(547b) Beyond Lcoe: Market-Based Design of Flexible Solar Thermal Systems

Authors: 
Dowling, A. W., University of Notre Dame
Zheng, T., University of Wisconsin-Madison
Peng, X., University of Wisconsin-Madison
Maravelias, C., University of Wisconsin-Madison
Root, T. W., University of Wisconsin - Madison
Zavala, V. M., University of Wisconsin-Madison
Fueled by diverse economic, social, and political factors, over 50% of electrical energy is expected to come from renewable sources by 2030 in states including California and New York. This poses new challenges for grid operators, as many renewable sources (e.g., solar, wind) are intermittent. New sources of flexibility are required to provide contingency and ensure generation always matches consumption.

Concentrated solar power (CSP) systems are a promising technology for providing both baseline and flexible energy to power grids. In brief, solar thermal energy is obtained by concentrating solar radiation into a heat transfer medium (e.g., Dowtherm A, molten salt, air) which ultimately used to drive a thermodynamic power cycle (e.g., Rankine, Brayton). Thermal energy storage (TES) is realized by installing insulated tanks. Thus, CSP systems with TES are dispatchable and able shift generation to times with intermittent solar irradiance. Current literature and Department of Energy evaluations focus on improving designs in order to reduce the so-called Levelized Cost of Electricity (LCOE) (i.e., the average electricity generation cost) [1]. We argue that this approach dramatically underutilizes the inherent flexibility of CSP systems, as it neglects the time value of electricity [2]. This is important because in many regions of the world electrical energy is traded in multiscale wholesale markets that exhibiti significant price fluctuations.

In this talk, we present a general framework to co-optimize design decisions (i.e., collector field and storage sizing) and operational decisions (i.e., generation schedules and market participation) for CSP systems [3,4]. We use the framework to compare different performance metrics, such as LCOE, power purchase agreements (PPA) with time-of-day (TOD) factors, and net present value (NPV) with market revenues. Using historical market and solar irradiation data for California, we compute optimal designs with 15 hours of storage and minimum LCOE of 147 $ / MWh. We show that the average value of the electricity delivered to the grid, computed from historic market prices, is only 31 $ / MWh, indicating an implicit incentive of 116 $ / MWh from the LCOE mechanism. As a reference, average market prices in CA range from 20 $ / MWh (in the early morning and solar noon) to 50 $ / MWh (in the late evening) [2]. Motivated by these results, we compute the necessary incentives to facilitate direct CSP participation in California energy markets. We then discuss the link between the financial metric and the optimal design. In summary, we find that LCOE favors larger storage systems (10+ hours), as the value of electricity is time-invariant. In contrast, market-based mechanisms favor smaller storage systems (4 to 6 hours) that can take advantage of the dominant evening price spikes and provide ancillary services throughout the rest of the day. In conclusion, we argue that the proposed new market-based design approach can provide CSP technologies with a better value proposition and differentiation from cheaper non-dispatchable photovoltaic technologies.

References:

[1] SunShot Initiative. https://energy.gov/eere/sunshot/sunshot-initiative

[2] Dowling, Zheng, Zavala (2017). Economic Assessment of Concentrated Solar Power Technologies: A Review. Renewable and Sustainable Energy Reviews 72, pg. 1019-1032.

[3] Dowling, Kumar, Zavala (2016). A Multi-Scale Optimization Framework for Electricity Market Participation. Applied Energy 190, pg. 147-164.

[4] Dowling, Dyreson, Miller, Zavala (2016). Economic Assessment and Optimal Operation of CSP System with TES in California Energy Markets. To Appear in Proceedings for SolarPACES 2016. Preprint available at zavalab.engr.wisc.edu.