(467b) Improved Price-Taker Optimization of Integrated Energy Systems Using Data-Driven Surrogate Models

Integrated energy systems (IESs) explore synergies between different energy resources and technologies, which provide operational flexibility and improve energy efficiency [1]. In order to obtain higher economic values while maintaining a feasible IES operation, optimization tools are widely employed in the techno-economic assessment of IES designs and operational strategies. However, the trade-off between model complexity and accuracy presents a significant challenge in comprehensive techno-economic analyses. Simplified models (e.g., levelized cost analysis) often overlook the integration of IES within the broader electric grid context, while rigorous models (e.g., production cost modeling) incorporating grid constraints are computationally expensive. Therefore, it is crucial to develop optimization models that improve accuracy while maintaining computational efficiency.

One strategy that sits in the middle of both model complexity and accuracy is the Price-taker (PT) [2, 3] assumption, which is a widely used optimization model to perform IES optimization. PT simplifies the dynamic and complicated wholesale electricity market as price signals. It approximates the electricity market as an infinity bus, which can take and give any amount of electricity without impacting the prices. In other words, electricity prices are considered as exogenous inputs to the optimization model and assume operation decisions of the IES won’t impact the market prices. Thus, while the PT model does consider dynamic electricity prices (more accurate than levelized cost models), it does not consider the impact of participating within the market (less accurate than production cost modeling).

To this point, previous literature [4-6] demonstrates that this PT assumption may lead to inaccurate estimations of IES economics. An alternative optimization tool is the production cost model (PCM). PCM mimics the wholesale electricity market by solving unit commitment (UC) [7] and economic dispatch [8] problems with a rolling horizon. PCM includes electric grid operation constraints, which is a more rigorous simulation than PT. However, PCM is computationally expensive due to the integer variables in the UC, and it needs the modeling information of the entire electric grid, including all generator characteristics, renewable generation, electricity demand and transmission lines. PT focuses on the process modeling and has much lower model complexity than PCM. Therefore, improving the PT formulation to include the impact of market participation while retaining less computational complexity than PCM bridges the gap between the medium and high accuracy solution strategies for IES optimization.

In this work, we propose an improved PT optimization framework, which increases the model accuracy while maintaining simplicity. A data-driven surrogate model is trained and embedded in the PT optimization to capture the impact of IES over electricity prices. The framework is organized as follows:

1. Perform PCM simulations of different IES designs
2. Use PCM simulation results to train data-driven surrogate models that map the IES design to the market behavior
3. Build and solve the IES optimization with data-driven surrogate models
4. Verify the optimization results using PCM (optional)

The surrogate-assisted optimization is applied to a nuclear power plant case study, where the nuclear power plant is retrofitted with a proton exchange membrane (PEM) electrolyzer to co-produce hydrogen. PEM electrolyzer provides operational flexibility for a baseload nuclear power plant to ramp down according to fluctuating market prices. We show the different design (PEM electrolyzer power) and operation (bidding price) decisions of the nuclear-PEM IES impact the market prices, and the data-driven surrogate model allows the price taker formulation to more accurately account for participation impact.

Acknowledgement and Disclaimer

This work was conducted as part of the Institute for the Design of Advanced Energy Systems (IDAES) with support from the U.S. Department of Energy’s Office of Fossil Energy and Carbon Management (FECM) through the Simulation-based Engineering Research Program.

This project was funded by the Department of Energy, National Energy Technology Laboratory an agency of the United States Government, through a support contract. Neither the United States Government nor any agency thereof, nor any of its employees, nor the support contractor, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

Reference

[1] Arent, Douglas J., Shannon M. Bragg-Sitton, David C. Miller, Thomas J. Tarka, Jill A. Engel-Cox, Richard D. Boardman, Peter C. Balash, Mark F. Ruth, Jordan Cox, and David J. Garfield. "Multi-input, multi-output hybrid energy systems." Joule 5, no. 1 (2021): 47-58.

[2] Laky, Daniel J., Nicole P. Cortes, John C. Eslick, Alexander A. Noring, Naresh Susarla, Chinedu Okoli, Miguel A. Zamarripa et al. "Market optimization and technoeconomic analysis of hydrogen-electricity coproduction systems." Energy & Environmental Science 17, no. 24 (2024): 9509-9525.

[3] Laky, Daniel, Gooty, Radhakrishna, Jaffe, Tyler, Holly, Marcus, Atia, Adam, Chen, Xinhe, & Dowling, Alexander (2024). IDAES-PSE Software Tools for Optimizing Energy Systems and Market Interactions. https://doi.org/10.69997/sct.137860

[4] Gao, Xian, Bernard Knueven, John D. Siirola, David C. Miller, and Alexander W. Dowling. "Multiscale simulation of integrated energy system and electricity market interactions." Applied Energy 316 (2022): 119017.

[5] Jalving, Jordan, Jaffer Ghouse, Nicole Cortes, Xian Gao, Bernard Knueven, Damian Agi, Shawn Martin et al. "Beyond price taker: Conceptual design and optimization of integrated energy systems using machine learning market surrogates." Applied Energy 351 (2023): 121767.

[6] Emmanuel, Michael Ikechi, and Paul Denholm. "A market feedback framework for improved estimates of the arbitrage value of energy storage using price-taker models." Applied Energy 310 (2022): 118250.

[7] Knueven, Bernard, James Ostrowski, and Jean-Paul Watson. "On mixed-integer programming formulations for the unit commitment problem." INFORMS Journal on Computing 32, no. 4 (2020): 857-876.

[8] Mahor, Amita, Vishnu Prasad, and Saroj Rangnekar. "Economic dispatch using particle swarm optimization: A review." Renewable and Sustainable Energy Reviews 13, no. 8 (2009): 2134-2141.