JPH02264303A - Optimum operating method for power generating system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is a method for controlling the steam supplied from a boiler and the amount of generated electricity (and the amount of purchased electricity) to a minimum cost in a factory that has an in-house power generation system. The present invention relates to an optimal operating method for a power generation system, which allows the private power generation system to operate optimally.

[Prior Art] Various plants in the petrochemical, steel and paper valve industries, etc., are equipped with in-house power generation systems to supply electricity and steam.

In this way, for factories with private power generation systems, there are two cases in which the power demand is covered only by the private power generation system;
There are cases where private power generation and power purchase are used together, taking into account situations where the private power generation system has to be shut down, high spot demand for electricity, or lower electricity purchase rates at night. be.

Therefore, efficient operation of private power generation systems has become important, and research is being carried out.

In particular, in the case of a system that uses both private power generation and purchased power, operating costs can be significantly reduced by efficiently supplying power through private generation and purchased power.
The question is how to operate the power supply most economically, taking into consideration factors such as load fluctuations on the plant side, fluctuations in power purchase rates depending on the time and day of the week, and the characteristics of the boiler and turbine. ) are being studied. Conventionally, it has been common to implement an optimal operation method based on nonlinear programming.

However, since nonlinear programming takes time to calculate, a method has been proposed that aims to shorten the calculation time (Japanese Patent Publication No. 18922/1982).

In an energy conversion system consisting of multiple boilers and energy conversion devices such as multiple turbines that are driven using the output of these boilers as energy, this method is used to group each type of energy conversion device group together. This is an operating method that uses an equivalent system replaced with a single equivalent device as a model.

[Issue B to be solved] The driving method shown in the above-mentioned Special Publication No. 18922/1982 is as follows:
By collectively replacing a group of energy conversion devices with a single equivalent device, the number of variables is reduced, thereby significantly shortening calculation time.

However, since a plurality of groups of energy conversion devices are replaced with a single device, there is a problem that optimization operation cannot be performed for each energy conversion device individually.

The present invention has been made in view of the above problems, and it is possible to perform the calculations necessary for optimized operation in a short time, and also to perform the optimum operation of individual devices constituting one power generation system. Perform the necessary control, and thereby 1
The purpose of this study is to provide an optimal operation method for a power generation system that enables even more efficient and optimized operation of the power generation system.

[Means for Solving the Problems] In order to solve the above problems, the method of the present invention is applied to optimal operation of a private power generation system.

In power generation systems that have multiple equipment such as boilers and power generation turbines and supply electricity to plants.

At least, data on the amount of steam used in the plant, the amount of electricity used in the plant, and the amount of electricity generated by each turbine are collected, and continuous linear programming and combinations are performed based on the collected data at regular intervals. This is a method of operating a power generation system by calculating optimally the steam and power balance of each turbine based on theory, and controlling the distribution of the amount of steam supplied to each turbine based on the results of this optimal calculation.

In addition, when the method of the present invention is applied to the optimal operation of a system that uses both private power generation and purchased power, in a power generation system that has multiple devices such as boilers and power generation turbines and supplies electricity to the plant, At least, collect data on the amount of steam used by the plant, the amount of electricity used by the plant, the amount of electricity generated by each turbine, and the amount of electricity purchased, and perform continuous linear planning based on the collected data at regular intervals. This is a method of operating a power generation system by performing optimal calculations of the steam and power balance of each turbine using methods and combination theory, and controlling the distribution of the amount of steam supplied to each of the turbines based on the results of this optimal calculation. .

Preferably, in any case, with respect to each of the plurality of turbines, the amount of steam supplied and the amount of electricity generated in the turbines other than the one turbine performing the optimal calculation are fixed within the limits of the previous continuous linear function, This is to perform optimal calculations for the one turbine in question.

Here, continuous linear programming (5uccessive Li
nearPrograg+sing) is an optimization method for static problems in which a nonlinear function is linearly approximated near a search point, and the area of release of the linear programming problem is sequentially moved to search for an optimal point.

This method maintains the characteristics of linear programming, and is more suitable for controlled optimization problems with large models than other nonlinear optimization methods.

[Operation] The present invention distributes the amount of steam to each turbine while keeping the required amount of steam and the amount of electric power constant by optimizing the operation of the power generation system based on continuous linear programming and combinatorial theory. This allows the power generation system to operate at its highest efficiency point.

[Example] An example in which the present invention is applied to an optimal operating method of a power generation system that uses both private power generation and purchased power will be described below.

FIG. 1 shows a block diagram of a model of a power generation system for carrying out the method for optimally operating a power generation system according to the present invention.

In the figure, Bl and B2 are the boilers, G1-H is the high-pressure turbine in the first power generation system, Gl-L is the low-pressure turbine, G2-H is the high-pressure turbine in the second power generation system, and G2-L is the same. A low pressure stage turbine, G3, is a turbine in the third power generation system. In addition, IH and IL are control valves for steam supplied to the high-pressure turbine G1-H and low-pressure turbine Gl-L of the first power generation system, and 2H and 2L are
A control valve for steam supplied to the high-pressure turbine G2-H and low-pressure turbine G2-L in the second power generation system; 3 is a control valve for steam supplied to the turbine G3 in the third power generation system.

4x, 4y, and 4z are lines 5a, 5b, and 6a for supplying steam to plants #1 to #3.

6b is a steam conversion valve installed between steam lines 4x, 4y, and 4z; 7.8 is a steam conversion valve provided between steam lines 4x, 4y, and 4z; and 7.8 is #4. #5 This is the wiring for supplying power to the plant.

9x, 9y, 9z are pressure gauges that measure the amount of steam flowing through the pipes 4x, 4y, 4z and used in plants #l to #3;
11, 12, and 13 are the first power generation system.

The wattmeter 14.15 is #4, which measures the amount of power generation in the second power generation system and the third power generation system. #5 A power meter that measures the power used in the plant, l 6a.

16b is a wattmeter for measuring the amount of electricity purchased from the outside.

20 is an arithmetic processing unit, which has a constraint value avoidance control function, an optimum calculation control function, an electricity purchase MIN control function, and the like.

Further, this arithmetic processing device 20 includes pressure gauges 9x, 9y, 9
Input measurement data from z and wattmeters 11.12°13.14, 15.16, etc., and input each control valve IH, I
Control signals are output to L, 2L, 3, conversion valves 5a, 5b, 6a, 6b, etc.

Next, one embodiment of the method of the present invention will be described based on the flowchart shown in FIG.

Step 1 Collect necessary data.

That is, the arithmetic processing device 20 is a pressure gauge 9x.

The data from 9y and 9z shows the amount of steam used in plants #l to #3, the data from wattmeter 11.12.13 shows the amount of power generated in the first, second and third power generation systems, and wattmeter 14.
．． Plant #4 with data from 15. The amount of power used in #5 and the amount of purchased power are collected using data from the wattmeters 16a and 16b.

Step 2 The arithmetic processing unit 20 examines the status of the control system based on the collected data, and determines whether the operational and equipment conditions are within the constraint range.

Here, the operational conditions refer to control system conditions that enable optimal control of steam and power balance using continuous linear programming and combinatorial theory, and the method of the present invention cannot be performed unless these conditions are within the constraint region. There will be no. Therefore, the collected data and the constraint values (upper and lower limits) are compared, and if the operational conditions are within the constraint area, proceed to the next step (step 4), and if they are outside the constraint area, the operational conditions are set within the constraint area. To restore the value, use the constraint value avoidance control (step 3) described below.
Do this.

In addition to the above, the operational conditions also include conditions for enabling control at the distributed instrumentation level, and when these conditions are not within the constraint range, operation is switched from automatic to manual. The constraint values in this case have a wider range of upper and lower limits than the constraint values that allow application of continuous linear programming and combinatorial theory.

In addition, equipment conditions are the conditions of the control system to prevent the destruction of various plant equipment, and when these conditions deviate from the constraint values (upper and lower limits), information such as generation of an alarm or suspension of operation is sent. emanate. In this case, the range of the upper and lower limit values of the constraint values is wider than that in the case where automatic operation is enabled by the above-mentioned distributed instrumentation level.

Normally, this does not occur when the 1M and R preparation conditions are out of their respective constraint values, but this may occur if the amount of steam used in the plant temporarily increases due to rain, etc.

Step 3: When the operational conditions are outside the limit values for enabling optimal control based on continuous linear programming and combinatorial theory, the arithmetic processing unit 120 sends a signal depending on the situation to the conversion valves 5a, 5b, 6a, 6b and either or all of the control valves 2H and 2L. And these conversion valves 5a, 5b.

By adjusting the control valves 6a, 6b and/or the control valves 2H, 2L, the amount of steam supplied to each plant and each power generation system is increased or decreased, and the operational conditions are controlled so as to avoid the limit value and return within the range. .

For example, if the amount of steam in #2 plant increases and the pressure in the steam line 4y falls below the lower limit value under operational conditions,
If there is a margin in the control valves IH and 2H, this control valve IH and 2H
Open H to increase the amount of private power generation while supplying steam from steam line 4X to steam line 4y of #2 plant. If there is no margin, open conversion valves 5a and 6a and supply steam from steam line 4x to steam line 4y. do.

Further, for example, when the amount of purchased electricity is small and the amount of purchased electricity falls below the lower limit value among operational conditions, the control valves IH, 2H, and 3 are closed to reduce the amount of private power generation and increase the amount of purchased electricity.

Step 4 If the control system conditions to enable optimal control using continuous linear programming and combinatorial theory are included in the constraint values, check whether the data collection was performed after a predetermined period of time has passed since the previous data collection. to judge. What is the sampling period for the collected data?

It can be changed arbitrarily depending on the situation.

Step 5 If the data is sampled before a predetermined period of time has elapsed, the arithmetic processing unit 2
0 outputs a signal to control valve 3.

The amount of steam supplied to turbine G3 in the third power generation system is increased to increase the amount of power generation.

In this case, the amount of electricity purchased is OJ! If it goes below, a reverse power transfer phenomenon will occur, so set the lower limit value of power purchase so that the amount of power purchased does not become ≦0 (for example, 500! IH) L/, power system le
Based on the data from control valves 3.a and 16b. IH,
Controls IL, 2H, and 2L.

Step 6 On the other hand, if data sampling is performed according to a predetermined period, the collected data is pre-processed (step 6-1) and then the optimal calculation ( Step 6-2) and determine whether there is an abnormality in the calculation result (Step 6-2).
-3) Do.

Step 6-1 The collected data contains various error factors, which become obstacles when performing optimal calculations using continuous linear programming and combinatorial theory. Therefore, data containing error factors is corrected in advance.

■Steam flow rate variation correction processing: Correction is made so that the amount of steam generated and the amount of steam consumed match in terms of variation.

■Pressure and temperature correction processing for turbine characteristic equations: Corrections are made so that the theoretical generator output from the turbine steam inlet amount matches the actual generator output.

Step 6-2 Based on the above collected data and partially modified data, optimal calculation of steam and power balance is performed using continuous linear programming and combinatorial theory.

Here, continuous linear programming is a linear programming problem (LP; LINEAR) that linearly approximates a nonlinear function near the search point.
This is an optimization method for static problems in which the LP release area is sequentially moved to search for the optimal point. In addition, combinatorial theory is a theory that uses a combination of multiple continuous linear programming solution methods, i.e., when the optimal solution obtained by continuous linear programming is the upper or lower limit within the limits of one continuous linear function ( For example, the third
If the solution found in the area between A and B in Figures is clearly applicable),
This is a theory in which continuous linear programming is performed again in an adjacent region (region between RO and C in FIG. 3).

In this way, the reason why continuous linear programming is combined with combinatorial theory is that when steam flow characteristics and turbine characteristics are linearized, if there are parts that cannot be linearly approximated, continuous linear programming cannot be applied beyond that point. This is because the optimum point may not be found. Therefore, if the optimal point cannot be found, it is necessary to apply continuous linear programming to the next linearization region beyond the part that cannot be approximated by a straight line, and for this reason, combinatorial theory is adopted. .

The calculation method of continuous linear programming in the present invention is as follows.

Independent variable (X)...manipulated variable formula %formula%) We handle the model represented by .

The nonlinear optimization problem searches for a condition X· that maximizes or minimizes the objective function Yobj under the above model function.

If the nonlinear quantity fiF is first-order TAYLAR expanded around one point XO, then Ft (Xo+AX) = Ft (Xo)+5
It is approximated by the linear form of F+/aX・AX.

Therefore, a small change in X (i
=1.2. It is given by...

Since formula (2) is a linear approximation formula for the amount of change from the reference point XO, if the value of Y at XO is YO and the limit in (1) is treated as the amount of change, YL-YO≦ΔY≦Yu − Y.

AX[≦ΔX≦ΔXυ (ΔY=Y-Yo) becomes.

Therefore, ΔXL and ΔXU are linear approximation allowable X widths AxHL
(HOVE LIHIT) Toreba, AX [=
max (Xt −Xo+ ++ΔXHL
)ΔXu=min (Xu −Xo+ AxH
L).

By solving the LP model according to equations (2) and (3), the optimal solution X- within the bond range of AxHL can be obtained from one point XO.

Xψ=XO+Δχ number Δx−+ Solution of r and p according to equations (2) and (3) Furthermore,
By replacing xe with XO and repeating the above calculation, the optimal solution within the restriction (1) is finally obtained.

If the highest efficiency point determined by the above calculation method is at the upper or lower limit within the limits of one continuous linear function, a provisional highest efficiency point is again determined by continuous linear programming in an adjacent region using the above combinatorial theory.

Then, this calculation is performed for each turbine G1-H, G2-H, G
This is performed for all of l-L, G2-L, and G3, and the final solution is determined to be the apparent highest efficiency point, and the optimal calculation is completed.

In this case, when determining the efficiency point of one turbine (for example, Gl-H), it is necessary to calculate the efficiency point of one turbine (G2-H, Gl-L).
, G2-L, G3) are fixed within the limits of the previous continuous linear function. Note that the solution for one turbine (Gl-H) is the upper or lower limit of the continuous linear function. Therefore, even when recalculating efficiency points in adjacent continuous linear function regions using combinatorial theory, for other turbines, the amount of steam supplied and the amount of electricity generated are fixed within the limits of the previous continuous linear function, and This is performed sequentially for each turbine to find the tentative maximum efficiency point for each turbine.

Step 6-3: It is determined whether the optimum calculation result can be used as a control output.

Here, when determining the optimal calculation result as abnormal, ■ If the calculation time exceeds and no output result is obtained, ■ If the output result does not change even though there is an input change, ■ If there is an abnormality due to other production rules (rules of thumb) Case 1: For example: (1) When the amount of electricity purchased is at the lower limit and the input steam amount of the turbine G3 is at the lower limit, the conversion valve 5a.

6a is closed (if it is normal, when the amount of electricity purchased and the power generation of turbine G3 are the lower limit values, the amount of electricity is sufficient, so the conversion valve 5a is closed.

6a to reduce the amount of power generated).

■When the power purchase cost is lower than the power generation cost in turbine G3 and at least one of the conversion valves 5a, 5b, 6a, 6b is open (if normal, turbine G3
Since power generation is expensive, converter valves 5a, 5b, 6a
, 6b are closed and the turbines Gl and G2 other than turbine G3 are
generate electricity).

(2) The input steam amount of the more efficient turbine among the 2 turbines other than turbine G3 is smaller than the other turbines (if it is normal, a large amount of steam is supplied to the more efficient turbine).

If the optimum calculation result is determined to be abnormal, an output is issued to maintain the previous value.

On the other hand, if the optimum calculation result is determined to be normal, it is considered as the true highest efficiency point, and the amount of steam to the power generation system is adjusted to operate at the highest efficiency point while keeping the required amount of steam and the required power constant. A control signal is output to the control valve for distribution.

For example, when the amount of steam to the #2 plant is AT/H and the amount of steam to the #3 plant is B T/H, the overall amount of steam is balanced, and moreover, the amount of oil in the first power generation system is equal to that of the Ale exhaust gas. The amount of oil is B1, the amount of oil in the second power generation system is A2, and the displacement is B
2, the respective flow rates are A3 and B3.
, A4, Bs, the highest efficiency point will be reached, then the control valve IH is used to distribute the amount of steam to the - and second power generation systems so that A=AI +A2 =A3+A4 B=Bl +82 =83+B4 , IL, 2B, and 2L.

When the load fluctuates, the value of A and/or B changes, so
! ! At regular times, the optimal efficiency balance is lost, but in this case, the optimal point is determined again after the load is stabilized, and the control signal is output again based on this.

The present invention can also be applied to an optimal operating method for a system that does not purchase electricity and only uses private power generation. In this case, the difference from the above embodiment is that power purchase data is not collected and power purchase lower limit value is not set and power purchase lower limit value MIN control is not performed. Therefore, when performing optimal calculation, the amount of purchased electricity is assumed to be zero, and optimal calculation control is performed in the same manner as in the above embodiment.

Note that it is also possible to implement this method taking into consideration the amount of steam distributed to one power generation system G3, or to implement this method only for one power generation system, G1 or G2. .

[Effects of the Invention] As described below, according to the optimal lotus rotation method of the power generation system of the present invention, calculations necessary for optimized operation can be performed in a short time, and each device constituting the power generation system can be Since this is also used as a control element during optimal operation, more efficient optimal operation is possible.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an example of a power generation system implementing the method of the present invention, FIG. 2 is a flowchart for explaining the present invention, and FIG. 3 is an explanatory diagram of an optimal calculation method. Bl, B2: Boiler 1, H, IL, 2H, 2L, 3: Control valves 5a, 5b, 6
a, 6b: Conversion valve 9x, 9y, 9z: Pressure gauge

Claims (3)

[Claims] (1) In a power generation system that has multiple equipment such as boilers and power generation turbines and supplies electricity to the plant, at least data on the amount of steam used by the plant, the amount of electricity used by the plant, and the amount of electricity generated by each turbine. At regular intervals, the steam and power balance of each turbine is calculated optimally using continuous linear programming and combinatorial theory based on the data collected above. Based on the results of this optimal calculation, An optimal method for operating a power generation system, characterized in that the power generation system comprising private power generation is operated by controlling the distribution of the amount of steam supplied to each of the turbines. (2) In a power generation system that has multiple equipment such as boilers and power generation turbines and supplies electricity to a plant, at least the amount of steam used by the plant, the amount of electricity used by the plant, the amount of electricity generated by each turbine, and Data on the amount of electricity purchased is collected, and at regular intervals, the optimal calculation of the steam and power balance for each turbine is performed based on the collected data using continuous linear programming and combinatorial theory. An optimal operating method for a power generation system, characterized in that the distribution of the amount of steam supplied to each of the turbines is controlled based on the results of the above to operate a system that uses both in-house power generation and purchased power. (3) For each of the plurality of turbines mentioned above, the amount of steam supplied and the amount of electricity generated in the turbines other than the one for which the optimum calculation is performed are fixed within the limits of the previous continuous linear function, and the optimum An optimal operating method for a power generation system according to claim 1 or 2, characterized in that calculation is performed.