A Scalable Method for Extracting Soiling Rates from PV Production Data

Michael Deceglie; Matthew Muller; Sarah Kurtz; Zoe Defreitas

doi:10.1109/PVSC.2016.7749992

A Scalable Method for Extracting Soiling Rates from PV Production Data

Materials Science

Ametek

Research output: Contribution to conference › Paper › peer-review

33 Scopus Citations

Abstract

We present a method for analyzing time series production data from photovoltaic systems to extract the rate at which energy yield is affected by the accumulation of dust, dirt, and other forms of soiling. We describe an approach that is based on prevailing methods, which consider the change in energy production during dry periods. The method described here builds upon these methods by considering a statistical sample of soiling intervals from each site under consideration and utilizing the robust Theil-Sen estimator for slope extraction from these intervals. The method enables straightforward application to a large number of sites with minimal parameterization or data-filtering requirements. Furthermore, it enables statistical confidence intervals and comparisons between sites.

Original language	American English
Pages	2061-2065
Number of pages	5
DOIs	https://doi.org/10.1109/PVSC.2016.7749992 https://doi.org/10.1109/PVSC.2016.7749992
State	Published - 18 Nov 2016
Event	43rd IEEE Photovoltaic Specialists Conference, PVSC 2016 - Portland, United States Duration: 5 Jun 2016 → 10 Jun 2016

Conference

Conference	43rd IEEE Photovoltaic Specialists Conference, PVSC 2016
Country/Territory	United States
City	Portland
Period	5/06/16 → 10/06/16

Bibliographical note

See NREL/CP-5J00-65763 for preprint

NLR Publication Number

NREL/CP-5J00-67930

Keywords

data mining
production
rain
soil
temperature measurement
time series analysis

Access to Document

Cite this

@conference{07d50bbc8e754a0c800004cb112c57c4,

title = "A Scalable Method for Extracting Soiling Rates from PV Production Data",

abstract = "We present a method for analyzing time series production data from photovoltaic systems to extract the rate at which energy yield is affected by the accumulation of dust, dirt, and other forms of soiling. We describe an approach that is based on prevailing methods, which consider the change in energy production during dry periods. The method described here builds upon these methods by considering a statistical sample of soiling intervals from each site under consideration and utilizing the robust Theil-Sen estimator for slope extraction from these intervals. The method enables straightforward application to a large number of sites with minimal parameterization or data-filtering requirements. Furthermore, it enables statistical confidence intervals and comparisons between sites.",

keywords = "data mining, production, rain, soil, temperature measurement, time series analysis",

author = "Michael Deceglie and Matthew Muller and Sarah Kurtz and Zoe Defreitas",

note = "See NREL/CP-5J00-65763 for preprint; 43rd IEEE Photovoltaic Specialists Conference, PVSC 2016 ; Conference date: 05-06-2016 Through 10-06-2016",

year = "2016",

month = nov,

day = "18",

doi = "10.1109/PVSC.2016.7749992",

language = "American English",

pages = "2061--2065",

}

TY - CONF

T1 - A Scalable Method for Extracting Soiling Rates from PV Production Data

AU - Deceglie, Michael

AU - Muller, Matthew

AU - Kurtz, Sarah

AU - Defreitas, Zoe

N1 - See NREL/CP-5J00-65763 for preprint

PY - 2016/11/18

Y1 - 2016/11/18

N2 - We present a method for analyzing time series production data from photovoltaic systems to extract the rate at which energy yield is affected by the accumulation of dust, dirt, and other forms of soiling. We describe an approach that is based on prevailing methods, which consider the change in energy production during dry periods. The method described here builds upon these methods by considering a statistical sample of soiling intervals from each site under consideration and utilizing the robust Theil-Sen estimator for slope extraction from these intervals. The method enables straightforward application to a large number of sites with minimal parameterization or data-filtering requirements. Furthermore, it enables statistical confidence intervals and comparisons between sites.

AB - We present a method for analyzing time series production data from photovoltaic systems to extract the rate at which energy yield is affected by the accumulation of dust, dirt, and other forms of soiling. We describe an approach that is based on prevailing methods, which consider the change in energy production during dry periods. The method described here builds upon these methods by considering a statistical sample of soiling intervals from each site under consideration and utilizing the robust Theil-Sen estimator for slope extraction from these intervals. The method enables straightforward application to a large number of sites with minimal parameterization or data-filtering requirements. Furthermore, it enables statistical confidence intervals and comparisons between sites.

KW - data mining

KW - production

KW - rain

KW - soil

KW - temperature measurement

KW - time series analysis

UR - https://www.scopus.com/pages/publications/85003481181

U2 - 10.1109/PVSC.2016.7749992

DO - 10.1109/PVSC.2016.7749992

M3 - Paper

AN - SCOPUS:85003481181

SP - 2061

EP - 2065

T2 - 43rd IEEE Photovoltaic Specialists Conference, PVSC 2016

Y2 - 5 June 2016 through 10 June 2016

ER -

A Scalable Method for Extracting Soiling Rates from PV Production Data

Abstract

Conference

Bibliographical note

NLR Publication Number

Keywords

Access to Document

Other files and links

Fingerprint

Cite this