(576a) Optimal Selection and Design of Energy Storage Technologies Integrated with Advanced Energy Plants

With the increasing interest and focus on renewable energy sources, various fossil-fueled (e.g., coal-, natural gas- and oil-fired) power plants that were designed to operate at base-load conditions are being forced to cycle their load. Cycling includes load-following (ramping of the power generated up/down), shutdown and startup, and operation under variable and minimum load conditions. Increased renewable penetration creates a need for power plants to incorporate fast cycling, fast ramp rates and flexible operation to ensure grid reliability. Cycling leads to high thermal stresses and high failure rates in critical components [1]. To this end, suitable integration of energy storage facilities with the power plants can be instrumental in achieving near-steady state operation of these power plants. Most importantly, decentralized deployment of energy storage at the power plant level can exploit the existing equipment items and facilities at the host power plant for improving the efficiency of the storage technologies and reducing the storage capacity. However, realization of these benefits will critically depend on novel configuration/integration strategies with the least impact on the power plant operation and its configuration. Most of the previous work in integrating energy storage in power generation considers large-scale storage at the grid level, resulting in high storage capacities and costs [2-4]. Furthermore, the storage technologies differ widely based on the maturity of technologies, their costs, size, life, availability, efficiency, footprint, safety and environmental hazards. The grid scale imbalance can significantly vary based on the location, time of the day and the year. Lastly, dynamics of the entire integrated system including both the power plant and the storage technologies must be taken into account to obtain the cost-optimal solution.

In this work, we develop a mathematical optimization-based methodology for downselecting technological choices for decentralized energy storage with the existing power plants. Dynamic models of several electro-chemical, mechanical, and thermal storage technologies are developed along with the dynamics models of the host fossil-fueled plants. A mixed integer nonlinear programming (MINLP)-based optimization frameworks is used to select the most promising storage technology/technologies for a given power plant considering various tradeoffs between ramping capacities, ramping rates, life span and capital costs, and load balances. The framework not only helps to select the best technologies, it also provides the target capacities and the designs decisions that contribute to the overall cost of the system.

Energy generation, storage and dispatch – these are inherently dynamic operations. Specifically, a storage system needs to dynamically follow the time-varying loads and demands. The dynamic behavior must be carefully considered and modeled in order to ensure efficient power flow. Therefore, once the MINLP-based framework selects the candidate storage technologies, we consider their optimal integration with the power plant. When the system dynamics is included, it gives rise to Mixed-Integer, Nonlinear and Ordinary Differential Equation (MINODE)-based model. We propose two solution strategies to solve the MINODE problem. The first strategy is based on a time-averaged approximation, which leads to a large-scale MINLP. This is solved using a commercial solver. The second strategy utilizes a data-driven black-box optimization algorithm [5]. Both strategies are applied to illustrate the benefits of optimal downselection and integration of storage technologies for two representative regional-scale hourly electric load profiles.

References:

[1] Kumar N, Besuner P, Lefton S, Agan D, Hilleman D. Power Plant Cycling Costs.; 2012.

[2] Denholm P, Ela E, Kirby B, Milligan M. Role of Energy Storage with Renewable Electricity Generation.; 2010.

[3] Das T, Krishnan V, McCalley JD. Assessing the benefits and economics of bulk energy storage technologies in the power grid. Appl Energy. 2015;139:104-118.

[4] Nyamdash B, Denny E, O’Malley M. The viability of balancing wind generation with large scale energy storage. Energy Policy. 2010;38(11):7200-7208.

[5] Bajaj I, Iyer SS, Hasan MMF. A trust region-based two phase algorithm for constrained black-box and grey-box optimization with infeasible initial point. Comput Chem Eng. 2018;116:306-321.