This paper was presented at the EIC Climate Change Technology Conference 2013. The introduction and conclusions are excerpted below. The full paper can be accessed by clicking on the link(s) at the bottom of this page.
Increasing wind penetrations can necessitate the use of fast ramping generation to manage the variability associated with wind energy production. In some jurisdictions system operators utilize contingency reserves -- generating assets standing ready and able to provide short-term electricity to the grid. These generators must be synchronized to the grid and operating at less than their nominal capability. For Alberta’s fossil fuel generators this means that the unit is operating at less than full efficiency. Strbac et al.  assume a linear relationship between loading levels and CO2 emissions, stating that "coal and [combined cycle] gas units operate less efficiently when part loaded, with an efficiency loss of about 20%" (p. 1217), we assume this applies to conventional coal and natural gas generating units. As wind capacity is added into the system, contingency reserve provision must increase. In this paper, the net CO2 implications of higher wind penetrations in a thermally dependent electric system are estimated using Alberta as a case study.
As a participant in the Northwest Power Pool, the Alberta Electric System Operator (AESO) is required to carry a minimum contingency reserve of 5% of its firm load responsibility served by hydro and wind resources, plus 7% served by thermal resources. On an average day, the system operator procures approximately 250 MW of spinning and supplemental reserves for balancing the system each hour of the day and maintains these levels of reserves throughout each hour . As more reserves are dispatched to balance the system the operator activates additional standby reserves to maintain the required reserve capability. These reserves are
assumed to be part-loaded and produce 0.72 tC02/MWh. Once called upon the generators operate at closer to their full-efficiency levels and produce 0.60 tCO2/MWh. For this study a simulation model was developed that replicates both the unit commitment and economic dispatch decisions of the system operator for various wind capacity levels and supply situations. Employing a stochastic wind generating process, CO2 emissions from energy dispatch are estimated as well as the additional CO2 emissions associated with reserve production to determine the total CO2 emissions from the grid given two wind capacity levels and three differing supply states: normal, shortage and surplus.
Alberta has integrated nearly 1.1 GW of wind into its existing, thermally dependent and deregulated electric grid. The effect of this wind on overall CO2 emissions appears to be negligible. By far, the largest contributor to Alberta’ CO2 emissions is from coal and gas generating units dispatched from the energy market merit order. The impacts of wind generation on overall emissions are marginal at best, and are the result of the non-dispatchable nature of wind generation. Further CO2 emissions reductions may be achieved through increased efficiencies of thermal generators, demand-side reductions, imports from British Columbia’s hydroelectric system or the development of nuclear facilities that could provide emissions-free, baseload generation. Increasing the frequency of intertie scheduling could allow British Columbia’s fast-ramping hydro generation to act as a ramping service to mitigate wind variability.