Estimating the Actual Cost of Transmission System Congestion

Abstract In its simplest form determination of the cost of transmission system congestion is rather straightforward. Assuming complete knowledge of the power system inputs, such as the loads at all system buses, and that the generators bid in their actual costs (i.e., they are not taking advantage of localized market power), then the hourly short-term cost of transmission congestion is the difference between the hourly cost of an optimal power flow (OPF) solution that includes transmission constraints and one that does not. If contingency constraints are to be included then transmission congestion is the difference between the security constrained OPF (SCOPF) solution and the unconstrained OPF. In the absence of energyconstrained generation (such as hydro) or generator ramping/shutdown/startup constraints, then congestion costs could be calculated by repeating the above analysis for each hour of the year. This paper describes a methodology that could be used by a utility to estimate the actual cost of congestion on its transmission system using limited, non-state estimator data. The assumed problem inputs are a power flow model of an entire interconnected grid (i.e., the Eastern Interconnect), costs for the utility’s generators, and then hourly values of the utility’s generation, load and tie-line flows over the study time period. Due to the common lack by most utilities of external measurements, the system is first equivalenced to retain only the utility’s own internal buses and a small subset of the external buses. Then, for each hour, the utility’s load and generation is set to match their historical values, while the external generation is adjusted to match the tie-line flows. Next, an economic dispatch is performed to determine the unconstrained cost. Finally, a security constrained OPF (SCOPF) is solved to take into account base case and contingent constraints. The methodology uses a complete ac power flow formulation to accurately estimate the impact of voltage constraints and the incremental impact of system losses. The inclusion of hydro generation is also considered. For illustrative purposes only, the methodology demonstrated on the TVA system using publicly available data transmission system data. A simple example of this approach is shown in Figures 1 and 2 using a three bus system that contains lossless lines with equal impedance and equal MVA limits, here set to 100 MVA. With a single 180 MW load at bus 3, a generator at bus 1 with a cost of $10/MWh and a generator at bus 2 with a cost of $12/MWh, the unconstrained and constrained dispatches are as shown in the figures. The hourly cost of transmission system congestion is then $120/hr, while the marginal cost of enforcing the bus 1 to bus 3 line constraint is $6/hr/MVA.