Performance, Reliability and Analysis of Photovoltaic Power Systems

1
Performance, Reliability and Analysis of
Photovoltaic Power Systems

• 5.4 Performance Prediction
• (solar irradiation part)
• Didier Mayer
• Vienna, 18-19 Septembre 2006

2
Activity 5.4 Performance prediction

• Overview of the existing activities in the field
  of solar irradiation forecasting and the
  corresponding on-going applications in the PV
  field.
• Identification of the potential of irradiation
  forecasting tools according to the requirements
  of possible target users.
• Use of the Task 2 work performed to date for
  systems performance prediction. What can be
  achieved ?
• New factors able to influence the performance of
  PV systems will be taken into account

3
Approach

• Connection with Solar Heating and Cooling
  programme dealing with Solar Resource Knowledge
  Management (task 36)
• Overview of on-going projects
• PV PERFORMANCE, PV SAT2, PV GIS,
• New developments on PV systems performance
  prediction, including factors other than
  irradiation
• Analyse and package the information of the Task2
  PDB and related activities to act as a  real
  life  support to performance prediction
  activities.

4
Solar irradiation

• Objectives
• evaluation of the global solar horizontal
  irradiation via satellite images and validation
  with ground based measurements (meteo stations)
• Calculation on an inclined surface and validation
  with ground based measurements at different time
  scales (PDB and other PV systems)

5
Evaluation of the solar resource

• HelioClim Solar irradiation
• Data base
• Europe - Africa
• http//www.soda-is.com

• Using data fusion, this method exploits satellite
  observation treatments, ground based measurements
  and other meteo and geographical information
• space step 10X10 km2
• daily time step since 1985 (helioclim1)
• hourly time step since 2004 (helioclim2)

6
Helioclim1 and Helioclim2 data base

• Helioclim 1
• Daily values of the global horizontal irradiance,
  calculated from one measurement every 3 hours,
  using a clear sky model
• One pixel 10X10 km2 over 50X50 km2
• Helioclim2
• hourly values of the global horizontal
  irradiance.
• One pixel 10X10 km2
• Starting in 2004

7
Advantages of HelioClim-2 vs HelioClim-1
Daily Values Large Triangles
Daily Profiles Now Available Better Sizing and
Analyses (Before, Modelling of Profile Was
Necessary)
Spatial Resolution every 5 arc angle (before,
B2 pixel and interpolation) Time
Sampling gt Hourly Values Available to Customers
(before, daily only) Space Sampling gt Better
Accuracy in Hourly Estimates 25 (before 32
) Space and Time Sampling gt Better Accuracy in
Daily Estimates 14 (before 20) Near
Real-Time Nevertheless and important,
consistency with the 21-years long HelioClim-1
database
C. Delamare, M. Albuisson, L. Wald - October 2004
8
Comparison between values from the PDB and from
the helioclim DB

• Comparison achieved for different locations with
  Helioclim1
• Rouvenac (FR) installation Bigou
• Lyon (FR) installation F1
• Bologna (IT) installation Bologna M
• Munich (DE) installation Munich Trade Fair
• Comparison achieved with helioclim2
• Hespul building near Lyon (FR)

9
Bigou 1994-1995
Accuracy 10-15 (yearly basis) 5-30 (monthly
basis)
10
Bigou - 1994
11
Bigou - 1995
12
F1 - 2004
Accuracy about 10 (yearly basis) 5-30
(monthly basis)
13
F1 - 2004
5 accuracy on a monthly basis
30 accuracy on a monthly basis
14
Bologna - 2001
Accuracy 15 (yearly basis) between 5 and 30
(monthly basis)
15
Munich trade fair - 2004
Accuracy 23 (yearly basis) 10-50 (monthly
basis) nearly 20 bias
16
Comparison with Helioclim2Hespul buliding
17
The new method Heliosat-3
The basic principles are different estimation
of the optical parameters of the atmosphere and
exploitation of a numerical model for radiative
transfer.  real time collaboration  with DLR in
Germany Advantages More physical, less
empirical adjustment Joint estimation of the
diffuse and direct components of the
irradiation Spectral distribution estimation
18
The new method Heliosat-3
Inconvenients Much more heavy to put in place
time and space consuming, computationally
speaking (x10) Benefits Better accuracy on the
estimation of the irradiation components Better
accuracy on the estimation of the spectral
distribution
19
Next steps

• Comparison of the calculated
  irradiation/irradiance on horizontal and tilted
  surfaces with PDB or other PV systems, using the
  helioclim2 DB
• hourly, daily, monthly
• From global horizontal using well known
  algorithms
• From diffuse and direct horizontal components and
  the calculation to the tilt. (heliosat3)

20
Deliverables

• Description of the potential transfer criteria of
  PV monitoring system insolation data into
  existing insolation mapping databases.
• Technical report
• PV systems monitoring input to PV systems
  performance prediction.

21
Technical reportPV systems monitoring input to
PV systems performance prediction

• Overview of the existing tools and possible
  trends
• with respect to the interest of identified
  target groups
• Selection of a number of PV systems in the PDB to
  compare measured irradiation data with existing
  ones in the SoDa DB.
• Forecasting methods and application

22
Schedule

• Description of the potential transfer criteria of
  PV monitoring system insolation data into
  existing insolation mapping databases
• (end of 2006- first semester 2007)
• Technical report
• PV systems monitoring input to PV systems
  performance prediction. (first version 12/2006,
  final draft 3/2007)