Skip to Content
  • Proceeding Paper
  • Open Access

2 September 2026

10 Pages

Adaptive Estimation of Risk in Gas Distribution Networks with an Extended Kalman Filter–Monte Carlo Framework †

Department of General Engineering, University of Telecommunications and Post, 1756 Sofia, Bulgaria
†
Presented at the International Conference on Electronics, Engineering Physics and Earth Science (EEPES2026), Bandirma, Turkey, 24–27 June 2026.

Abstract

Static forecast models used by gas distribution operators in regulatory submissions provide point estimates but cannot quantify the probability that approved targets will be met. This paper proposes an Extended Kalman Filter–Monte Carlo (EKF–MC) framework that (i) jointly estimates the residential client state and the parameters of the logistic S-curve growth model in a sequential Bayesian setting, and (ii) propagates the posterior parameter uncertainty through 100,000 Monte Carlo draws to construct calibrated probability distributions for the 2026–2027 regulatory forecast horizon. Applied to five years of regulatory submission data (2021–2025) for a licensed gas distribution operator in Sofia Province, Bulgaria, the EKF refines the saturation ceiling to M = 1995 ± 30 and the growth rate to r = 0.561 ± 0.083. The Monte Carlo analysis yields 90% forecast intervals of [1930–2002] for 2026 and [1939–2015] for 2027. Both intervals lie entirely below the regulator-approved targets (2039 and 2150), demonstrating a structural over-forecasting tendency in the regulatory approval process. The framework provides operators with a computationally efficient, auditable tool for quantifying forecast uncertainty in rate-case submissions and capital investment planning.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.