CHEMICAL ENGINEERING TRANSACTIONS VOL. 81, 2020 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Qiuwang Wang, Min Zeng, Panos Seferlis, Ting Ma, Jiří J. Klemeš Copyright © 2020, AIDIC Servizi S.r.l. ISBN 978-88-95608-79-2; ISSN 2283-9216 Spatial Decomposition Of Drivers Of Household Energy Use in Metro Manila Using LMDI Ana Karmela Lopeza,*, Neil Stephen Lopeza, Anthony Chiub, Jose Bienvenido Manuel Bionaa,c aMechanical Engineering Department, De La Salle University, Manila, Philippines bIndustrial Engineering Department, De La Salle University, Philippines cEnrique K. Razon Logistics Insitute, De La Salle University, Philippines ana_karmela_lopex@dlsu.edu.ph The trend on the migration to home-based work environment as an option for the global workforce implies that the Philippines’ residential energy use is as significant as in the industrial and commercial sectors. Improving efficiency and reducing energy intensity in these sectors are as important in the context of sustainability. Little to no studies have looked into the potential drivers to household energy consumption in a developing country such as the Philippines. Using a spatial decomposition analysis approach, various drivers to increasing household energy consumption in Metro Manila are estimated. Household size, income levels, energy intensity and energy use structure are considered in the study. The use of spatial decomposition techniques provides a new way to compare drivers between individual regions. The results are used to identify priority areas for electricity management and to draw insights for curbing household energy use. Policy implications are provided to conclude the study. 1. Introduction Energy utilization efficiency impacts environment and economy on a global scale. At an average yearly growth rate of 2.4 % for energy per capita from 2007 to 2017 (Department of Energy, 2017a), developing economies such as the Philippines benefit from understanding energy consumption efficiency. Existing government data show the consumption for various sectors, specifically; industry, residential, transport and commerce. Analyses of information is done on key areas of improvement for reasons of: (1) technological development and (2) policy making, (3) augmentation and (4) implementation. In the same year, the Philippine Energy Efficiency Roadmap (Department of Energy, 2017b) focused on these four areas. In 2017, DOE launched policy bridges to foster energy efficiency on a national level. Plans for execution of the roadmap shows actual monitoring of real-time data and the goal of initializing enabling technologies to make the proper changes. The 10 y data (2007 to 2017) shows that bulk of the total energy fuel consumption is allocated to transportation, industry, and residential sector. Transportation sector, ranges from 34.4 % in 2007 to 34.8 % in 2017; the industrial sector total fuel energy consumption remains relatively the same at 34.4 % to 34.9 %; and the residential total fuel energy consumption ranges from 31.7 to 27.2 %. This trend shows consistency for the averaged data. The total of approximately 85 % to 89 %, impact on the overall energy efficiency management of all three sectors require a detailed analyses. Evaluation of energy efficiency and economic performances provide insights on consumer behaviour, and other driving mechanisms that may have an effect on future projections. Impacts of these factors, tied in with their environmental implications and causes give way to how sustainable end goals can be achieved. Previous studies have focused on drivers to energy use and emissions in transportation (Lopez et al. 2018) and industrial (Sumabat et al., 2016) sectors. The importance of studying residential energy use is much more evident in the face of a globalized economy. Shifting demographics for some developing and developed countries, tied with the increased demand for a DOI: 10.3303/CET2081129 Paper Received: 30/04/2020; Revised: 05/06/2002; Accepted: 09/06/2020 Please cite this article as: Lopez A.K., Lopez N.S., Chiu A., Biona J.B.M., 2020, Spatial decomposition of drivers of household energy use in Metro Manila using LMDI, Chemical Engineering Transactions, 81, 769-774 DOI:10.3303/CET2081129 769 better lifestyle suggests a need allowed for shifts from office to home working environments to be done. Case studies have been proven that some employees who prefer to work at home can improve productivity that translates to benefits for both employer and employee (Bloom et al., 2015). In relation to this, Lopez and Biona (2019) assess accessibility in Metro Manila using a cumulative approach. Special cases such as need for isolation, lack of transportation, infrastructure limitations, domestic requirements and other prospects that could limit productivity from office-related factors can be remedied by household-based work. These cases, could also give rise to increased residential energy consumptions at any given period in time. So for these cases, industry-related emissions and energy consumption are offsetted towards residential sector. Methodology of analyses such as the logarithmic mean Divisia index (LMDI) help to develop policies and design products to direct towards goals for energy efficiency. Various decomposition techniques have already been used to identify areas for policy and technological changes. Methods linked to Divisia Index and Laspeyres Index have been used for various policy recommendations by institutions, academics and various agencies alike (Ang, BW., 2004). Other than the sectors, spatial-related effects and their interactions with known driving forces can be confirmed using the spatial variety of decomposition methods. Ang et al. (2015) discuss this variety of LMDI for spatial decomposition. Previously, Rapal et al. (2017) performed a spatial analysis for site selected in renewable energy projects in the Philippines. This study aims to uncover the characteristics or driving factors influencing high household emissions in various cities in Metro Manila, Philippines using the spatial LMDI method. Differences in economic and energy efficiency performance between cities offer discernment on how to best prioritize implementation of policies with economic and environmental cost-benefit in mind. 2. Methods and data The household energy consumption data is modeled by the authors from the Family Income and Expenditure Survey (Philippine Statistics Authority, 2015). The identity function used for decomposition analysis is shown in Eq(1). ∆𝐻𝐻𝑒 = ∑𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 × 𝑖𝑛𝑐𝑜𝑚𝑒 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 × 𝑒𝑛𝑒𝑟𝑔𝑦 𝑖𝑛𝑐𝑜𝑚𝑒 × 𝑒𝑛𝑒𝑟𝑔𝑦𝑖 𝑒𝑛𝑒𝑟𝑔𝑦 = 𝑖 ∑𝑝𝑜𝑝 × 𝑖𝑛𝑐 × 𝑖𝑛𝑡 × 𝑠𝑡𝑟 𝑖 (1) Where ∆HHe is the change in household energy consumption; pop refers to population; inc refers to household income per capita; int refers to household energy intensity; and str refers to the household energy use structure, i.e. the breakdown of energy use into components such as electricity use, direct use for transport (private), indirect use transport (public), and other miscellaneuos fuel use. The individual effects are calculated as follows in Eq(2) to Eq(5). The variable C refers to a specific city. As seen below, in spatial decomposition analysis, the comparison is between places instead of between time periods. The variable μ is the average of all cities considered in the study. This approach is taken from Ang et al. (2015). ∆𝑝𝑜𝑝 = 𝐻𝐻𝑒𝐶 − 𝐻𝐻𝑒𝜇 𝑙𝑛𝐻𝐻𝑒𝐶 − 𝑙𝑛𝐻𝐻𝑒𝜇 𝑙𝑛 ( 𝑝𝑜𝑝𝐶 𝑝𝑜𝑝𝜇 ) (2) ∆𝑖𝑛𝑐 = 𝐻𝐻𝑒𝐶 − 𝐻𝐻𝑒𝜇 𝑙𝑛 𝐻𝐻𝑒𝐶 − 𝑙𝑛𝐻𝐻𝑒𝜇 𝑙𝑛 ( 𝑖𝑛𝑐𝐶 𝑖𝑛𝑐𝜇 ) (3) ∆𝑖𝑛𝑡 = 𝐻𝐻𝑒𝐶 − 𝐻𝐻𝑒𝜇 𝑙𝑛𝐻𝐻𝑒𝐶 − 𝑙𝑛 𝐻𝐻𝑒𝜇 𝑙𝑛( 𝑖𝑛𝑡𝐶 𝑖𝑛𝑡𝜇 ) (4) ∆𝑠𝑡𝑟 = ∑ 𝐻𝐻𝑒𝑖 𝐶 − 𝐻𝐻𝑒𝑖 𝜇 𝑙𝑛𝐻𝐻𝑒𝑖 𝐶 − 𝑙𝑛𝐻𝐻𝑒 𝑖 𝜇 𝑙𝑛( 𝑠𝑡𝑟𝑖 𝐶 𝑠𝑡𝑟𝜇 ) 𝑖 (5) The direct comparison between two cities is done using Eq(6), ∆𝐻𝐻𝑒𝑗 (𝐶1−𝐶2) = ∆𝐻𝐻𝑒𝑗 (𝐶1−𝜇) − ∆𝐻𝐻𝑒𝑗 (𝐶2−𝜇) (6) 3. Results and discussion Figure 1 shows the estimated effects from the study per city. The main driving factors are income per person and energy intensity. The most significant effects are not the same for all cities. Figure 2 breaks down energy 770 use per city into electricity, miscellaneous, transport (direct) and transport (indirect). The impact of structure difference is apparent in the cities of Pateros, San Juan, and Paranaque. It is observed that these are the top three cities with the most direct transportation energy consumption as can be seen in Figures 1 and 2. Variations in the trend for energy use per city suggests that the drivers for each city. Figure 1: Estimated effects per city, measured against the group average Figure 2: Breakdown of energy use per city, modelled from Family Income and Expenditure Survey (PSA, 2015) Proximity of residents to major areas of commerce may attribute to the negative intensity offsetting the energy consumed of households in the area. Low use of miscellaneous type of energy source may suggest availability of infrastructure for electricity distribution. This finding contrasts high miscellaneous energy use of areas in the city at the outskirts of Manila. These are areas of development with infrastructure improvements implemented only after 2015. In Figure 3, it is observed that as income per capita and energy intensity increases, household energy use also increases. Las Piñas, Muntinlupa, Parañaque and Quezon City have above average miscellaneous energy -15,000 -10,000 -5,000 0 5,000 10,000 15,000 CLC LPN MKT MLB MND MNL MKN MNT NVT PNQ PSY PSG PTS QZN SJN TGG VLZ ∆ H H E n e rg y U se ( G J) ∆pop ∆inc ∆int ∆str ∆HH Energy Use 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 H H E n e rg y U se , to ta l ( G J) H H E n e rg y U se , b y t y p e ( G J) Electricity Misc Transport (Direct) Transport (Indirect) Electricity (Ave.) Misc (Ave.) Transport (Direct, Ave.) Transport (Indirect, Ave.) HH Energy Use (Total) HH Energy Use (Total, Ave.) 771 consumption. Possible interactions in the proximity of areas of commerce and type of industries existing in and neighboring areas could potentially contribute to the positive effects of income and household energy use. Figure 3: Difference in household energy use versus difference in income and difference in energy intensity The drivers to household energy use are further investigated in Figures 4a to 4c. In Figure 4a, it is surprising to see that electricity consumption is not the primary driver of high energy consumption in households. Makati is the only city with both above-average energy use and above-average electricity share in total consumption. Based on Figure 4b, household energy consumption is more directly proportional to direct energy consumption from transport. The household energy consumption is inversely proportional to indirect energy consumption from transport (i.e. from patronage of public transport) according to Figure 4c. This means that the primary characteristics of most high energy consuming households are high private transport use and low public transport use. Interesting areas of comparison are cities with overlapping values for effect of structure. Marikina and Navotas are cities at the far ends of Manila. Low energy use may be attributed to electricity infrastructure developments. Muntinlupa City shows high sensitivity on the stuctural effect (public transportation) as the highest energy consumer for this sector. The benefits of using private transportation in this case may be offsetted by distance-to-work, traffic route conditions and availability of public transportation options. Infrastructure developments from 2015 onwards should allow for improvements for the cities at the southern part of Metro Manila. Neighboring cities such as Mandaluyong and Manila also have negative structural effect on direct transportation this potentially resulted in low energy use. This shows that there are varying primary drivers for each city. Mechanisms for each driver may also include behavior-related factors. These mechanisms may manifest in some areas in the city. The conclusive identification of the effect of these factors is not directly investigated in this study. The study of these factors may be more relevant for areas not behaving according to trend observed. CLC LPN MKT MLB MND MNL MKN MNT NVT PNQ PSY PSG PTS QZN SJN TGG VLZ MKT MLB MND MNL MKN MNT NVT PNQ PSY PSG PTS QZN SJN TGG VLZ -8,000 -6,000 -4,000 -2,000 0 2,000 4,000 6,000 8,000 10,000 -15,000 -10,000 -5,000 0 5,000 10,000 15,000 ∆ H H E n e rg y U se ,G J ∆inc, Php ∆int, Php/GDP 772 Figure 4: Difference in household energy use versus energy use structure: a) energy use as electricity ; b) direct energy use for transport (private transport); and c) indirect energy use for transport (public transport) CLC LPN MKT MLB MND MNL MKN MNT NVT PNQ PSY PSG PTS QZN SJN TGG VLZ -8,000 -6,000 -4,000 -2,000 0 2,000 4,000 6,000 8,000 10,000 -3,000 -2,000 -1,000 0 1,000 2,000 3,000 4,000 ∆ H H E n e rg y U se ,G J ∆str-elec CLC LPN MKT MLB MND MNL MKN MNT NVT PNQ PSYPSG PTS QZN SJN TGG VLZ -8,000 -6,000 -4,000 -2,000 0 2,000 4,000 6,000 8,000 10,000 -3,000 -2,000 -1,000 0 1,000 2,000 3,000 4,000 5,000 6,000 ∆ H H E n e rg y U se ,G J ∆str-tdir CLC LPN MKT MLB MNDMNL MKN MNT NVT PNQ PSY PSG PTS QZN SJN TGG VLZ -8,000 -6,000 -4,000 -2,000 0 2,000 4,000 6,000 8,000 10,000 -2,000 -1,500 -1,000 -500 0 500 1,000 1,500 ∆ H H E n e rg y U se ,G J ∆str-tind (c) (a) (b) 773 4. Conclusion In this study, the use of spatial decomposition using LMDI to analyze drivers of high household energy consumption in Metro Manila was demonstrated. High household energy consumption is driven primarily by high income per capita and high energy intensity, but it is important to note that the primary driver for each city is not the same. Looking at the utilization, it is surprising to see that electricity consumption is not the primary driver, but rather private transport use. Potential interactions for cities in proximity with each other may also be looked into. With this information, should the government aim to manage energy consumption in households, the correct driver can be targeted. High energy intensity in households should be addressed. Other than high private transport use, it is also observed that energy-intensive households consume a lot of fuelwood and charcoal. In succeeding studies, more validating information can be discussed to support the findings and the study can be extended to estimate emissions. Potential applications of this study may be done to developing megacities. Acknowledgement The authors would like to acknowledge the financial support in the conduct of this study from the Department of Science and Technology, Philippines, Engineering Research and Development for Technology program, the Office of the Vice Chancellor for Research and Innovation of De La Salle University, and the Commission on Higher Education, Philippines, Grants-in-Aid program. Nomenclature CLC Caloocan PNQ Parañaque LPN Las Piñas PSY Pasay MKT Makati PSG Pasig MLB Malabon PTS Pateros MND Mandaluyong QZN Quezon MNL Manila SJN San Juan MKN Marikina TGG Taguig MNT Muntinlupa VLZ Valenzuela NVT Navotas References Ang B.W., 2004, Decomposition analysis for policymaking in energy: which is the preferred method? Energy Policy, 32, 1131-1139. Ang B.W., 2015, LMDI decomposition approach: A guide for implementation, Energy Policy, 86, 233-238. Ang B.W., Xu X.Y., Su B., 2015, Multi-country comparisons of energy performance: The index decomposition analysis approach, Energy Economics, 47, 68-76. Bloom N., Liang J.,Roberts J., Zhichun J.Y., 2015, Does Working from Home Work? Evidence from a Chinese Experiment, Quarterly journal of Economics, 130(1), 165-218. Department of Energy, 2017a, Key Energy Statistics accessed 29.04.2020. Department of Energy, 2017b, Philippine Energy Efficiency Roadmap 2017-2040 accessed 30.04.2020. Lopez N.S., Soliman J., Biona J.B.M., 2020b, Cost-benefit analysis of alternative vehicles in the Philippines using immediate and distant future scenarios, Transportation Research Part D, 82, 102308. Lopez N.S., Biona J.B.M., 2019, Assessing transport equity through a cumulative accessibility measure and Google Maps: a case for healthcare in Metro Manila, Philippines. In R. Hickman et al. (Eds.), A Companion to Transport, Space and Equity, Edward Elgar Publishing, UK, 136-146 Lopez N.S., Chiu A.S.F., Biona J.B.M., 2018, Decomposing drivers of transportation energy consumption and carbon dioxide emissions for the Philippines: the case of developing countries, Frontiers in Energy, 12, 3, 389-399. Philippine Statistics Authority, 2015, Family Income and Expenditure Survey (FIES) accessed 06.06.2020. Rapal B.K.A.L., Sumabat A.K.R., Lopez N.S.A., 2017, Analytic hierarchy process for multi-criteria site selection of utility-scale solar and wind projects, Chemical Engineering Transactions, 61, 1255-1260. Sumabat A.K., Lopez N.S., Yu K.D., Hao H., Li R., Geng Y., Chiu A.S.F., 2016, Decomposition analysis of Philippine CO2 emissions from fuel combustion and electricity generation, Applied Energy, 164, 795-804. 774 https://www.doe.gov.ph/sites/default/files/ https://psa.gov.ph/sites/default/files/FIES%202015%20Final%20Report.pdf