Due to concerns about dwindling fossil fuel reserves and climate change, and to the falling costs of solar photovoltaic (PV) energy, solar power is being increasingly integrated into power grids. Unfortunately, several factors involving the composition of current grids limit solar power’s ability to be fully utilized. One such limitation is that the peak hours of solar availability do not coincide with the peak hours of demand. Another limitation is the current grid’s inflexibility, or inability to increase or reduce output from conventional energy sources to coincide with variable fluctuations in solar output. A solution to this problem has been proposed in the form of thermal energy storage (TES) deployed with concentrating solar power (CSP) (Denholm and Mehos 2011). This technology complements PV by storing otherwise wasted solar capacity as heat, which can be stored and used as output during periods where PV generation is minimal. Furthermore, PV with CSP stabilizes solar output to a firmer, more dependable source less subject to random fluctuations. This allows for solar to replace a portion of the grid currently occupied by conventional energy generation, and thus furthers the percentage of potential energy output due to renewables.—Donald Hamnett
Paul Denholm, Mark Mehos, and colleagues at the National Renewable Energy Laboratory used a REFlex model to predict the ability of CSP to increase grid flexibility and solar penetration in the Southwestern United States. This model compares the hourly load with renewable resources to calculate energy curtailment, based on the grid’s flexibility, or ability to change generator output to accommodate variable renewable energy sources. To determine the limits of PV, the researchers used weather data from 2005 and 2006 in the System Advisor Model (SAM), which converts the data into hourly PV output. These data were in turn used to model the interaction between solar and wind generated energy, using simulated data for 2005 and 2006 from the Western Wind and Solar Integration Study (WWSIS). Simulations of energy generation per hour were conducted over a year and the four-day period, April 7-10, was displayed as the paper’s example. The area under the curve was split into different sections representing contributions from the various energy sources. Simulations were run for the current PV system with an assumed 80% flexibility, and a PV with CSP system. In the latter example, CSP was added to the REFlex simulation using SAM produced hourly generation values.
The simulation under a PV-only scenario showed that a significant proportion of annual PV production, 5%, is curtailed due to a lower energy demand, since energy demand is low when PV output is high, and other generators cannot slow output down at fast enough ramp rates. In this situation, the PV curtailment and cost increase exponentially as the percentage of energy from PV increases. Clearly, this poses an obstacle to switching to an increasingly renewable grid because it puts a limit on the proportion of solar energy that is practical. In order for this PV to be utilized, a more flexible grid is required. When the researchers included CSP into the REFlex model, the data suggested that flexibility improved and curtailment reduced.
The CSP model was based on wet-cooled trough plant technology (Wagner and Gilman 2011). In this scenario, solar energy that would have otherwise been curtailed during low load hours was instead stored thermally during the day. At the same time, PV energy was incorporated into the grid; however, as PV lessened in the night hours and load increased, the stored CSP energy was run through the grid. Energy that would otherwise have been wasted during low load hours could now be stored and used during peak hours. This decreased the annual solar curtailment to less than 2%. Also, it increases the solar energy contribution to 25%, 15% PV and 10% CSP.
Denholm and Mehos also found that CSP/TES helped to lower the baseline amount of conventional energy needed, because CSP allows for solar energy to be utilized at all hours. This result has cascading effects, as the implementation of CSP lowers the minimum conventional generation requirements. CSP allows for additional solar power to be added at much lower marginal curtailment rates. Lowering the minimum load that relies on conventional energy opens up greater probabilities of increased solar and wind components. The researchers, through their complex grid analysis, determined that CSP technology has the ability to both increase solar efficiency and the possibility of the implementation of greater variable energy inputs.