 |
| Titel |
Evaluation of the performance of a meso-scale NWP model to forecast solar irradiance on Reunion Island for photovoltaic power applications |
| VerfasserIn |
Natacha Kalecinski, Martial Haeffelin, Jordi Badosa, Christophe Periard |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250082816
|
|
|
|
|
|
| Zusammenfassung |
| Solar photovoltaic power is a predominant source of electrical power on Reunion Island,
regularly providing near 30% of electrical power demand for a few hours per day. However
solar power on Reunion Island is strongly modulated by clouds in small temporal
and spatial scales. Today regional regulations require that new solar photovoltaic
plants be combined with storage systems to reduce electrical power fluctuations
on the grid. Hence cloud and solar irradiance forecasting becomes an important
tool to help optimize the operation of new solar photovoltaic plants on Reunion
Island.
Reunion Island, located in the South West of the Indian Ocean, is exposed to
persistent trade winds, most of all in winter. In summer, the southward motion of the
ITCZ brings atmospheric instabilities on the island and weakens trade winds. This
context together with the complex topography of Reunion Island, which is about
60 km wide, with two high summits (3070 and 2512 m) connected by a 1500 m
plateau, makes cloudiness very heterogeneous. High cloudiness variability is found
between mountain and coastal areas and between the windward, leeward and lateral
regions defined with respect to the synoptic wind direction. A detailed study of local
dynamics variability is necessary to better understand cloud life cycles around the
island.
In the presented work, our approach to explore the short-term solar irradiance forecast at
local scales is to use the deterministic output from a meso-scale numerical weather prediction
(NWP) model, AROME, developed by Meteo France.
To start we evaluate the performance of the deterministic forecast from AROME
by using meteorological measurements from 21 meteorological ground stations
widely spread around the island (and with altitudes from 8 to 2245 m). Ground
measurements include solar irradiation, wind speed and direction, relative humidity, air
temperature, precipitation and pressure. Secondly we study in the model the local
dynamics and thermodynamics that control cloud development and solar irradiance in
order to define new predictors to improve probabilistic forecast of solar irradiance. |
|
|
|
|
|
|