Lucas Haas

Combined-cycle gas turbine (CCGT) power plants play an important role in reducing greenhouse gas emissions while maintaining high thermal efficiency. Accurate thermodynamic models are essential for evaluating plant performance under different operating conditions and supporting improvements in energy resource management. This study presents a Python-based thermodynamic model of an isolated natural gas-fired CCGT power plant, applied to 1027 steady-state operating points recorded over one month of operation. The model integrates energy, exergy, and exergoeconomic analyses based on the Specific Exergy Costing (SPECO) methodology to assess plant operation with and without supplementary firing. The model was validated against the measured steam production, with a mean absolute percentage error of 2.8%. The proposed methodology enables the evaluation of component efficiencies, exergy destruction, cost formation, and the influence of different operating modes on electricity production. The plant achieved median gross first- and second-law efficiencies of 46.6% and 44.6%, respectively (45.1% and 43.2% on a net basis). The combustion chamber exhibited the highest exergy destruction in both operating modes, while the supplementary burner became one of the major sources of irreversibility when supplementary firing was activated. The specific exergy cost of electricity decreased from about 21.0 to 14.2 $/GJ as the electrical output increased from 11.3 to 20.9 MW, with a median of 16.47 $/GJ in the period. The supplementary burner presented the highest relative cost difference (median of 1.06), whereas the combustion chamber had the highest exergy destruction rate and cost rate of exergy destruction. The proposed model provides a flexible and reproducible framework for evaluating CCGT power plants and can support operational assessment and decision-making aimed at improving thermodynamic and economic performance.

The recent efforts to increase renewable-based generation in power systems have also prompted the integration of Battery Energy Storage Systems (BESSs). In deregulated electricity markets, BESS investments are increasingly driven by private, profit-oriented agents. Although existing literature offers optimization-based models for BESS allocation within profit-based frameworks, these models often overlook the strategic decision of when to invest from the perspective of a price-maker aggregator. Optimizing this timing requires balancing future declines in BESS capital costs against the time value of money and the opportunity costs of foregone short-term profits. Furthermore, accurately representing electricity price behavior over the planning horizon is vital, especially when energy arbitrage is the primary revenue source. Traditional clustering techniques, such as K-means, often yield over-conservative representative prices, leading agents to avoid investing. To address these gaps, this paper proposes a multi-year bilevel optimization framework to jointly determine the optimal siting, sizing, and timing of BESS investments. The problem is formulated as a Stackelberg game where the leader (BESS aggregator) maximizes its Net Present Value (NPV) over a 20-year horizon, while the follower (Distribution System Operator - DSO) minimizes network operational costs. The model is reformulated as a mathematical program with equilibrium constraints (MPEC), linearized, and solved via mixed-integer linear programming (MILP). Additionally, a K-medoids clustering approach is applied to preserve historical wholesale price volatility. Numerical results on a 34-node test system demonstrate that the framework effectively captures the dynamic economic trade-offs of investment timing. By leveraging declining technology costs and rising electricity prices, the aggregator postpones investments by 6 years to achieve a cumulative cash flow more than 21 times higher than under traditional first-year investment approaches.

The efficient operation, monitoring, and maintenance of a photovoltaic (PV) plant are intrinsically linked to data accessibility and reliability, which, in turn, rely on the robustness of the communication system. As new technologies arise and newer equipment is integrated into the PV plants, such as the battery energy storage system (BESS) that transform the PV plant into a dispatchable plant and the all-sky camera (ASC) that enables the prediction of shading events, the communication system faces new challenges, that are described in this work. In this paper, two communication systems were developed using only open-source software, in which the first was designed for seamless communication between the PV and BESS equipment, while the second was tailored to collect photos from ASC devices. The requirements of the communication systems were defined based on the applications that control the PV plant, and on the industry-standard IEC-61724-1 norm for PV data. After being developed, the communication systems were installed in a PV plant, and the interaction between the data obtained from these two systems is discussed and presented.

The efficiency of a photovoltaic module decreases with the increase in the temperature of the photovoltaic cell, which makes the determination of this temperature one of the determining factors in the modeling of a photovoltaic module. The objective of this work is to make a detailed study about the input parameters of a developed thermal-electrical model of a photovoltaic module, based on a sensitivity analysis. The model includes a thermal part, in which the temperature varies along the five layers of the module, and an electrical part, which is based on a single diode five parameters model. The coupling of these models into a thermal-electrical model is described. Through differential sensitivity analysis, the parameters that have a significant impact on estimating the photovoltaic cell temperature were determined. The variations of the sensitivity coefficients throughout a simulated day are presented. Parameters with the highest absolute sensitivity coefficients, such as the transmittance-absorptance product and open circuit voltage, are crucial for accurate temperature estimation and were further researched and determined with greater precision, in order to improve the model reliability. Conversely, parameters with lower absolute sensitivity coefficients, such as albedo and parallel resistance, were considered less important and could be simplified to enhance computational efficiency.