EURO 2025 Leeds
Abstract Submission

2590. FORECASTING THE DIFFUSION OF RESIDENTIAL PHOTOVOLTAIC SYSTEMS IN AUSTRIA: AN AGENT-BASED MODELING APPROACH

Invited abstract in session TB-46: Tackling Energy Problems with ML and Scarce Data, stream Energy Economics & Management.

Tuesday, 10:30-12:00
Room: Newlyn 1.07

Authors (first author is the speaker)

1. Kagan Yüksel
School of Business and Economics, RWTH Aachen University
2. Reinhard Madlener
School of Business and Economics / E.ON Energy Research Center, RWTH Aachen University
3. Seif Marji
RWTH Aachen University

Abstract

Austria aims for 100% renewable electricity by 2030, placing residential solar PV at the center. However, PV adoption modeling is hampered by inconsistent records and limited policy insights, leading to data scarcity. We address these issues through data augmentation and scenario-based analyses in an agent-based OR framework that accounts for financial, social, and environmental factors influencing household decisions across diverse demographic and geographic contexts. Drawing on fragmented time-series (2008–2023), our method reconstructs missing information and refines parameters, enabling robust forecasts of PV adoption from rural to urban regions through 2050.
Results indicate that falling system costs and subsidy programs will expand PV uptake, though scenarios vary with shifts in subsidies and household incomes. Enhanced datasets improve predictive accuracy, particularly in underrepresented urban areas, revealing substantial opportunities as costs decline. By reconciling data gaps and evaluating multiple adoption trajectories, this framework provides actionable insights for policymakers seeking to design effective incentives and target high-potential regions.
This study demonstrates how innovative OR techniques, including data augmentation under scarce conditions, deliver evidence-based guidance for energy transitions. The findings underscore the need for flexible, data-focused strategies to sustain Austria’s renewable goals despite some serious market hurdles.

Keywords

Status: accepted

Back to the list of papers