Volume 4 Number 1 .
International Assulta of Research and Engagement
(IARE)
January Ae June 2026 Pages: 23-31 https://edujavare.
com/index.
php/IARE/ Prediction of Renewable Energy Potential to Prevent Greenflation Using Bayesian Structural Time Series: BSTS with JASP Software to Predict PLTMH Potential Bunga Mardhotillah1.
Shally Yanova2.
Abdul Manab3.
Yosi Riduas Hais4.
Edi Saputra5.
Ade Adriadi6.
Ade Nurdin7 Mathematics Department.
Universitas Jambi.
Jambi.
Indonesia Environmental Science Department.
Universitas Jambi.
Jambi.
Indonesia Electrical Engineering Department.
Universitas Jambi.
Jambi.
Indonesia Infromatics Department.
Universitas Jambi.
Jambi.
Indonesia Biology Department.
Universitas Jambi.
Jambi.
Indonesia Civil Engineering Department.
Universitas Jambi.
Jambi.
Indonesia Email: bunga.
mstat08@unja.
Article history Submitted: 2025/12/27.
Abstract The global transition toward renewable energy is increasingly urgent to mitigate climate change and reduce dependence on fossil fuels.
however, it also introduces economic risks such as greenflation, driven by rising demand for green commodities.
In Indonesia, renewable energy development has become a national priority, with Jambi Province identified as a strategic region due to its significant micro-hydropower (PLTMH) potential.
This study aims to predict PLTMH potential as a means of supporting energy transition planning and preventing greenflation through data-driven policy The research employs a Bayesian Structural Time Series (BSTS) approach using JASP software, integrating Kalman Filter, spike-and-slab regression, and Bayesian Model Averaging.
Time series data from 2008Ae2024 were analyzed with 2,000 MCMC draws and a 1% burn-in to ensure estimation stability.
The results demonstrate a strong upward trend in PLTMH capacity, with high model accuracy indicated by an RA value of 0.
low residual standard deviation, and acceptable prediction uncertainty.
Forecasts suggest continued growth in PLTMH capacity over the next two decades before reaching a steady state.
The study concludes that BSTS is a robust and reliable method for predicting renewable energy potential and supporting counterfactual policy analysis.
This research contributes empirically to applied Bayesian time series modeling and practically to renewable energy policy planning, offering evidence-based insights to enhance energy security and mitigate greenflation risks.
Bayesian Structural Time Series.
Energy Policy.
Greenflation.
MicroHydropower.
Renewable Energy.
Keywords Revised: 2026/01/14.
Accepted: 2026/01/31 A 2026 by the authors.
This is an open-access publication under the terms and conditions of the Creative Commons Attribution 4.
0 International (CC BY SA) license, https://creativecommons.
org/licenses/by-sa/4.
INTRODUCTION
The energy transition to renewable resources is an increasingly urgent global agenda in addressing the climate crisis.
This effort aims to reduce carbon emissions while reducing Published by Edujavare Publishing International Copernicus Indeks Prediction of Renewable Energy Potential to Prevent Greenflation Using Bayesian Structural Time Series: BSTS with JASP Software A Bunga Mardhotillah et al,.
dependence on fossil fuels, which have dominated the global energy system.
Thus, the energy transition is not merely a technical issue but also part of the international commitment to sustainable development.
However, the energy transition process often has complex economic One emerging phenomenon is greenflation, which is price pressure caused by increased demand for green commodities, such as metals for solar panels or electric vehicle This phenomenon demonstrates that renewable energy policies cannot be separated from global market dynamics (Nidianti & Wijayanto, 2019.
Nurwani, 2016.
Rizki, 2021.
Sasono.
In Indonesia the energy transition has become part of national policy.
The 2019Ae2028 Electricity Supply Business Plan (RUPTL) targets the development of large-scale renewable energy power plants.
This policy reflects the government's commitment to expanding the energy mix while supporting the carbon emission reduction agenda.
Jambi Province is one of the regions projected to play a key role in the implementation of renewable energy policies.
The potential of hydropower, particularly through Microhydro Power Plants (PLTMH), is considered significant enough to support the community's electricity needs while strengthening regional energy security.
However, the energy transition is not without economic challenges: increased investment, infrastructure costs, and fluctuating commodity prices can pose risks to local and national economic stability.
Therefore, analytical methods that can more accurately predict the impact of policies are needed (Selle, et al, 2022 & Insani.
Bayesian Structural Time Series (BSTS) presents a relevant statistical approach to address these challenges.
BSTS is able to capture trends, seasonal patterns, and the influence of external factors in time-series data.
These advantages make it an effective tool for understanding renewable energy dynamics.
BSTS allows for counterfactual analysis, namely simulations of different conditions if a policy were not implemented.
In this way, the impact of energy policies can be measured more objectively, thus providing an empirical basis for decision-making.
BSTS can be used to predict the potential electricity production from microhydro power plants (PLTMH).
This analysis not only helps estimate the energy capacity that can be generated but also assesses the extent to which micro-hydro power plants can contribute to achieving the RUPTL targets (Wulandari, 2024 & Mokilane, et al, 2.
Research on predicting the potential of micro-hydro power plants (PLTMH) using BSTS is expected to provide a clearer picture of the renewable energy prospects in the region.
The research results can serve as a reference for local governments, investors, and the community in planning sustainable energy strategies.
This research is expected to not only contribute to the development of applied statistics but also have practical implications for energy policy in Indonesia.
Through the BSTS approach, it is hoped that the energy transition to renewable resources can be more measurable, efficient, and sustainable, particularly in supporting green energy development.
International Assulta of Research and Engagement (IARE) METHODS Bayesian Structural Time Series (BSTS) is a modern statistical approach designed to analyze highly complex time series data.
The main advantage of BSTS lies in its ability to separate various components within the data, such as long-term trends, seasonal patterns, and the influence of external variables.
Thus.
BSTS can provide a more realistic picture of the dynamics of a phenomenon.
The first component in BSTS is the Kalman Filter.
This method functions as an algorithm that iteratively updates trend, seasonality, and regression estimates based on incoming data.
The Kalman Filter is crucial because it reduces noise in the data, resulting in more stable and accurate predictions.
The Kalman Filter also enables dynamic system modeling.
This means that changes in trends or seasonal patterns can be adaptively captured according to the latest data developments.
This makes BSTS superior to classical methods, which tend to be static (Wulandari, 2024 & Mokilane, et al, 2.
The second component is Spike-and-slab Regression.
This approach is used to select relevant predictor variables from a large number of candidates.
With its variable selection mechanism.
BSTS is able to avoid the overfitting problem that often occurs in models with a large number of predictors.
Spike-and-slab Regression works by assigning a probability to each variable, determining whether it is worthy of inclusion in the model.
Variables with significant contributions are retained, while less relevant variables are eliminated.
This process improves model efficiency while strengthening the interpretation of the results.
The third component is Bayesian Model Averaging.
This technique combines prediction results from various possible models.
Instead of selecting a single best model.
BSTS leverages information from multiple models to produce more robust predictions (Wulandari, 2024 & Mokilane, et al.
Bayesian Model Averaging also reflects the Bayesian philosophy, which emphasizes By combining multiple possible models, prediction results become more robust to biases and errors specific to a single model.
In this study, the analysis was conducted using JASP software.
JASP was chosen because it provides a user-friendly interface and comprehensive support for Bayesian analysis.
Using this software makes it easier for researchers to implement BSTS without having to write complex code.
To ensure the reliability of the results, 2,000 MCMC draws were conducted.
Markov Chain Monte Carlo (MCMC) is a simulation technique used to estimate posterior distributions in Bayesian analysis.
The greater the number of draws, the more stable the resulting estimates.
1% of the data was discarded as burn-in.
The burn-in process aims to eliminate the initial influence of the Markov chain, which may not yet have reached stability.
Thus, the analysis results more closely reflect the true state of the posterior distribution.
Model evaluation is a crucial step in ensuring the quality of Without proper evaluation, analysis results risk misleading policymakers or Therefore, evaluation indicators are carefully selected to reflect model The first indicator is the coefficient of determination (RA).
The RA value indicates Prediction of Renewable Energy Potential to Prevent Greenflation Using Bayesian Structural Time Series: BSTS with JASP Software A Bunga Mardhotillah et al,.
how much variation in the data the model can explain.
The higher the RA value, the better the model's ability to capture data patterns (Rohmaningsih, et al, 2016.
Maulana, 2018.
Ross, 2.
The second indicator is the residual standard deviation.
Residual standard deviation measures the magnitude of the prediction error compared to the actual data.
A low residual value indicates that the model is able to produce predictions close to reality.
The third indicator is the prediction standard deviation.
Unlike the residual standard deviation, this indicator assesses the uncertainty in predictions.
A low prediction standard deviation indicates that the model produces consistent and reliable predictions.
The combined analysis consisting of the Kalman Filter.
Spike-and-slab Regression, and Bayesian Model Averaging makes BSTS a powerful approach in time series analysis.
With the support of JASP software and rigorous model evaluation, this research is expected to produce accurate predictions that are useful for the development of renewable energy policies and other fields requiring long-term data analysis (Wulandari, 2024 & Mokilane, et al, 2.
FINDINGS AND DISCUSSION
Jambi Province is among the 10 provinces with the largest renewable energy capacity development plans in Indonesia, reaching 2,189 MW according to the 2019Ae2028 RUPTL (Regional Development Pla.
This figure demonstrates the government's commitment to making Jambi one of the centers of green energy development in Sumatra.
The planned power plants include hydroelectric power plants (PLTA), microhydro power plants (PLTMH), solar power plants (PLTS), and biomass power plants (PLTBi.
This diversification is important to ensure energy security while reducing dependence on a single resource.
PLTA and PLTMH are the primary choices due to Jambi's significant hydrological potential.
The region's rivers provide a stable water flow, allowing it to be utilized to generate electricity sustainably.
PLTS
also has bright prospects given the relatively high solar radiation intensity in Jambi throughout the year.
With increasingly efficient solar panel technology.
PLTS can be a clean energy solution for both rural and urban areas.
PLTBio utilizes agricultural and plantation waste, particularly palm oil, a primary commodity in Jambi.
Utilization of biomass not only produces energy, but also reduces the environmental impact of organic waste.
To support this planning, microhydro power plant capacity predictions were conducted using JASP software using the Bayesian Structural Time Series (BSTS) approach.
This method was chosen because it comprehensively captures long-term trends, seasonal patterns, and external influences.
The BSTS prediction results indicate an increasing trend in microhydro power plant capacity in Jambi.
This confirms that the region's microhydro energy potential can continue to grow along with investment and infrastructure development.
International Assulta of Research and Engagement (IARE) Table 1.
Model Summary of Bayesian State Space Model Model validation is a crucial step in ensuring the reliability of prediction results.
The BSTS model demonstrated high accuracy with a coefficient of determination (RA) of 0.
This value indicates that almost all data variation can be explained by the model.
The residual standard deviation (RSD) of 385.
2 indicates that the prediction error is relatively small compared to the projected energy capacity.
Figure 1.
Bayesian Structural Time Series Plots This strengthens confidence that the model is able to capture data patterns well.
The prediction standard deviation (PSD) of 720 indicates the model's consistency in producing Despite the uncertainty, this value is still within reasonable limits for time series Prediction of Renewable Energy Potential to Prevent Greenflation Using Bayesian Structural Time Series: BSTS with JASP Software A Bunga Mardhotillah et al,.
analysis with complex variables.
Time Series Data for predict PLTMH potential start from 2008 - 2024 The results show that PLTMH capacity in Indonesia has experienced a significant upward trend from 2018 to 2028, and steady state in 2029 Ae 2038 as shown in Figure 2.
Figure 2.
BSTS Prediction Plot for PLTMH Potentials The research confirms that micro hydro power plants (PLTMH) have significant potential as a stable renewable energy source in Indonesia.
The advantage of micro hydro power plants lies in their relatively small scale, making them easier to implement in remote The increasing capacity trend indicates positive prospects for supporting the national energy transition.
The prediction graph using the BSTS shows that the potential of micro hydro
power plants will increase over the next 20 years and reach a steady state within the next 30
CONCLUSION
BSTS has proven effective in predicting renewable energy potential.
Its high accuracy and ability to generate counterfactual analysis make this method relevant to support energy policy planning.
Counterfactual analysis enables policymakers to understand the impact of a policy if it is not implemented.
This allows for more objective and data-driven energy The policy implications of this research are significant.
Regulations that support renewable energy development will strengthen electricity efficiency and reduce dependence on fossil fuels.
With appropriate policies, the risk of greenflation can be minimized.
Energy prices can be maintained stable despite increasing demand for green commodities.
The green energy transition in Indonesia, particularly in Jambi Province, also has socio-economic The development of micro-hydro power plants (PLTMH) and other power plants can International Assulta of Research and Engagement (IARE) create new jobs, increase community incomes, and strengthen the regional economy.
Renewable energy development supports the global agenda of climate change mitigation.
Jambi Province is expected to contribute to carbon emission reduction.
With a data-driven approach like the BSTS, energy planning can be carried out in a more measurable and sustainable manner.
This research shows that the combination of local potential, modern analytical methods, and policy support can accelerate the energy transition in Indonesia.
Jambi has a significant opportunity to become a green energy pioneer, while strengthening national energy security in the future.
Acknowledgments This research would not have been possible without the assistance of various parties.
The research team would like to express its gratitude to the Rector of Universitas Jambi, the Head of the LPPM Universitas Jambi, the Dean of the Faculty of Science and Technology of Universitas Jambi, and the readers of this article.
REFERENCES