negative role on the decrease of United States ' EPN during 2001' EPN during -2009. For the import dependence effect, its direction of action was consistent with the direction of change in total net imports of embodied petroleum. For the economic effect, its direction of action also has changed, which played a promoting role during 2001-2007 and played an inhibiting role during 2007-2009. As shown in Fig. 6, the volume of embodied petroleum in net imports of Japan showed a trend of increase or decrease in volatility. However, from the cumulative effect, Japan's total net imports volume decreased by 154 million TJ during 2001-2009, which was mainly driven by the import dependence effect and the petroleum intensity effect. As shown in Fig. 7, the petroleum intensity effect always played the inhibitory effect during 2001-2009, which constituted 54.25% of Japan's total EPN and was the largest contributor to the reduction in embodied petroleum imports of Japan. Comparing the direction of the import dependence effect and the total effect, the direction of the annual embodied petroleum in net imports volume depended on the direction of the import dependence effect. However, with the results in the change of direction of action, the import dependence effect was the second largest contributor to the reduction in EPN of Japan. In addition, the economic effect and the population effect was relatively small, and has always promoted the increase of Japan's EPN. Figure 6 shows that Germany's EPN fell by a total of 45.24 million TJ during the study period. From the cumulative effect, the impact of the import dependence effect, the petroleum intensity effect, and the population effect played the promoting role for the increase of Germany's embodied petroleum imports. As shown in Fig. 7, although the population effect contributed only 2.57% to total embodied petroleum imports changes, it always played positive role on the decline in total net imports except for 2001-2003. As for the economic effect, although its direction of action has been changing, it contributed to the growth of the EPN from the cumulative perspective. Unlike the USA and Japan, the petroleum intensity effect of Germany did not always inhibit the EPN during 2001-2009, which was the second most influential factor effect after the import dependence effect. For the import dependence effect, it was the largest contributor for the annual changes of total EPN export for 2006-2007. During 2006-2007, the inhibiting effect of the petroleum intensity effect exceeded that of the import dependence effect.
As shown in Fig. 6, from the cumulative effect, the largest contributors to the changes of embodied petroleum net imports in the USA, Japan, and Germany were the petroleum intensity effect. Similar to the top 3 net importers, in the UK, the petroleum intensity effect also played the significant role on the changes of EPN, which accounted for about 149.26% of total imports decrease. The second largest contributor to the changes of EPN was the economic effect from the absolute value perspective, which accounted for about − 41.86% of total net imports decrease. Followed by the economic effect, the population effect also played the inhibiting effect on the decrease of EPN, which accounted for about − 27.43% of total imports decrease. Figure 7 shows that although the import dependence effect has the greatest impact on annual net imports changes except for 2002-2003 and 2005-2006, the cumulative effect was indeed the lowest, accounting for only 20.04% of the total imports decrease.
The embodied petroleum net imports of the top 4 net importers showed the downward trend during the period 2001-2009, while France, as the fifth largest net importer of embodied petroleum, accumulated a cumulative increase of 168 million TJ during the study period (Fig. 6). The embodied petroleum net imports in France have always shown a growing trend during the study period except for 2005-2006. As shown in Fig. 7, during 2001-2005and 2006-2008, the increase of embodied petroleum net imports was mainly driven by the import dependence effect; while during 2005-2006, the decrease of embodied petroleum net imports was mainly driven by the petroleum intensity effect. The population effect always played an active role on the increase of the embodied petroleum net imports during 2001-2009, which constitute 7.71% of France's total imports from the cumulative perspective. The economic effect's direction of action has changed, which played a promoting role in 2001-2007, and played an inhibiting role in the remaining years.

Conclusions
This paper explores the international embodied petroleum trade pattern and embodied petroleum trade flows in global world economy based on MRIO and LMDI method. A major contribution beyond previous research is the provision of time series analysis of the international embodied petroleum trade pattern in the global supply chain through MRIO method. This analysis reveals the top 5 net importers of embodied petroleum and analyzes the regional and sectoral trading patterns in international trade. Another contribution is to explore the drivers of the embodied petroleum trade volume of the top 5 net importers through the LMDI approach. This analysis reveals the extent to which the import dependence effect, the petroleum intensity effect, the economic effect, and the population effect affect different countries. Fig. 6 The changes of EPN and cumulative effect of decomposition for top 5 largest net importers during [2001][2002][2003][2004][2005][2006][2007][2008][2009] With this paper, the main conclusions can be shown as follows: For the world as a whole, the total embodied petroleum trade volume increased from 82.9 million TJ in 2001 to 103 million TJ in 2009 with the average annual growth rate of 2.81%. The USA was found to be the largest embodied petroleum importer and was also the largest embodied petroleum exporter during the whole study period. From the perspective of average value, the USA, Japan, Germany, China, and the UK were the five largest importers of embodied petroleum, in contrast to the USA, China, Russia, South Korea, and Germany as the top five exporters, and the USA, Japan, Germany, the UK, and France as the top five net importers. International trade causes trade flows of visual petroleum between regions; the trade flow associated with ROW is relatively larger than other countries resulted from ROW under the WIRD division contains more complex countries. If regardless of the trade flows of ROW, both largest and second largest trade flows were exported to the USA, the former came from China and the latter came from Japan. Moreover, the third and fourth trade flows of embodied petroleum were exports from EU21, 1.89 million TJ of embodied petroleum was exported to Germany and 1.06 million TJ of embodied petroleum was exported to the USA.
As for the sectoral trading patterns of embodied petroleum top five net importers, the commonality was that both imports and exports are driven primarily by the heavy manufacturing sector and the transport sector, which indicated that petroleum is the blood of industry. As for the regional trading patterns of embodied petroleum top five net importers, the main distribution markets for imports and exports in each country are basically different. For the USA and Japan, ROW was both their largest import market and the largest export market. Whether it is import or export, Asia Pacific was the second largest embodied petroleum trade market of Japan. However, the embodied petroleum trade markets of Germany, the UK, and France was mainly distributed in Europe & Eurasia resulted from they are all European countries.
Among the top five net importers of embodied petroleum, the volume of net embodied petroleum imports of top four net importers showed a down tendency during the whole study period. From the cumulative effect, the largest contributor to the changes of embodied petroleum imports in USA, Japan, Germany, and the UK was the petroleum intensity effect, which played an inhibiting role on the increase of their embodied petroleum net imports. However, embodied petroleum net imports in France has always shown a growing trend during the study period except for 2005-2006, which was manly driven by the import dependence effect. The import dependence effect was the largest contributor to the increase of its embodied petroleum net imports.

Policy implications
Because of the importance of petroleum in the country's economic politics, and as one of the main sources of air pollution, the results of this study also have important implications for petroleum security risk aversion and environmental protection policy formulation. Under international trade, if a country's petroleum is mainly dependent on imports, once the importing country stops supplying it, it will endanger national security. Moreover, under international trade, the petroleum use in a country is transferred internationally, resulting in its actual petroleum demand and petroleum import dependence are always underestimated or overestimated. Therefore, considering only direct petroleum consumption analysis is not conducive to the formulation and implementation of national petroleum security strategy.
The substantial increase in embodied petroleum imports and the continuous expansion of import dependence make economic and social development and national security face new challenges. Therefore, in order to ensure petroleum supply and national security, these countries must actively take measures to avoid petroleum security risk aversion. According to the research results of this paper, for the major importers of embodied petroleum, while increasing domestic petroleum and gas exploration and development and ensuring stable and increased domestic petroleum production, efforts should be made to do the following three aspects: (1) The regional trade pattern of embodied petroleum shows that the trading partners of embodied net petroleum importers are concentrated, which greatly increases the risk of petroleum security. Therefore, in order to ensure stable petroleum supply and diversify petroleum import risks, it is necessary to implement a diversified import strategy and build a new pattern of diversified petroleum imports. Specifically, these countries should actively develop strategic petroleum partnerships with other countries, strive to open new petroleum supply areas, and reduce excessive dependence on their original petroleum cooperative partners.
(2) Given that the petroleum intensity effect is the most powerful factor restraining the growth of embodied net petroleum imports, decreasing embodied petroleum intensity can be an effective way to achieve petroleum security risk aversion. The key to improving the efficiency of oil utilization lies in technological innovation. Therefore, the government should actively promote technological innovations related to oil utilization and increase the economic growth brought about by unit oil consumption.
(3) Given that the import dependence effect is the largest contributor to the increase of its embodied petroleum net imports, net importing countries can actively reduce import dependence through adjusting energy structure to effectively avoid oil security risks. The focus of energy restructuring is to increase the proportion of natural gas and renewable energy consumption. Specifically, the government should vigorously develop the natural gas industry and encourage the application of renewable energy such as wind and solar energy to reduce the intensity of dependence on petroleum.
Although our study has made some contributions, there are still limitations. Although this paper focuses on the net importing countries of embodied petroleum consumption, only the top five countries are selected in the specific analysis, and future research can expand the research object or conduct a classification and exploration of global countries according to income level. In addition, although this paper decomposes the embodied net petroleum imports based on the Kaya formula and the LMDI method, the selected indicators still have room for expansion.   Appendix 3 Table 1   Table 2   Table 3