Computational optimization of solar thermal and natural gas power systems

Advisors

Adam Brandt, co-advisor
Louis Durlofsky, co-advisor
Chris Edwards

Abstract

In order to facilitate the rapid adoption of renewable energy, it is necessary to identify low cost renewable-based systems to compete with fossil-based electricity generation. In this work, we use computational optimization to explore the integration of solar thermal energy into new or existing fossil energy systems. We use combined design and operations optimization to incorporate the time-varying aspects of solar thermal operation into the design decision. We explore two distinct systems. The first is a carbon capture retrofit to a coal-fired power plant, where we consider supplying the necessary auxiliary heat for carbon capture from both solar thermal and natural gas systems. The second is an integrated solar combined cycle, where solar and natural gas resources are used together for electricity production. These systems are both modeled with the Hybrid Power Plant Optimization (HyPPO) model (expanded in this work), which is a flexible model that uses modular representations of different interacting systems. These systems include gas turbines, multi-pressure heat recovery steam generators, steam turbines, and solar thermal fields, where interactions are modeled through mass and energy balances. In each of the power plant designs considered, we use HyPPO to perform optimizations under a variety of economic scenarios in order to assess the viability of solar thermal integration in different potential markets. Post combustion carbon capture using amine-based solvents, considered in the first part of this work, requires a significant thermal load. In a traditional retrofit carbon capture system, the efficiency penalty associated with extracting steam to meet this thermal demand is high. Consequently, we explore supplying this thermal load with an auxiliary power plant using either solar thermal energy, natural gas energy, or both. The solar thermal field considered is a low-temperature, enclosed-trough solar thermal system originally designed to provide steam for enhanced oil recovery. We examine a power plant that is limited by a CO2 emissions constraint of 499 kg/MWh, which is consistent with state-level legislation in multiple locations throughout the United States. We optimize the nonlinear, time-dependent operations over a series of days representative of a year, along with large-scale system design choices. The decision to use solar thermal, natural gas, or both was found to be sensitive to economic conditions -- particularly the electricity clearing price, natural gas price, and discount rate. Designs utilizing solar thermal energy are found to be profitable in a variety of economic scenarios, including high electricity clearing prices and low discount rates. The second power plant type examined is an integrated solar combined cycle (ISCC), which combines thermal streams from both a natural gas and a solar thermal system to produce electricity. In order to explore ISCC designs using computational optimization, we expand HyPPO to include a proxy model of a solar thermal field that, when combined with the heat recovery steam generator and natural gas turbine modules, allows for fast ISCC simulations. We use this model to explore the optimal operations of an ISCC. In order to explore ISCC design variables, which include heat exchange element sizes, turbine sizes, system pressures, and the integration strategy between solar thermal and natural gas streams within the heat recovery steam generator, we develop a computationally efficient reduced-form model. We use this reduced-form model for combined design and operations optimization, and find that a preferred integration strategy is to use the solar thermal system to supplement the low-pressure water stream preheating and evaporation. By applying bi-objective optimization, we explore the tradeoff between ISCC economics (net present value) and CO2 emissions intensity. We then quantify, under a variety of economic scenarios, the cost of CO2 avoided. The results for the cost of CO2 avoided suggest that our optimal ISCC designs may be competitive with other approaches for low-carbon electricity generation, such as carbon capture and storage.