Stochastic modelling of solar irradiance on horizontal and vertical planes at a northerly location

Craggs, C., Conway, E. and Pearsall, Nicola (1999) Stochastic modelling of solar irradiance on horizontal and vertical planes at a northerly location. Renewable Energy, 18 (4). pp. 445-463. ISSN 0960 1481

Full text not available from this repository.
Official URL: https:/


Sunshine levels incident in the plane of a photovoltaic panel are the overriding influence on electrical output, and modelling solar irradiance is therefore an essential step in the design and performance prediction of solar energy conversion systems. This study aims to assess the efficacy of SARIMA models and their potential for short-term prediction at a northerly latitude. Data was collected from a monitoring site on the roof of a 5-storey building at a city centre location in Newcastle upon Tyne, UK (latitude 55°N). Hourly and ten minute data relating to 13 and 15 day periods in two winters (1993, 1994) and two summers (1994, 1995) were utilised. Univariate stochastic modelling, using SARIMA models, is carried out for horizontal and south facing vertical solar irradiance. Results showed that these models provided a good fit for the ten minute averaged horizontal and vertical irradiance, with, on average, 82% and 85% of total variation being accounted for respectively. Use of hourly averaged data in these models gave a substantial reduction in the fit. Models for the winter data were a poorer fit than for summer for both orientations. It is concluded that the SARIMA approach can be used to develop prediction methods and to study rapid and large changes in PV output from extensive areas of solar cladding.

Item Type: Article
Subjects: F300 Physics
H600 Electronic and Electrical Engineering
Department: Faculties > Engineering and Environment > Mathematics, Physics and Electrical Engineering
Depositing User: Becky Skoyles
Date Deposited: 18 Feb 2015 10:56
Last Modified: 06 Feb 2020 09:12

Actions (login required)

View Item View Item


Downloads per month over past year

View more statistics