THE INTEGRATION OF SOLAR GENERATION ON A POWER SYSTEM: OPERATIONAL AND ECONOMIC EVALUATION
In order to distinguish essays and pre-prints from academic theses, we have a separate category. These are often much longer text based documents than a paper.
In recent years, the accelerated deployment of renewable electricity generation resources, especially wind and photovoltaic (PV) solar, has added challenges to the operation and planning of the power grid. One of the challenges is that the variability of solar and wind power output may increase the variation of the load that must be followed by dispatchable resources and increase the ramping capacity needs. Moreover, the decision about the configuration of a PV solar generation systems has operational and economic implications because peak solar energy production does not always precisely occur when the wholesale electricity prices of the system are highest. Therefore, as the renewable capacity levels grow, it becomes increasingly important to examine the potential impacts on the system cost and portfolio of conventional generating units to respond to the intermittent nature of some renewable generation technologies. Three related analyses explored in this dissertation address some of the challenges of integrating utility-scale PV solar and wind projects into a power system using a case study for Indiana.
The first analysis identifies the optimal azimuth and tilt angles of solar PV installations that alternatively maximize the annual electricity generation or the economic value of the resource. The economic implications of the configuration of solar PV installations within Indiana are estimated based on wholesale prices of electricity and simulated solar output for different combinations of angles and types of array installations. The results show that solar projects across the state would need to have azimuth angles within the 177 and 182 degrees range to obtain maximum annual energy and 180 to 190.5 degrees to maximize annual value, independently of their array types. Furthermore, southern and northwestern zones showed the highest impacts from using an optimal angle configuration of the solar installations. Nevertheless, on average, the benefits in annual electricity generated or economic value from their reconfiguration across the state are minor, amounting to less than one percent.
The second analysis explores the effects of additional solar and wind power investments on the 2035 requirements for baseload and peaking generation capacity, the amount of energy supplied by various types of generation technologies and the costs of Indiana’s electric supply system. From a capacity planning and unit commitment/dispatch perspective, the results of this analysis indicated that with a portfolio that includes more solar and/or wind power generation, there would be need to add new peaking generation units. However, the total need for additional peaking resources declines as more renewables are added to the generation mix. Because Indiana still heavily relies on coal and other baseload resources to generate electricity, no new baseload capacity is required in the future. Generally, additions of PV solar and wind capacity amplify the variation in load net of renewable generation and create greater needs for ramping services from conventional units. However, results of the analysis show that the existing portfolio of conventional generation resources in Indiana would have sufficient operational flexibility to be able to accommodate ramping requirements even with PV solar and wind capacity penetration levels as high as 30% of total electricity generation. However, at those levels of renewables capacity there are a times during the year when the optimal operational strategy is to curtail solar and wind generation. From a technical perspective, the results indicated that larger thermal generating units are used more for load following and turned on and off (cycled) more frequently with the additional renewables than without them but mainly during days with low levels of demand and high levels of generation from renewable technologies. From the cost perspective, the results of the model support the idea that it would be cheaper in the long-term to invest in a combination of solar and wind generation resources than in solar generation resources alone. Moreover, the reductions in variable costs, driven by the zero variable cost added to the system by the additional solar and wind capacity, were not sufficient to outweigh the increases in capital costs regardless of the levels of capacity additions.
For the third analysis, the proposed capacity expansion model was used to estimate the value of capacity of PV solar and PV solar in combination with wind capacity in terms of baseload/peaking resources from a deterministic system peak load reliability perspective and for various penetration levels of these resources. The capacity values of solar, which refer to the contribution of PV solar plants to reliably meeting the system peak demand, for all the wind capacity levels analyzed, fall as the amount of solar capacity increases. This is because as solar generation increases and closely coincides with the occurrence of the system peak load, there is a shift of the peak load net of renewable generation time to later afternoon hours, when solar installations begin to reduce their production, therefore decreasing their contribution to reliably meeting system peak demand. The calculated solar capacity values are between 2.7% and 67.3% of the corresponding solar nameplate capacity considering all zones and types of PV solar arrays in Indiana, and vary with the level of solar penetration. The range of values obtained are in line with the ones found in other studies using stochastic reliability-based methods.
This dissertation contributes to the literature on the interaction between PV solar with other generation resources and to their economic, operational and policy implications. Furthermore, it provides another decision-making tool from a planning perspective for policymakers, utility companies and project developers.