Assessing the Quality of Natural Gas Consumption Forecasting: An Application to the Italian Residential Sector
Open Access
Abstract
(1) Background: The present paper aims at estimating the quality of the forecasts obtained by using one equation models. In particular, the focus is on the effect that the explanatory variables have on the forecasted quantity. The analysis is performed on the long term natural gas consumption in the Italian residential sector, but the same methodology can be applied to other contexts; (2) Methods: Different ex ante knowledge scenarios are built by associating different levels of confidence to the same set of explanatory variables. Forecasting results, coming from a standard regression algorithm and confirmed by a Kalman filter, are analyzed by means of covariance matrix propagation to assess the quality of the provided estimates; (3) Results: The outcomes show that one-equation models are very sensitive to the quality of the explanatory variables, therefore their erroneous estimation may have a relevant detrimental effect on the predictive accuracy of the model; (4) Conclusions: The overall ex ante forecasting accuracy of an example of one equation model is assessed. It has emerged that long-term forecasts need particular attention when the covered time horizon spans over decades. The information contained in the present paper is of interest for energy planners, supply network managers and policy makers in order to support their decisions.Keywords
This publication has 31 references indexed in Scilit:
- Long term outlook of primary energy consumption of the Italian thermoelectric sector: Impact of fuel and carbon pricesEnergy, 2015
- Comparison of static and adaptive models for short-term residential natural gas forecasting in CroatiaApplied Energy, 2014
- Review of energy models to the development of an efficient industrial energy modelRenewable and Sustainable Energy Reviews, 2014
- Long term energy demand projection and potential for energy savings of Croatian tourism–catering trade sectorEnergy, 2012
- Linear Regression Models to Forecast Electricity Consumption in ItalyEnergy Sources, Part B: Economics, Planning, and Policy, 2012
- Energy models for demand forecasting—A reviewRenewable and Sustainable Energy Reviews, 2012
- Mid-Term Energy Demand Forecasting by Hybrid Neuro-Fuzzy ModelsEnergies, 2011
- Industrial natural gas consumption in the United States: An empirical model for evaluating future trendsEnergy Economics, 2007
- Functional forms in energy demand modelingEnergy Economics, 2003
- Sensitivity of electricity and natural gas consumption to climate in the U.S.A.—Methodology and results for eight statesEnergy, 1997