University of Sheffield
Measuring the uncertainty associated with estimating national photovoltaic electricity generation: A Great Britain case study
Abstract
dc:description.abstractMonitoring near real-time national solar PV output is an increasingly important part of operating an electricity system. Monitoring PV output requires knowledge of the output of many PV systems embedded in the distribution network whose generation are not directly visible through existing transmission system metering. In this thesis a review of 27 national solar PV monitoring services which provide national PV output estimates for 20 different countries was performed showing that every service follows the same general approach. First, the PV yield is modelled using a set of data from reference PV systems providing data in real-time. Then the modelled PV yield is scaled by an estimate of the national solar PV capacity to estimate the national PV output. National PV output is then used, along with similar measurements for other embedded technologies such as wind, to train and validate electricity forecasts which ensure efficient electricity market operation. Using Great Britain as a case study, the total error and uncertainty associated with the estimates from a national PV monitoring service are analysed. There are three main sources of error which contribute to the overall error in the national PV output estimates; the sample bias error, the statistical error in the yield model, and the error in the national capacity estimate. For the GB PV monitoring service, the domestic sample was shown to be unbiased for estimating national PV output. However, at a regional level the domestic sample used in the GB service is biased for estimating commercial/utility PV systems. The statistical error in the yield model was shown to be $\pm1\%$ providing that a sample size of at least 6000 was used. The error in the GB national capacity estimate was shown to be $\pm5\%$. I can conclude that, the capacity error, at $\pm5\%$, dominates the yield calculation error, at $< \pm1\%$ and leads to an overall error in GB solar PV output estimates of $\pm5.1\%$. I also conclude that solar PV measurements, and consequently national electricity demand forecasts, are currently limited by the state of national PV capacity registers.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D
- Level dc:type.qualificationlevel
- doctoral
- Grantor dc:publisher.institution
- University of Sheffield
- Year dc:date.issued
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Huxley, Owen Thomas
- Advisors dc:contributor.advisor
-
- Buckley, Alastair
- Taylor, Jamie
- ORCID person:identifier.orcid
Rights
- Language dc:language.dcterms_iso639_2
- eng
Identifiers
dc:identifier.*- Identifier
- uk.bl.ethos.855676
- Dc Identifier Grantnumber
-
NIA_NGSO0008
EP/N509735/1