Unit commitment deals with unit generation scheduling in a power system: the determination of the start-up and shut-down times of thermal units towards minimizing operating cost and satisfying various constraints, such as load balance, spinning reserve margin and reactive power flow. [2]  This problem involves a large number of logic type variables, such as the up/down status of the unit, as well as continuous variables expressing generation output, and is therefore a very difficult optimization problem to solve. The present study is aimed at finding a sub-optimum solution efficiently.

 

The unit commitment method that has been widely used solves the problem without consideration of voltage, reactive power and transmission constraints. With the unit commitment schedule, generating companies satisfy customer load demands and maintain transmission flows and bus voltages within their permissible limits. However, for power flow with heavily loaded lines and transformers located far from loads, transmission flow limits and voltage drops throughout the system may become troublesome. 

 

Unit generation, phase shifters and transformer tap controls can be used to control power distribution, alleviate transmission overflows and improve voltage profile in the system for a given unit commitment. However, as the system becomes more constrained, generation level adjustments by these controls may deviate significantly from the predetermined constrained economic dispatch, [3], and [4]. 

 

Furthermore, if control devices cannot alleviate violations, optimal power flow or constrained economic dispatch based on unit commitment will have no solutions owing to excessive transmission flows or violation of voltage constraints.

 

In this presentation, a solution is proposed to the unit commitment problem with voltage and transmission constraints, based on a heuristic method and Optimal Power Flow (OPF), [5], and [6]. The proposed algorithm can handle network constraints such as line flow, voltage magnitude and reactive power limits, easily.

 

In this method, initial unit status is determined from random numbers and the feasibility is checked for minimum start-up/shut-down time and demand-generation balance. If the solution is infeasible, the initial solution will be regenerated until a feasible solution can be found. Next, OPF is applied for each time period with the temporary unit status. Then, we find the units to can shut down to have lees cost, in the most power systems units are designate for this purpose when the energy is sufficient. Note that the criteria to shut down units do not necessarily include those that have less energy contribution to the system, as example is the nuclear plant which is rarely shut down, regardless of their energy contribution to the system.

 

For to identify risk or constraints violations on the system is necessary verify the following:

 

1. Each generator is operating in a range greater than a specified minimum power, and less than its maximum power design limit. 

Pmin i ≤ Pi ≤ Pmax i  (1)

 

2. The node voltages are within the range established by a voltage stability study. 

|V|min i ≤ |V|i ≤ |V|max i  (2)

 

3. The reactive power at each node is limited between minimum and maximum values:

Qmin i ≤ Q i ≤ Qmax i   (3)

 

4. The apparent power flow on each of the transmission lines is limited to a maximum value: 

Si ≤ Smax i  (4)

 

At the same time, if there are constraint violations, appropriate units will be started-up to alleviate these violations. For example, if reactive power flow constraints are violated in any node, the author of Unit Commitment must change the proposal for release of generation units or only settings in devices to controlling reactive power. This may include FACTS devices, such as SVCS, or some switch capacitors. This process will be repeated until a sub-optimal solution is obtained. 

 

Since the authors have implemented a heuristic design, as with Monte Carlo design, the large number of random contingencies mandates a very fast simulator, such as the real-time simulator of OPAL–RT®. 

 

A section of the Mexican Power System has been used for their study. This study includes eight equivalents of generation units in eight different nodes; the system is formed by forty-three transmission lines of 400kV and sixteen nodes have load installed. 

 

Fig. 1 shows the configuration of the system with the solution of power flow in the initial condition. Our initial condition considers a power system with all devices in operation, and the minimum value from the shot-term forecast study which was made considering a summer day. The load was chosen as starting point and all day variations. 

Fig. 2 shows active power behavior for load nodes. Notice the dimension that refers to variations at 24 hours, so it is visible the load consumed on every node through the day. Fig. 3 shows nodal voltages when the charge is incremented as presented in Fig. 2. This visual procedure is part of the strategy to detect risky conditions and for implementing control actions. By the moment, power load behavior was defined and basic maneuvers are integrated, separating effect of contingencies.

 

The implementation in Real–Time is prepared to have a random list of contingences to apply in nodes, lines and generators. These contingences provoke topology changes and power flow changes in the system. The presentation also includes numerical results of the algorithm for monitoring security in real time under different contingencies applied to the test system.

See more about Electrical and Power Systems

 

[1] Gabriela Hug-Glanzmann, Göran Anderson, “Decentralized Optimal Power Flow Control for Overlapping Areas in Power Systems”, IEEE Transactions on Power Systems, Vol. 24, No. 1, February 2009.

[2] Haili Ma, S. M. Shahidehpour, “Unit Commitment with transmission security and voltage constrains”, IEEE Transactions on Power Systems, Vol.14, No. 2, May 1999.

[3] Farrokh Aminifar, Mahmud Fotuhi-Firuzabad, Mohammad Shahidehpour, “Unit Commitment with probabilistic spinning reserve and interruptible load considerations”, IEEE Transactions on Power Apparatus and Systems, Vol.24, No 1, February 2009

[4] Federico Milano, Claudio A. Cañizares and Antonio J. Conejo, “Sensitivity-Based Security-Constrained OPF Market Clearing Model”, IEEE Transactions On Power Systems, Vol. 20, No. 4, November 2005.

[5] Hermann W. Dommel, William F. Tinney, “Optimal Power Flow Solutions”, IEEE Transactions on Power Apparatus and Systems, Vol. Pas-87, No 10, October 1968.

[6] F. J. Nogales, F. J. Prieto, and A. J. Conejo, “A decomposition methodology applied to the multi-area optimal power flow problem,” Ann. Oper. Res., no. 120, pp. 99–116, 2003.