Available online at www.sciencedirect.com
ScienceDirect
Energy Procedia 57 (2014) 1100 – 1109
2013 ISES Solar World Congress
Development and test of gap filling procedures for solar
radiation data of the Indian SRRA measurement network
Marko Schwandta*, Kaushal Chhatbara, Richard Meyera,
Katharina Frossb Indradip Mitrab, Ramadhan Vashisthab,
Godugunur Giridharc, S. Gomathinayagamc, Ashvini Kumard
a
Suntrace GmbH, Brandstwiete 46, Hamburg, 20457, Germany
b
GIZ GmbH, Chennai, India
c
Centre for Wind Energy Technology, Chennai, India, Ministry of New and Renewable Energy, Chennai, India
d
Solar Energy Corporation of India, New Delhi, India
Abstract
Solar radiation measurements as most time-series data suffer from interruptions. Gaps may occur due to loss of
power, misalignment, failure of instruments, insufficient cleaning or other reasons. Quality check procedures identify
such malfunctioning and mark untrustworthy data by flags. Even well maintained stations with good equipment
usually show gaps. In the case of the Indian SRRA network with its 51 stations operating since 2011, typically around
7% of the data are flagged as potentially erroneous or missing. Duration of gaps ranges from few minutes to several
days. However many applications such as solar energy performance simulations need continuous time-series.
Therefore it is required to fill the measurement gaps with reasonable data. Depending on duration and type of missing
parameters various procedures can be used to fill gaps. This paper describes a set of procedures called ‘basic gap
filling’ for solar irradiance, which can be applied without having available additional data. From the over-determined
set of global, diffuse and direct radiation a single missing parameter can be calculated from the other two. When two
or more solar irradiance components are missing for short gaps, clearness indices are derived to calculate the missing
irradiance components. Basic gap filling procedure is applied as part of the SRRA/SolMap projects. The accuracy of
the applied basic gap filling methodology is tested and the results show a mean bias of ca. 3 % over GHI, DNI and
DHI over all types of gaps.
© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2013 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection
Selection and/or
and/or peer-review
peer-review under responsibility
under responsibility of ISES. of ISES
Keywords: Solar irradiance; gap filling procedure; DNI; GHI; satellite based gap filling; solar radiation; quality control;SRRA
* Corresponding author. Tel.: +34-622-298-766.
E-mail address: .
1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and/or peer-review under responsibility of ISES.
doi:10.1016/j.egypro.2014.10.096
, Marko Schwandt et al. / Energy Procedia 57 (2014) 1100 – 1109 1101
1. Introduction
Solar radiation measurements as most time-series data suffer from interruptions. Gaps may occur due
to loss of power, misalignment, failure of instruments, insufficient cleaning or other reasons. Quality
check procedures identify such malfunctioning and mark untrustworthy data by flags. Even well
maintained stations with good equipment usually show gaps. Duration of gaps can range from few
minutes to several days. Depending on duration and type of missing parameters various procedures can be
used to fill gaps. Many applications such as solar energy performance simulations need continuous time-
series. Therefore it is required to fill the measurement gaps with reasonably accurate data.
The Indian Ministry of New and Renewable Energy (MNRE) of Government of India (GoI) has
awarded a project to Centre for Wind Energy Technology (C-WET), Chennai in the year 2011 to set up 51
Solar Radiation Resource Assessment (SRRA) stations using the state-of –the-art equipment in various
parts of the country, especially at sites with high potential for solar power [1]. The GoI SRRA project has
synergy with SolMap project, which is implemented by the Deutsche Gesellschaft für Internationale
Zusammenarbeit (GIZ) in cooperation with MNRE. SolMap project contributes to SRRA project in
establishing quality checks on the data obtained as per International protocols and helping data processing
to generate investment grade data. SolMap project also aims to develop a solar radiation atlas for the
country. Various quality control tests are applied that check the plausibility of data, identify correctly
measured data and differentiate them from erroneous data.
Each of the 51 SRRA stations is equipped with one secondary standard pyranometer to measure Global
Horizontal Irradiance (GHI), one secondary standard pyranometer to measure Diffuse Horizontal
Irradiance (DHI) and a first class pyrheliometer to measure Direct Normal Irradiance (DNI). A two-axis
solar tracker is used to track the sun with the pyrheliometer and a shading assembly for the DHI
pyranometer. Apart from solar radiation parameters, these stations also measure other auxiliary
meteorological parameters like ambient temperature, wind speed and direction, humidity, pressure, rain
rate etc. All data were previously averaged in 10-minute time resolution. Since August 2012 they are
measured in 1 s and integrated to 1 min.
At present there are no standard procedures/protocols for gap filling of solar radiation data. Some
applied research is being carried out in this direction as part of the IEA Task 46, IEA SolarPACES and
EU research project ENDORSE. However, such quality check and gap filling procedures are not yet
applied to data from high research quality networks like BSRN, GAW, etc. This paper describes a set of
procedures called ‘basic gap filling’, which can be applied without having available additional data and
introduces the concept of satellite-based gap filling, which needs overlapping satellite-derived data.
2. Quality check and quality control of SRRA data
One of the main aims of SRRA is to provide investment grade bankable solar radiation data to the
solar industry, project developers, decision makers in the financing institutions and policy and also to the
scientific community. It is envisaged that this data will also be used for improvement and validation of
satellite-derived solar radiation data for India. Under such circumstances, quality check and control of
data forms the backbone of this data collection and monitoring system, ensuring proper operation and
maintenance of the system. Various quality control tests are applied that check the plausibility of data,
identify correctly measured data and differentiate them from erroneous data. The tests applied here follow
international best practices like those established by NREL’s SERI-QC [2], WMO’s BSRN [3], [4] and
those used in the EU-project MESOR [5].
A data flagging system is implemented to identify, differentiate and quantify different types of errors.
Such flags give feedback to users for identification of possible types of errors, which prove useful for
ScienceDirect
Energy Procedia 57 (2014) 1100 – 1109
2013 ISES Solar World Congress
Development and test of gap filling procedures for solar
radiation data of the Indian SRRA measurement network
Marko Schwandta*, Kaushal Chhatbara, Richard Meyera,
Katharina Frossb Indradip Mitrab, Ramadhan Vashisthab,
Godugunur Giridharc, S. Gomathinayagamc, Ashvini Kumard
a
Suntrace GmbH, Brandstwiete 46, Hamburg, 20457, Germany
b
GIZ GmbH, Chennai, India
c
Centre for Wind Energy Technology, Chennai, India, Ministry of New and Renewable Energy, Chennai, India
d
Solar Energy Corporation of India, New Delhi, India
Abstract
Solar radiation measurements as most time-series data suffer from interruptions. Gaps may occur due to loss of
power, misalignment, failure of instruments, insufficient cleaning or other reasons. Quality check procedures identify
such malfunctioning and mark untrustworthy data by flags. Even well maintained stations with good equipment
usually show gaps. In the case of the Indian SRRA network with its 51 stations operating since 2011, typically around
7% of the data are flagged as potentially erroneous or missing. Duration of gaps ranges from few minutes to several
days. However many applications such as solar energy performance simulations need continuous time-series.
Therefore it is required to fill the measurement gaps with reasonable data. Depending on duration and type of missing
parameters various procedures can be used to fill gaps. This paper describes a set of procedures called ‘basic gap
filling’ for solar irradiance, which can be applied without having available additional data. From the over-determined
set of global, diffuse and direct radiation a single missing parameter can be calculated from the other two. When two
or more solar irradiance components are missing for short gaps, clearness indices are derived to calculate the missing
irradiance components. Basic gap filling procedure is applied as part of the SRRA/SolMap projects. The accuracy of
the applied basic gap filling methodology is tested and the results show a mean bias of ca. 3 % over GHI, DNI and
DHI over all types of gaps.
© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2013 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection
Selection and/or
and/or peer-review
peer-review under responsibility
under responsibility of ISES. of ISES
Keywords: Solar irradiance; gap filling procedure; DNI; GHI; satellite based gap filling; solar radiation; quality control;SRRA
* Corresponding author. Tel.: +34-622-298-766.
E-mail address: .
1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and/or peer-review under responsibility of ISES.
doi:10.1016/j.egypro.2014.10.096
, Marko Schwandt et al. / Energy Procedia 57 (2014) 1100 – 1109 1101
1. Introduction
Solar radiation measurements as most time-series data suffer from interruptions. Gaps may occur due
to loss of power, misalignment, failure of instruments, insufficient cleaning or other reasons. Quality
check procedures identify such malfunctioning and mark untrustworthy data by flags. Even well
maintained stations with good equipment usually show gaps. Duration of gaps can range from few
minutes to several days. Depending on duration and type of missing parameters various procedures can be
used to fill gaps. Many applications such as solar energy performance simulations need continuous time-
series. Therefore it is required to fill the measurement gaps with reasonably accurate data.
The Indian Ministry of New and Renewable Energy (MNRE) of Government of India (GoI) has
awarded a project to Centre for Wind Energy Technology (C-WET), Chennai in the year 2011 to set up 51
Solar Radiation Resource Assessment (SRRA) stations using the state-of –the-art equipment in various
parts of the country, especially at sites with high potential for solar power [1]. The GoI SRRA project has
synergy with SolMap project, which is implemented by the Deutsche Gesellschaft für Internationale
Zusammenarbeit (GIZ) in cooperation with MNRE. SolMap project contributes to SRRA project in
establishing quality checks on the data obtained as per International protocols and helping data processing
to generate investment grade data. SolMap project also aims to develop a solar radiation atlas for the
country. Various quality control tests are applied that check the plausibility of data, identify correctly
measured data and differentiate them from erroneous data.
Each of the 51 SRRA stations is equipped with one secondary standard pyranometer to measure Global
Horizontal Irradiance (GHI), one secondary standard pyranometer to measure Diffuse Horizontal
Irradiance (DHI) and a first class pyrheliometer to measure Direct Normal Irradiance (DNI). A two-axis
solar tracker is used to track the sun with the pyrheliometer and a shading assembly for the DHI
pyranometer. Apart from solar radiation parameters, these stations also measure other auxiliary
meteorological parameters like ambient temperature, wind speed and direction, humidity, pressure, rain
rate etc. All data were previously averaged in 10-minute time resolution. Since August 2012 they are
measured in 1 s and integrated to 1 min.
At present there are no standard procedures/protocols for gap filling of solar radiation data. Some
applied research is being carried out in this direction as part of the IEA Task 46, IEA SolarPACES and
EU research project ENDORSE. However, such quality check and gap filling procedures are not yet
applied to data from high research quality networks like BSRN, GAW, etc. This paper describes a set of
procedures called ‘basic gap filling’, which can be applied without having available additional data and
introduces the concept of satellite-based gap filling, which needs overlapping satellite-derived data.
2. Quality check and quality control of SRRA data
One of the main aims of SRRA is to provide investment grade bankable solar radiation data to the
solar industry, project developers, decision makers in the financing institutions and policy and also to the
scientific community. It is envisaged that this data will also be used for improvement and validation of
satellite-derived solar radiation data for India. Under such circumstances, quality check and control of
data forms the backbone of this data collection and monitoring system, ensuring proper operation and
maintenance of the system. Various quality control tests are applied that check the plausibility of data,
identify correctly measured data and differentiate them from erroneous data. The tests applied here follow
international best practices like those established by NREL’s SERI-QC [2], WMO’s BSRN [3], [4] and
those used in the EU-project MESOR [5].
A data flagging system is implemented to identify, differentiate and quantify different types of errors.
Such flags give feedback to users for identification of possible types of errors, which prove useful for
