PGU-450 network heaters modes modelling with CCGT participation in regulating the electrical load in the heating mode

The features of a mathematical model for optimizing the distribution of heat and electricity at a large thermal power plant with a complex composition of equipment as part of traditional heating units and a heating CCGT are considered. The selection and justification of optimization criteria at different stages of preparation and entry of the station to the electricity and capacity market is given. The disadvantages of the previously proposed optimal distribution algorithms are analyzed in relation to thermal power plants with a complex composition of equipment and with a complex scheme for the supply of electricity and heat. A method and algorithm for solving the problem are proposed based on the equivalence of the CHP equipment and the decomposition of the problem taking into account the schemes of electricity and heat output. The description of mathematical optimization methods is given, taking into account the peculiarities of the CCGT operating modes at reduced loads. The requirements for information support when integrating the developed algorithm into the application software of the automated process control system based on the PTC are given.


Introduction
The transition of the national economy to market relations has led to certain structural changes in the electric power industry. Thus, there was a decrease in the share of industrial consumers with 2-3 shift modes of operation and a significant increase in the share of consumers with sharply alternating modes of power consumption. The absence of highly manoeuvrable units required for these modes during the passage of power consumption schedules dips led to forced deep unloading or shutdown of equipment designed for the basic operating mode. Statistics of operational modes of power plants show that even powerful power units of 300, 500 and 800 MWt for supercritical steam parameters on a number of power systems during the passage of night load dips are daily unloaded up to 30-50% of the installed capacity. The problem of heating power plants is even more complicated, because during the heating period, the passage of night dips in power consumption, as a rule, coincides with the maximum of the heat consumption schedule. Thus, at the power plants of JSC "Mosenergo", the thermal load in the winter period at night is 80-100% of the installed thermal capacity at 30-50% of the installed electric capacity. The problem is also complicated by the fact that during the hours of dips due to the low cost 1 edik_arakelyan@inbox.ru 3. Approaches to solving the problem There are two approaches to solving this problem: 1. Complete calculation of the thermal scheme of the CCGT power unit each time with actual (increased) underheating in network heaters and the use of the results obtained during optimization calculations. The disadvantage of this approach is that, as a rule, for its implementation, in addition to the software product, a significant amount of measured and calculated initial data is required (efficiency and parameters of the flow part of turbines, the value of steam pressure in the selections for network heaters, real values of underheating in network heaters, etc.), which require verification and refinement taking into account measurement errors.
2. The use of digital models of the power unit and the heating plant in the form of regression dependencies obtained on the basis of experimental and computational studies. Methodological approaches and algorithms for the implementation of the second approach will be considered on the example of a combined-cycle power unit PGU-450T, a simplified scheme of which is shown in Fig.1. The thermal scheme of the PGU-450T is designed in such a way as to provide any combination of electrical and thermal loads from the technical range of electrical loads and from the maximum value of the thermal load to its complete absence (condensation mode). Let's consider the heating mode of operation of a steam turbine with partial discharge of steam into the turbine condenser. The digital model of the CCGT is a set of regression equations showing the dependence of the output parameters of gas turbines, the heat recovery boiler, the steam turbine and the unit as a whole on the outdoor air temperature, the required electrical and thermal loads, etc. [6]. Thus, the dependence of fuel consumption on these parameters is represented as: Where B f -fuel consumption, t a -outdoor air temperature , N -electrical load of the unit, Q -thermal load of the unit, k mod -a set of parameters that determine the operating mode of the CCGT. With this representation of the consumption characteristics of the CCGT, the optimization problem can be solved directly by increasing the dimension of the problem with the inclusion of discrete parameters (mode number), but this significantly complicates the optimization algorithm and does not always lead to high efficiency of the solution.
In a simplified form, the operation of a unit with one or two gas turbines can be considered as possible modes of operation of the CCGT, since this feature has the greatest impact on the performance of the CCGT. Then the consumption characteristic will take the form: where n gt is the number of working gas turbines. This circumstance significantly simplifies the solution of the optimization problem at the stage of choosing the composition of the working equipment, since the matrix of possible operating modes of the power equipment of the station is significantly simplified. When considering the operation of the CCGT in the condensation mode, the consumption characteristic will take the form: In this case, the solution can also be obtained by the method described above for a simplified version of accounting for the operating modes of the CCGT.

The initial stage of solving the task
The approximation of the calculated data was performed in the MathCad software for the operation of the unit at a fixed outdoor temperature -2 o C. When approximating the calculated data with a second-order polynomial, the maximum deviations of the approximating consumption characteristics from the calculated data used do not exceed 1% for the fuel consumption of the CCGT over the entire range of load changes [7]. When the PGU-450T unit operates with one gas turbine, the characteristic has the form: .
When operating the installation with two gas turbines, the flow characteristic has the form: In (4) and (5) ) , ( When choosing the operating modes of the heating plant at the stage of selecting the composition of the generating equipment 2-4 days before the operational ones and when carrying out the optimal distribution of heat and electric loads when preparing the station's proposals for entering the market "a day ahead", it is necessary to evaluate the efficiency of the heating unit (turbine) when using one -or two-stage heating of mains water within one turbine. For this purpose, you can use a digital model of a heating plant, which is a set of material and balance equations and the dependence of fuel consumption per unit (or heat consumption per turbine) for single-stage (T1 mode) and two-stage (T2 mode) heating of mains water, taking into account the actual preheating of mains water in them: (7) where a 0 , ..., a 5 , b 0 , ..., b 5 are the coefficients of the regression equation (5) When the CCGT operates in the mode of passing the load dip according to the electric load, the minimum power of the steam turbine can serve as a criterion under the same conditions. The following algorithm for solving the problem is proposed: -for the daytime operation of the CCGT:  when Nunit, Qunit set, for one of the modes, for example, for T2, taking into account the actual underheating in the network heaters, the fuel consumption per unit is calculated:  under the conditions (9) It is not difficult to obtain from equation (6) (10)  If , then the T2 mode is effective and vice versa. The proposed algorithm is based on the predicted values of thermal and electrical loads and is therefore recommended as the initial stage of solving the problem.

Operational optimization algorithm
When performing operational optimization during the execution of the dispatching schedule of loads, a simpler, but more accurate approach is proposed. Its algorithm includes the following steps: 1. In the steady-state operation mode of the heating plant at the specified conditions unit N , unit Q , the actual underheating in the network heaters is calculated as the difference between the saturation temperature of the selected steam by its pressure in front of the heater and the temperature of the network water behind the corresponding network heater; 2. When the heating plant is operating in the mode of two-stage heating of mains water at specified unit N , unit Q , the fuel consumption for the CCGT is calculated by the expression (4) when the CCGT is operating with one gas turbine or (5) when working with two gas turbines; 3. The network installation is switched to the mode of single-stage heating of the network water and the parameters of the selected steam are set to provide a given thermal power, for which, taking into account the underheating calculated in stage 1, the temperature and pressure of the selected steam are determined in front of the network heater and at the outlet of the steam turbine, the power of the steam turbine and the CCGT as a whole are fixed.
4. The fuel consumption is calculated as in stage 2 at unit Q the new power of the CCGT, compared with the result of stage 2. 5. If the B CCGTstage4 < B CCGTstage2 , then the single-stage heating mode is more effective and vice versa. It should be noted that the real value of underheating in network heaters, as in all steam-water heaters, in addition to the noted dependence on time, also depends on the value of the current water flow, in this case, on the flow of network water. Such a dependence can be obtained experimentally. In the absence of experimental data, we can use the empirical dependence [8]: -we calculate the fuel consumption according to the dependence (7): (17) To simplify the calculation procedure, the calculation can be carried out using the expression: which, for a dependence in the form of a second-degree polynomial, is written as: (19) The disadvantage of this method is that the regression dependences of fuel consumption per unit in the form of polynomials do not always reflect the real state of the unit's equipment and require constant refinement.