By Jason Deign

Eight industrial concerns have submitted a list of needs for DNICast, a pan-European project to improve CSP output forecasting.

The requirements will form the basis of a research programme lasting until 2017, says Dr Luis Fernando Zarzalejo Tirado, a Spanish investigator working with the project.

DNICast is being coordinated by the Observatoire Méditerranéen de l’Energie in Paris, France, and was launched in October last year to look for ways of enhancing current forecasting techniques for CSP.

“The principal objective is to offer a final user, in CSP or PV, a prediction of solar irradiation with a high spatiotemporal resolution, say to within a within a few metres and less than an hour,” says Zarzalejo.

The programme will aim to improve the information gleaned from existing methods of data collection rather than develop new forecasting technologies, he adds.

Commercial entities that have contributed to the project objectives include the plant developers Abengoa, Areva, Constructions Industrielles de la Méditerranée, Torresol Energy, Total and TSK Flagsol, along with IP-Solar, a CSP operational surveillance provider.

The Spanish Photovoltaic Union (Unión Española Fotovoltaica or UNEF) have also contributed. For the remainder of the four-year project, the objectives will be pursued within six work packages spread across a dozen European research centres.

These include bodies such as the Spanish Centre for Energy, Environmental and Technology Investigation (Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas or CIEMAT in Spanish), the University of Patras and Genossenschaft Meteotest.

Forecast uncertainties

“Currently, forecasts are made by several techniques, which have their own merits and drawbacks,” says DNICast on its web site. “The uncertainty in the forecast of the DNI is still too large and must be reduced.

“Therefore, we propose a concept of portfolio of innovative or improved methods and possibly hardware that can be assembled by company experts to answer the specific needs of a given plant.”

Users' feedback on the advances will be collected in the course of the project, says DNICast, and a workshop will be held for the demonstration of the final version of the methods and their combinations.

A key line of research will be to correlate satellite data and atmospheric models with information from whole sky cameras and ground-based sensors. “There are different models but none offer really high spatiotemporal resolution,” says Zarzalejo, of CIEMAT.

As part of the work, Zarzalejo and his team is linking up a network of 19 solar measurement stations across the Almería Solar Platform, managed by CIEMAT in the south of Spain. The network will be used to validate the results of any models emerging from the project.

If successful, the DNICast project could help power tower operators improve production by, for example, moving mobile heliostats out of shaded spots.

For parabolic trough plants, meanwhile, more accurate forecasting could help operators predict where temperature gradients might occur across the plant, and take remedial action.

More importantly, the work should allow all plant owners to provide a more accurate estimate of power production for the grid operator.

“It can help on the power despatch side because you can know more accurately how much you are going to deliver,” says Josefin Berg, senior solar analyst at IHS. “The grid operator appreciates that greatly.”

Availability of DNI information

In markets such as Italy, plant owners have to provide an accurate forecast of production a day in advance and are penalised if they do not meet the prediction to within predefined limits, she adds.

CSP experts acknowledge that availability of accurate direct normal irradiance (DNI) information is a problem for the industry.

In India, a lack of correlation between satellite forecasts and on-the-ground conditions, mainly caused by the presence of airborne particulates such as dust and soot, has been highlighted as one of the factors holding back growth of the CSP sector.

Commenting in a recent discussion on the CSP Today LinkedIn group, Dr Klaus Pottler, managing director at CSP Services, said: “We often see too little knowledge in the rather high uncertainty of satellite data regarding DNI on the side of project developers.

“Satellite data has to be calibrated by precise ground measurements. With Rotating Shadowband Radiometers, excellent sensors for DNI measurements are available at low price.”

He adds: “Many project developers use such stations but resist in spending some more money on a station control and data service that is urgently needed to secure high data quality. Especially in low-income regions, money is saved for resource checks.

“Much more money is lost later when the plant, surprisingly, does not deliver the projected energy output.”

DNICast will be holding a workshop on May 7 at the GENERA International Energy and Environment Fair in Madrid, Spain.

The meeting will aim to collect expressions of interest for a “DNI nowcasting system”, document current DNI forecasting practices and link the project up with the World Meteorological Organisation’s Sand and Dust Storm Warning Advisory and Assessment System.

To respond to this article, please write to the author, Jason Deign.