Laporkan Masalah

Analisis Aliran Daya Probabilistik Pada Jaringan Distribusi Dengan Integrasi Pembangkit Fotovoltaik Dan SPKLU Menggunakan Pendekatan Markov Chain

Satria Bhekti Akhisnu, Dr. Ir. Jimmy Trio Putra, S.T., M.Eng., IPM

2026 | Tugas Akhir | D4 TEKNOLOGI REKAYASA ELEKTRO

Integrasi pembangkit Fotovoltaik dan Stasiun Pengisian Kendaraan Listrik Umum (SPKLU) pada jaringan distribusi menimbulkan ketidakpastian operasional akibat sifatnya yang stokastik. Penelitian ini menganalisis pengaruh integrasi Fotovoltaik dan SPKLU terhadap profil tegangan dan rugi-rugi daya menggunakan simulasi Probabilistic Power Flow (PPF) dengan pendekatan Markov Chain pada horizon operasi 24 jam. Pemodelan probabilistik dibangun menggunakan data Global Horizontal Irradiance (GHI) NASA POWER periode 2015–2025 dan data transaksi SPKLU di wilayah Surakarta periode Januari–Desember 2025. Metode K-Means Clustering diterapkan untuk membentuk profil representatif Fotovoltaik dan SPKLU, kemudian Markov Chain orde pertama digunakan untuk memodelkan probabilitas transisi antar-state. Simulasi dilakukan menggunakan metode Forward Backward Sweep (FBS) pada empat skenario operasi terhadap Penyulang GI Jajar 01 UP3 Surakarta sebagai sistem utama dan IEEE 33-Bus sebagai sistem uji. Hasil menunjukkan bahwa skenario Fotovoltaik Only menghasilkan kinerja terbaik dengan rerata tegangan 0,9771 p.u., tegangan minimum 0,9398 p.u., dan total rugi-rugi harian terendah 7,72 MWh (turun 34,1?ri skenario SPKLU Only). Sebaliknya, skenario Swap Fotovoltaik-SPKLU menghasilkan performa terburuk dengan probabilitas undervoltage tertinggi 41,35%, tegangan minimum 0,9082 p.u., dan rugi-rugi harian terbesar 15,09 MWh melampaui skenario SPKLU Only (11,71 MWh) yang tidak mengintegrasikan Fotovoltaik sama sekali. Analisis sensitivitas menunjukkan bahwa kapasitas SPKLU adalah parameter paling berpengaruh, sedangkan lokasi penempatan Distributed Energy Resources (DER) berperan sebagai faktor pengali yang dapat meniadakan seluruh manfaat integrasi Fotovoltaik. Penelitian ini membuktikan bahwa pendekatan PPF berbasis Markov Chain mampu merepresentasikan ketidakpastian sistem distribusi secara realistis sebagai dasar perencanaan integrasi Fotovoltaik dan SPKLU pada jaringan distribusi modern.

The integration of Photovoltaic (PV) generation and Electric Vehicle Charging Stations (EVCS) into distribution networks introduces operational uncertainty due to their stochastic characteristics. This study analyzes the impact of PV and EVCS integration on voltage profiles and power losses using a Probabilistic Power Flow (PPF) simulation based on a Markov Chain approach over a 24-hour operating horizon. The probabilistic model was developed using NASA POWER Global Horizontal Irradiance (GHI) data from 2015–2025 and EVCS transaction data from the Surakarta region collected between January and December 2025. K-Means Clustering was employed to generate representative PV and EVCS profiles, while a first-order Markov Chain was used to model transition probabilities between states. Simulations were performed using the Forward Backward Sweep (FBS) method under four operating scenarios on the GI Jajar 01 feeder of PLN UP3 Surakarta as the primary system and the IEEE 33-Bus system as a benchmark. The results indicate that the PV Only scenario achieved the best performance, with an average voltage of 0.9771 p.u., a minimum voltage of 0.9398 p.u., and the lowest total daily power loss of 7.72 MWh, representing a 34.1% reduction compared to the EVCS Only scenario. In contrast, the PV-EVCS Swap scenario exhibited the worst performance, with the highest undervoltage probability of 41.35%, a minimum voltage of 0.9082 p.u., and the largest daily power loss of 15.09 MWh, exceeding that of the EVCS Only scenario (11.71 MWh), which did not integrate PV generation. Sensitivity analysis revealed that EVCS capacity is the most influential parameter, while the placement of Distributed Energy Resources (DER) acts as a multiplier factor capable of eliminating the benefits of PV integration. The findings demonstrate that the Markov Chain-based PPF approach can realistically represent uncertainty in distribution systems and provide a reliable basis for planning PV and EVCS integration in modern distribution networks.

Kata Kunci : Fotovoltaik; jaringan distribusi; K-Means Clustering; Markov Chain; Probabilistic Power Flow; Stasiun Pengisian Kendaraan Listrik Umum

  1. D4-2026-505927-abstract.pdf  
  2. D4-2026-505927-bibliography.pdf  
  3. D4-2026-505927-tableofcontent.pdf  
  4. D4-2026-505927-title.pdf