• Energy Research

This study estimates the influence of income, electricity price, oil price, urbanization, and Asian economic crisis to electricity demands in residential, industrial, and commercial sectors of Indonesia using autoregressive distributed lag (ARDL) method. Our estimations reveal that electricity demands in all sectors during the period of 1969–2015 are significantly affected by income and urbanization in which the demands in all sectors are elastic to urbanization but inelastic to income. On the other hand, electricity price only significantly affects residential electricity demand while oil price does not have significant influence on electricity demands in all sectors. The electricity demands in residential and commercial sectors are also significantly inelastic to the number of electricity customers while Asian economic crisis in 1998 only affects industrial electricity demand. These findings are like most previous studies for other country cases but not to previous Indonesian studies that had suffered from the cointegration issues.

Abstract Indonesia has large renewable energy resources that are not always located in regions where they are needed. Sub-sea power transmission cables, or island links, could connect Indonesia’s high-demand islands, like Java, to large-resource islands. However, the role of island links in Indonesia’s energy transition has been explored in a limited fashion. Considering Indonesia’s current fossil fuel dependency, this is a critical knowledge gap. Here we assess the role of island links in Indonesia’s full power sector decarbonisation via energy system optimisation modelling and an extensive scenario and sensitivity analysis. We find that island links could be crucial by providing access to the most cost-effective resources across the country, like onshore photovoltaics (PV) and hydropower from Kalimantan and geothermal from Sumatera. In 2050, 43 GW of inter-island transmission lines enable 410 GWp of PV providing half of total generation, coupled with 100 GW of storage, at levelised system costs of 60 US$(2021)/MWh. Without island links, Java could still be supplied locally, but at 15% higher costs due to larger offshore floating PV and storage capacity requirements. Regardless of the degree of interconnection, biomass, large hydro, and geothermal remain important dispatchable generators with at least 62 GW and 23% of total generation throughout all tested scenarios. Full decarbonisation by 2040 mitigates an additional 464 MtCO2e compared to decarbonisation by 2050, but poses more challenges for renewables upscaling and fossil capacity retirement.

The Indonesian government has proposed energy conservation and renewable energy as the main strategies to decarbonise the electricity sector. Energy conservation is expected to reduce 17% of total energy consumption at business as usual projections by 2025 while renewable energy is anticipated to grow from 20.3 million tonnes of oil equivalent (MTOE) in 2015 to 92.2 MTOE by 2025. Nevertheless, the implementation of these strategies has been hindered by recent unexpected policy changes. In 2017, the country withdrew its energy conservation regulations to solve financial problems caused by low electricity demand growths and overinvestment in new power plants. Similarly, the premium feed-in tariffs (FIT) for pulling up renewable energy investments had been slashed to provide tariffs that are lower than the fossil-fuelled electricity generation costs. This was a direct response to low electricity tariffs resulting from solar farm auctions in Dubai and electricity subsidy constraints. These circumstances forced a reboot of clean energy policies in Indonesia.This research was undertaken to seek and define more effective and efficient clean-energy policy options using a diverse methodology approach consisting of a systematic literature review, time-series analysis, renewable energy projection evaluations, and hybrid energy models for distributed and centralised renewable energy systems. The systematic literature review discussed various renewable energy policies that are globally practised in on- and off-grid systems. The review also identified structural differences between developed and developing countries and the implications of using analytical tools for developing countries. The autoregressive distributed lag (ARDL) model, a time-series analysis, is used to understand the nexus of electricity demands, incomes and electricity prices in the residential, industrial, and commercial sectors. In addition, urbanisation and the number of electricity customers (representing dynamic transitions in developing countries) were used as control variables.Regarding renewable energy, their projection errors in developed countries were analysed to comprehend the most achievable renewables target and inaccurate assumptions applied in the projections. Thereafter, two hybrid energy models were developed which combined techno-economic analysis (TEA), input-output analysis (IOA) and life-cycle analysis (LCA) to assess the impact of proposed policies to the economy and natural environment. The first model, called the Agent-based Renewables model for Integrated Sustainable Energy (ARISE), used socioeconomic data to estimate photovoltaic (PV) market potential emerged from alternative policy interventions. The second model, Power Generator – Agent-Based Modelling (PowerGen-ABM), included linear programming (LP) approach to optimise power plant expansions in 15 main electricity systems under emission reduction targets. PowerGen-ABM uses the results of the ARDL estimations and the evaluations of renewable energy projections.The findings suggested several policy implications on electricity demand and supply sides. The ARDL estimations revealed that electricity demands in all sectors during the period from 1969 to 2015 were significantly affected by income and urbanisation. This finding provided an opportunity for controlling electricity demand growth by mitigating urbanisation effects, e.g., higher electrical appliance ownership. Evaluation of the renewable energy projections concluded that solar energy has the lowest level of uncertainty as it has the most reachable capacity projections. However, other renewable energies entail policies that are more effective, and further researches are needed for the advancement of reliable technology and accurate weather predictions. This thesis also provided ranges for the projection uncertainties of six renewable energy technologies, drawing attention to ways that the dominant errors in these renewable energy projections may be rectified. The results of the ARISE simulations indicated that it would be beneficial to scrap the PV donor gift policy for rural households without electricity access, and instead to improve the production efficiency of the PV industry and establish after-sales services and rural financing institutions. Net metering is the most effective policy for encouraging those in urban areas to invest in PV in a climate where fossil energy prices are on the rise while PV prices are dropping. Lastly, the PowerGen-ABM simulations recommended the utilisations of geothermal, large hydro, micro-hydro, and wind energy as these are the most cost-effective technologies that can be used to meet emission reduction targets.This research offered two contributions to existing literature in renewable energy systems and modelling. The ARDL estimations addressed the limitations of previous studies, which could not verify co-integrated relationships between electricity demand, income and electricity prices in Indonesia. Moreover, to the best of author’s knowledge, ARISE is the first energy model that integrates the triple bottom line of the economy - energy - environment (E3) to social behaviour analysis. PowerGen-ABM is also the first energy model that implements linear programming and E3 in the agent-based modelling (ABM) framework for analysing optimal expansions of power plants. As a pioneering model, ARISE and PowerGen-ABM have potentials to benefit from further improvements and integration of ARISE to PowerGen-ABM.

Abstract Indonesia has a target of achieving 23% of renewable energy share in the total energy mix in 2025. However, Indonesia does not have accurate and comprehensive data on renewable energy potentials, especially wind energy. This article aims to assess the theoretical potential of wind speed and to visualize the wind speed by province for the entire Indonesia. Our assessment relied on the Weather Research and Forecasting (WRF) model using Four-Dimensional Data Assimilation technique, also known as Nudging Newtonian relaxation. The robustness of our analysis is confirmed by using high-resolution data from the National Centers for Environmental Prediction–Final (NCEP - FNL) and Cross-Calibrated Multi-Platform (CCMP) Reanalysis satellite data. This study shows the WRF method is a feasible option to estimate wind speed data.

Data is used for Agent-based Renewables model for Integrated Sustainable Energy (ARISE), a solar energy policy model designed for use in developing countries. However, the data is for Indonesia case. The GIS files contains data for socio-economic characteristics and renewable energy potential; meanwhile m input output 6.txt is Leontief Inverse Matrix from Indonesian input output table 2010. Please refers the Appendix for more detailed description of the data.

Abstract Ecological concerns stemming from local energy usage generates more severe impacts on mountain economies, such as the Hindu Kush Himalaya (HKH) — a region consisting of eight nation boundaries. However, little is known about the regional energy consumption pattern and trends in the HKH due to data scarcity making it difficult to devise effective evidence-based energy policy and planning strategies. The purpose of this study is to bridge the existing data deficit gap in the HKH region. We first estimate the HKH-specific energy consumption patterns in 2013 and then forecast the sectoral energy consumption of the region for the period 2013–2030 by using the Long-range Energy Alternatives Planning (LEAP) system model. Our results show the dominance of residential sector (57%) and conventional biomass energy (50%) in the energy consumption landscape of the HKH. Industrial activities in HKH regions are still relatively underdeveloped compared to overall national averages. Energy consumption in the HKH region is around 5% of that in the HKH countries but is projected to increase from 5.89 million terajoule (TJ) in 2013 to 7.04 million TJ in 2030. The findings emphasise the importance of further investments and research in creating and periodically updating an HKH-specific energy database to overcome data-driven development challenges.

This study develops a hybrid energy agent-based model that integrates the input-output analysis, environmental factors and socioeconomic characteristics of rural and urban households in Indonesia. We use the model to estimate the effects of four solar energy policy interventions on photovoltaic (PV) investments, government expenditure, economic outputs, CO2e emissions and the uses of steel, aluminium, concrete and energy. The results of our analysis call for the abolition of the PV donor gift policy, the improvement of production efficiency in the PV industry and the establishment of after-sales services and rural financing institutions. A 100 W peak (Wp) PV under this recommendation would be affordable for 80.6% of rural households that are projected to be without access to electricity in 2029. Net metering is the most effective policy for encouraging urban people to invest in PV in a situation where fossil energy prices are increasing and PV prices are declining. A donor gift policy may induce USD 51.9 new economic outputs for every Wp of PV operating to capacity in 2029, but would require a subsidy of USD 18.6/Wp. The recommended policies do not require subsidies and reduce CO2eq emissions and the consumption of aluminium, energy, steel and concrete by between 83.1% and 89.7% more than the existing policy. Several policy implications are discussed in response to these findings. As a contribution to energy modelling literature, the model can be used for other developing countries by merely changing its data.