Enhancing Accuracy in Curtailment Loss and Profitability Predictions for Oversized Hybrid PV-Wind Power Plants Using High, Instead of Hourly, Resolution Data
03 HYB24-51
Presented by: Øyvind Klyve
Hybrid PV-Wind Power Plants (HyPPs) are gaining increased attention since such power plants might experience reduced capital and operational costs, as they are co-localized and share the point of interconnection (POI) with the grid, compared to standalone PV and wind power plants. However, for HyPPs where the POI's capacity is smaller than the combined installed PV and wind capacities, there will likely be periods where curtailing one or both assets is necessary. Such curtailment could result in revenue losses. Typically, when deriving optimal designs for HyPPs, PV and wind generation data on hourly resolution are used, as this is what is commonly available in databases. However, for an oversized HyPP, the curtailment losses, and thus the related revenue losses, will be underestimated when using hourly instead of using data on higher resolution.
This study undertakes a comparative analysis to explore how different data resolutions impact curtailment and revenue losses, in addition to the optimal techno-economic designs of HyPPs. We use operational data on 5 sec resolution from an existing HyPP in Eastern Germany, and compare the curtailment losses estimated with this resolution with those which are estimated using hourly resolution data. For a HyPP topology where the POI, PV and wind capacities are equal, meaning the POI is undersized relative to the total hybrid power plant capacity, curtailment losses are estimated to be 1.45%, 1.43%, and 1.09% when using data resolutions of 5 seconds, 1 minute, and 1 hour, respectively. Additionally, for this topology, using one-hour data instead of one-minute data leads to an overestimation of the total Net Present Value (NPV) by 1.86%. When the capacities of the PV and wind components are increased relative to the POI capacity, the curtailment loss and project NPV estimation error increase as well.
In our work, we generalize our finding by thoroughly describing the site where the data is collected from, such that the results can be extrapolated to other sites where e.g. the PV and wind capacities differ from the ones for the investigated site.
This study undertakes a comparative analysis to explore how different data resolutions impact curtailment and revenue losses, in addition to the optimal techno-economic designs of HyPPs. We use operational data on 5 sec resolution from an existing HyPP in Eastern Germany, and compare the curtailment losses estimated with this resolution with those which are estimated using hourly resolution data. For a HyPP topology where the POI, PV and wind capacities are equal, meaning the POI is undersized relative to the total hybrid power plant capacity, curtailment losses are estimated to be 1.45%, 1.43%, and 1.09% when using data resolutions of 5 seconds, 1 minute, and 1 hour, respectively. Additionally, for this topology, using one-hour data instead of one-minute data leads to an overestimation of the total Net Present Value (NPV) by 1.86%. When the capacities of the PV and wind components are increased relative to the POI capacity, the curtailment loss and project NPV estimation error increase as well.
In our work, we generalize our finding by thoroughly describing the site where the data is collected from, such that the results can be extrapolated to other sites where e.g. the PV and wind capacities differ from the ones for the investigated site.