Date of Award

Spring 4-2016

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Engineering and Public Policy


Paulina Jaramilo


This thesis analyzes two potential means of mitigating the cost increase of ancillary services that is expected with the decarbonization of the U.S. electricity network. The first method, balancing area consolidation, addresses this cost rise by reducing the demand for ancillary services. This research quantifies the economic benefit of consolidation in the frequency regulation market by estimating the resulting reductions in frequency regulation requirements and cost. The results show that this policy leads to a reduction in frequency regulation cost of approximately $0.1 per MWh of total load. These results do not significantly change with the inclusion of 20% wind, suggesting that in the near term, wind’s interaction in the frequency regulation market is not a prime motivation for consolidation. This analysis does not consider all the benefits or costs of BA consolidation, and is not meant as an assessment of net-benefits. Though the results show consolidation could lead to an increase in emissions of some air pollutants, which suggest that there may be significant trade-offs associated with the decision to consolidate balancing areas. The second means of addressing the expected increase in ancillary services costs is to increase the supply of ancillary services by leveraging residential demand response. We developed methods that optimally schedule ancillary service capacity on demand response resources while accounting for the risk of customer response fatigue. The model is used to test the efficacy of hourly caps on demand response penetration in ancillary service markets. The results show that residential demand response could provide a significant portion of the total ancillary service requirements attributable to residential loads: between 50% and 75%. Hourly caps on demand response participation are shown to be economically inefficient. With a 25% market cap, residential demand response is scheduled to 6 provide 25% of the hourly total market value, while the risked-based optimization schedules residential demand response to provide 82% of the total value. Methods like the ones presented in this paper, that can appropriately weight the benefits and risks of committing residential demand response will be critical to efficiently and effectively use this resource for ancillary services.