Alberto J. Lamadrid
Asst. Prof. Economics
with Tim Mount, Ray Zimmerman and Daniel Munoz-Alvarez
The electricity industry landscape in the 1980s was in what can be categorized as the calm before the storm: the changes seen in other industries, most notably airlines, were still making ripples across the regulatory spectrum, and what could be considered a mature industry after close to 100 years since the Pearl Street Station, was the object of study and intense. It is from this time that the conceptual re-imagining of the way this industry operates dates, integration engineering and economic solutions to suggest changes in the way in which system operators and regulators balance supply availability and electricity demand for planning purposes [Schweppe et al.(1989)], [Gellings and Smith(1989)]. It is now hard to conceive a situation in which the planning of the bulk electricity system does not consider some form of demand response, and both researchers and practitioners have dedicated some attention to this topic (see e.g. [Braithwait(2010)], and the full issue of the Power and Energy Magazine dedicated to Demand Response).
Our research integrates two burgeoning fields with symbiotic characteristics: demand response, and the adoption of renewable energy sources (RES) into the electricity system. This integration provides the opportunity to reevaluate the fundamental ways in which one of the most complex engineered systems is managed. With higher RES shares, the dispatching ability of the generation resources is no longer applicable. This lack of control on dispatch for RES is due to the dependency of these resources on atmospheric conditions. However, the availability of energy storage systems (ESS) and the procurement of thermal demand services provide physical mechanisms to modify both the supply and demand of energy services.
In our framework we start with what is known as an Optimal Power Flow (OPF) [Carpentier(1962)], whose objective is to minimize a function (e.g. losses, costs), subject to the network equations obeying Kirchoff’s voltage and current laws for an AC flow. The uncertainty in the state of nature at the moment of dispatch prompted the development of models that incorporated varying inputs into the system (e.g. [Wang and Shahidehpour(1995)]). We then integrate this with the co-optimization (co-opt) of energy and different types of reserves needed for secure operation of the system ([Ela et al.(2011)], [Chen et al.(2005)], [Thomas et al.(2008)]). The reliability in the system is preserved by specifying appropriate probabilities for contingencies, and procuring them at the individual generator level. Besides generators, other sources can also provide reserves [Sioshansi and Denholm(2010)].
We apply this framework to a reduced version of the Northeastern Power Coordinating Council [Allen et al.(2008)], to study the effect of close to 25% Wind penetration [NREL(2010)], and the deployment of controllable demand as an specific demand resource available to the System Operator (SO). The demand considered controllable is determined using an econometric analysis of demand for New York and New England (NYNE) [Mo(2011)], calculating the proportions of the temperature sensitive demand (TSD) in for these two regions. TSD is that portion of the total demand used to provide thermo services, such as heating, ventilation and air conditioning (HVAC). Thanks to technologies like ice batteries [Dorgan(2000)], the time of electricity purchase can be decoupled from the time of service delivery for the energy required to maintain a certain comfort level.
We contrast controllable (deferrable) demand with utility-scale storage, placed in the same buses where the wind farms are, to analyze the effect of location and the effects of congestion in the system due to creation of load pockets [Lesieutre et al.(2005)]. In total we simulate 6 different policy case studies, for a representative day with a relatively high peak system load. Each one of the cases is evaluated using three main measures of performance: (1) the out-of-pocket operating costs of the the conventional generators. (2), the amount of wind generation dispatched and (3) the maximum conventional generation capacity needed to cover the peak demand and maintain system reliability.
We provide some insights about the economic viability of some of the most commonly discussed options for a constrained network as the northeastern grid of the United States. Using demand side mechanisms has the potential to reduce the overall costs of the electricity system, and at the same time reduce the cost of capital due to the requirements for conventional capacity. A tacit function played by the controllable demand is the change in demand profiles, decreasing the need for ancillary services such as load following reserve.
Regulators often favor upgrades of the transmission infrastructure, as these allow operating the system closer to traditional supply-side management paradigms. The results in this study show that in fact, using storage located in the same buses as the wind buses allows reducing the costs of operation to the lowest level given a realistic depiction of the variable resources in the system. However, the congestion that tends to build up at peak times is not eliminated, and the system may end up fragmented, with generators exercising local power in pockets of demand. Transmission upgrades allow having en economy order dispatch. However, the cost of capital of this option is very expensive. While siting new transmission paths is unlikely to happen, the upgrade of existing corridors is a possible option. But even for existing corridors, crossing state lines usually meets with both public and local regulators opposition.
A very important barrier to controllable demand is that the current structure of the retail rates that most customers pay does not reflect the correct economic incentives. To date, local regulators in most states have not shown much initiative in designing more appropriate rate structures or in educating the public in the potential benefits of the smart grid. Unless this situation changes, it seems unlikely that customers will see the economic benefits that they deserve from the smart grid and the utility industry will continue to depend on supply-side solutions for problems and assume that regulators will ensure that customers pay the bill.
[Allen et al.(2008)] Allen, E., Lang, J., Ilic, M., Aug. 2008. A combined equivalenced-electric, economic, and market representation of the northeastern power co- ordinating council u.s. electric power system. Power Systems, IEEE Transactions on 23 (3), 896–907.
[Braithwait(2010)] Braithwait, S., may-june 2010. Behavior modification. Power and Energy Magazine, IEEE 8 (3), 36 –45.
[Carpentier(1962)] Carpentier, J., 1962. Contribution a l’etude du dispatching economique. Bul- letin de la Societe Francaise des Electriciens 3, 431–447.
[Chen et al.(2005)] Chen, J., Mount, T. D., Thorp, J. S., Thomas, R. J., 2005. Location-based scheduling and pricing for energy and reserves: a re- sponsive reserve market proposal. Decis. Support Syst. 40 (3-4), 563–577.
[Dorgan(2000)] Dorgan, C., 2000. Cool storage technology guide. Tech. rep., EPRI, 7601 Ganser Way Madison, Wisconsin 53719-2074.
[Ela et al.(2011)] Ela, E., Milligan, M., Kirby, B., 2011. Operating re- serves and variable generation. Tech. rep., National Renewable Energy Laboratory. URL http://www.nrel.gov/docs/fy11osti/51978.pdf
[Gellings and Smith(1989)] Gellings, C., Smith, W., jun 1989. Integrating demand-side manage- ment into utility planning. Proceedings of the IEEE 77 (6), 908 –918.
[Lesieutre et al.(2005)] Lesieutre, B. C., Thomas, R. J., Mount, T. D., October 2005. Identification of load pockets and market power in electric power sys- tems. Decis. Support Syst. 40, 517–528. URL http://dx.doi.org/10.1016/j.dss.2004.09.010
[Mo(2011)] Mo, J. Y., 2011. Economic analyses of plug-in hybrid electric vechicles, carbon mar- kets and temperature- sensitive loads. Ph.D. thesis, Cornell University.
[NREL(2010)] NREL, January 2010. Eastern wind integration and transmission study: Executive summary and project overview. Tech. rep., EnerNex Corporation, The National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401.
[Schweppe et al.(1989)] Schweppe, F., Daryanian, B., Tabors, R., may 1989. Algorithms for a spot price responding residential load controller. Power Sys- tems, IEEE Transactions on 4 (2), 507 –516.
[Sioshansi and Denholm(2010)] Sioshansi, R., Denholm, P., 2010. The value of plug-in hybrid electric vehicles as grid resources. The Energy Journal 0 (Number 3). URL http://ideas.repec.org/a/aen/journl/2010v31-03-a01.html
[Thomas et al.(2008)] Thomas, R., Murillo-Sanchez, C., Zimmerman, R., July 2008. An advanced security constrained opf that produces correct market-based pricing. In: Power and Energy Society General Meeting - Conversion and De- livery of Electrical Energy in the 21st Century, 2008 IEEE. pp. 1–6.
[Wang and Shahidehpour(1995)] Wang, C., Shahidehpour, S., Feb. 1995. Optimal generation scheduling with ramping costs. Power Systems, IEEE Transactions on 10 (1), 60 –67.
 firstname.lastname@example.org, 451 RBC Hall, Lehigh University, Bethlehem, PA, 18015