DE19819973C2 - Controller circuit for a steam power plant block - Google Patents

Description

The invention relates to a controller circuit for simultaneous primary fre frequency support and power control, with the controller scarf as the output signal device a turbine control valve control signal is formed.

In power plants often used controller circuits for turbine speed control for the purpose of primary frequency support and for power control have as a common principle that a P-speed controller is connected to a PI power controller, as shown in Fig. 2.

The basic circuit shown in FIG. 2 contains a proportional (P) controller 1 as a speed controller and a proportional integral (PI) controller 2 as a generator power controller. Input signal of the speed controller 1 is a speed deviation Δn between the target speed n _S and the actual speed n. According to the set gain factor K, z. B. K = 20, the speed controller 1 provides an output signal K. Δn.

At a first addition point 3 , a signal P _NR , which is supplied by a network frequency controller as a power setpoint component, is linked to a specifiable power setpoint P ₀ . The output signal of the first addition point 3 is fed to a second addition point 4 , whose further input signals are the output signal K. Δn of the speed controller 1 and the generator power P _G supplied with a negative sign are.

It is known that in such a regulated power plant block in the event of a sudden Performance reduction a so-called false control effect occurs. Because of the For example, false control effects can cause problems when separating a force Work blocks occur from a supply network, with a subnet or island network is to be supplied by the power plant block and usually a difference between between the power generated and the power consumed in the subnet. In the such a situation, increasing network frequency deviation from the setpoint, z. B. 50 Hz, must be canceled quickly by the primary frequency control.

The false control effect is shown in FIGS . 3a and 3b. There a reduction of the load P _L by 20% of the nominal power is assumed. It is shown in Fig. 3a that the internal turbine power P _{T is} initially increased, although it should be reduced. FIG. 3b shows that the control signal Y for the Turbinenregefventile and the boiler output WL are first adjusted in the wrong direction. The result is an unnecessarily large deviation in the turbine speed n, possibly even a failure of the turbine set.

The cause of the false control effect lies in the physical condition of the circuit shown in FIG. 2: A change in the mains frequency or change in the turboset speed follows a load change with a delay. It is the result of an ongoing difference between the power generated and the power consumed. The generator performance controller therefore acts on the turbine control valves earlier than is possible by the turbine speed controller. The quick effect of the generator power controller is unfortunately in the wrong direction: In the initial phase, the generator power controller tries to adjust the reduced generator power P _L by means of the modified internal turbine power P _T to the power setpoint that has not yet been corrected by the speed controller. The speed setpoint is delayedly reduced by the speed controller, thereby reducing the turbine output. This results in a deteriorated control quality of the turbine speed compared to a case without false rain effect and possibly also a failure of the turboset or disintegration of an island network.

In the circuit shown in FIG. 2, no false control effect would occur if the generator power controller could be switched off at the time the load changed. However, it can usually not be switched off if the main generator switch remains closed. The switch-off signal for the generator power controller does not appear and the turbine speed control cannot determine the activation of the turbine control valves prior to the power controller.

So that the turbine speed control can gain the upper hand over the power control even when the generator power controller is not switched off, the one from the Deutsche Verbundgesellschaft DVG in the DVG brochure: Dynamic behavior of steam turbine controllers, taking into account the short and medium-term dynamics of the network, Deutsche Verbundgesellschaft e. V., Heidelberg, May 1988 to follow recommended measures. With regard to the occurrence of a false control effect, simulation studies of some turbine control circuits have been carried out. As the best circuit, the DVG recommends the sum circuit shown in FIG. 5 of a speed controller 10 and a generator power controller 20 . The sum of the two controller output signals forms the control signal Y of the turbine control valves. The controllers 10 , 20 are connected in parallel on the input side. If the proportion of the power controller 20 is kept small overall or only slow changes occur, the false control effect should be strongly suppressed during the above transition to operation on an island network with previously unknown power consumption. With reference numerals 30 , 40 , 50 and 60 addition points are referred to.

According to the DVG recommendation, generator power controller 20 should only be used as an I type with an integral time constant T _N <20 s. It is assumed that a power controller should only be used to ensure the linear steady-state characteristic curve required by the supply network between the turbine speed and the generator power deviation.

A power controller should also better regulate the influence of the Boiler existing heating faults on the turbo unit output. The ver Electrical power fed into the supply network should only cause minimal interference exhibit ripple. In this way, the power regulator is intended to calm the Supply network operation contribute. Aside from the recommended slow Effect of the I-power controller, but no suitable interference behavior can be Reach the help of an I-power controller.

The invention is therefore based on the object of a regulator circuit for rotation Number and power control in a steam power plant block, with the not only - while avoiding the false control effect - the desired linear characteristic line in the steady state between the speed and power deviation ge is guaranteed, but also the influence of heating disturbances on the electrical Performance is reduced.

This object is specified by a regulator circuit with in claim 1 resolved characteristics. Advantageous embodiments are in further claims specified.

With the invention - in deviation from the DVG recommendation - it is proposed to start from the regulator circuit shown in FIG. 2, but to regulate the internal turbine output instead of the generator output. The internal turbine power cannot be measured directly, but can be approximated with sufficient accuracy. Different approaches are available for this.

A further description of the invention is given below with reference to the drawing figures.  

It shows:

. Figure 1 is a controller circuit of the invention;

Fig. 2 shows a controller circuit according to the prior art;

Stung Figure 3a shows the course of the turbine power P _T in the case of the circuit of Figure 2, and a sudden load reduction by 20% of the Nennlei..;

Fig. 3b, the course of the turbine valve position signal Y and the heat output WL in the case shown in Fig. 3a;

Stung Figure 4a shows the course of the turbine power P _T in the case of the circuit of Figure 1 and a sudden load reduction by 20% of the Nennlei..;

Hydrochloric derating Figure 4b shows the course of the thermal output and the turbine valve control signal in the case of the circuit of Figure 1 and 20%..;

5 shows a circuit according to DVG recommendation.

Fig. 6 is a block diagram of a controlled system "steam generation and turbo rate" in a power plant block.

Fig. 7 is a diagram P * | T-determination based on measurement data acquired at relevant points of the turbine;

Fig. 8 arrangement for determining the inner turbine performance based on the torque detected by measurement and the rotational speed;

FIGS. 9 and 10 are flowcharts to variants of the determination of the approximated inner turbine performance.

Fig. 1 shows a controller circuit according to the invention, which largely corresponds to the circuit shown in Fig. 2 Darge. . Notwithstanding the known circuit of Figure 1, however, the power regulator 2 regulates the inner Turbinenlei stung P _T, for which the second addition point 4 instead of the generator power, the approximated inner turbine power P * | T is fed. The third addition point 7 supplies a current power setpoint P _{S, act} . The approximated inner turbine power P * | T is determined in a module 6 as a function of the condensate flow _K be. This determination can be made in different ways, as set out below.

By regulating the internal turbine output, a false control effect is avoided. The course of the internal turbine power P _T thus achieved without an initial increase in the event of a sudden load reduction by 20% of the nominal power is shown in FIG. 4a. Fig. 4b shows the corresponding curves of the heat output WL and the control signal Y.

In order to make the difference between the physical variables generator power P _G and turbine power P _T clear, both variables are entered in a block diagram of a controlled system 57 in FIG. 6. The controlled system 57 contains function elements 51 to 55 with dynamic behavior for simulating steam generation and a turbo set. A block 56 forms the electrical supply network. Functional element 51 is a simulation of the effective internal turbine power and consists of two power components. Control variables are the fuel setpoint _B or the boiler heat output WL of an equivalent controlled boiler and the control signal Y of the turbine control valves. 52 is the replica of the turbo set rotors. With 53 the simulation of the pole wheel angle δ and with 54 its implementation in the generator power is designated. The dynamic feedback of slip n-f is designated 55 as power component P _D. From this it can be seen that the turbine power P _T is invariant against a load change P _L , while the same does not apply to the generator power P _G. By changing the load, the network frequency f, as a result of the rotor angle δ and by him the generator power P _G , does not change the inner turbine power.

The internal turbine power P _T arises as an effect of the expanding steam in the individual turbine stages. The internal turbine power is reduced by the mechanical power loss (especially through bearing losses), which is taken into account by means of the mechanical efficiency η _m . The effective drive power of the turbogenerator, also called the effective turbine power P _{T, eff} , is then the product of P _T and η _m , which is simultaneously the product of the effective torque M _{T, eff} on the turbine shaft and its angular velocity ω = 2πn / 60 [1 / s] equals. The internal turbine power P _T is only determined by the boiler heat output WL and the position of the turbine control valves Y. Since the turbine power is not influenced by the generator power P _G or the load P _L , the false control effect or the impairment of the mains frequency stability cannot occur. The desired load dependency is created in a targeted manner, namely only through the effect of the turbine speed control (primary frequency control), where the output is adapted to the current load.

The internal turbine power P _T cannot be measured directly. For the control-technical purpose, however, it can be recorded with sufficient precision using known means of control technology: vapor pressures and temperatures can be measured at relevant points in the turbine, as shown in FIG. 7. On the basis of these measurement data, the steam mass flows in individual turbine sections and the corresponding enthalpy gradient can be calculated. The sum of the products from the partial mass flows and enthalpy drops represents an approximated internal turbine power P _T ^* . The calculation is carried out using a system module that can also be subsequently integrated in a turbine control system. If the turbo set has devices for a condensate stop, the measured condensate flow _K can possibly also be used for a more accurate approximation of P _T.

The internal turbine power P _T can be detected with different levels of accuracy. Four variants for recording P _T ^* are specified below. In order to cancel the effect of the heating disturbances on the electrical power, it is sufficient for example if only the vapor pressures (cf. FIG. 6) are recorded in the turbine. The existing dynamic error can be kept within the permissible tolerance band.

An exact determination of the internal turbine power P _T , ie the greater computational effort required for this, can be worthwhile, since at the same time the position of the expansion line of the steam in the turbine in the region of the superheated steam is averaged. It can be used to monitor the thermodynamic process in the turbine. The change in the flow cross-section (wear in the flow part) of the HD and MD turbine compared to the state at time zero can be recorded. The time of a necessary general inspection can then be determined more objectively or economically than has been practiced up to now.

Four variants are specified for the determination of the signal P _T ^* to be controlled, which approximates the internal turbine power P _T with different accuracy. In a tolerance band, the time profile of P _T ^* corresponds to the profile of the physical quantity of the internal turbine power P _T, which cannot be measured directly.

version 1

A first variant represents the simplest implementation of a determination of the signal P * | T, but at the same time the roughest approximation of P _T. The signal P _T ^{* to be} controlled is replaced by the vapor pressure p ₁ (see FIG. 7) before the first total blading of the high-pressure turbine part. However, the vapor pressure p ₁ is by no means to be equated with the internal turbine output. It is therefore not sensible to specify the width of a tolerance band. If the pressure p ₁ (= P _T ^* ≠ P _T ) according to FIG. 7 is regulated instead of the power ( FIG. 1), the effect of the disturbances in the steam generation Ke on the electrical power is hereby canceled. The disturbances of the turbine power, which occur in the flow part of the turbine downstream of the reheater, such as in the case of heat extraction or condensate freezing, are poorly or not at all compensated for. The pressure p ₁ is also disturbed, but dynamically and in terms of the gain factor differently than the turbine performance.

The controlled variable p ₁ can, however, be regarded as a substitute for the controlled variable turbine power if it is a condensation turbo set without coupling out heat or without a condensate stop.

Variant 2

If the signal vapor pressure p ₁ (variant 1) in a turbo set with a reheater is initially via a dynamic link with the transfer function

is fed to the PI controller as controlled variable P _T ^* (≅ P _T ), the width of the tolerance band can already be specified, in which its course over time will run close to the course of the internal turbine power (P _T ). In the above relationship, a is a power-dependent coefficient that is determined on the basis of the document on the thermal turbine design at zero operating time. This variant is identical to variant 1 with regard to the possible regulation of faults.

Variant 3

So that the turbulence of the turbine output caused by heat decoupling or condensate freezing can also be corrected, steam mass flows and the corresponding enthalpy gradient must be calculated from section by section. From this, the turbine partial powers are calculated, which in total emulate the approximated inner turbine power P _T ^* (≅ P _T ) with the best possible accuracy that can be achieved at the time of zero turbine operation.

The steam mass flows in the HP, MD and LP turbine sections are calculated according to the relationships ( 1 ), ( 2 ) and ( 3 ) below. The additional index N means the nominal state of the respective physical quantity. Since the ratio of the - absolute - temperatures (Θ = (T + 273) ° K) differs only slightly from the value one when the temperature changes in the turbine, this can usually be taken as one in ( 1 ) to ( 3 ) . The steam mass flows are therefore only dependent on the pressures before and after the respective turbine part. In turbines with heat extraction or condensate stop, the above turbine parts still have to be divided into subsections and the steam mass flows have to be calculated in the same way. The subsections are determined by the position of the steam tapping points for the heat extraction or for the condensate stop.

The individual enthalpy drops Δh _j can be determined in two different ways:

• a) With this procedure, one starts from the load-dependent expansion line of the steam in the turbine, through which the thermodynamic process in the turbine is documented stationary at zero operating time. With the help of this document, the thermodynamic efficiencies η _{th, i of} the individual turbine sections i = 1 to n are known and the corresponding polytropen exponent k _{i is} determined off-line according to the relationship ( 4 ). Using the known k _i , the ratio of the enthalpy gradient is calculated according to equation (5). From ( 5 ) it follows that the enthalpy gradient Δh _i depends on the ratio of the outlet p _ai and inlet pressure p _{ei of} the i-th section and on the enthalpy gradient at the nominal power Δh _iN .
  The turbine power in the i-th section is now calculated according to ( 6 ):
  The controlled variable P _T ^* (≅ P _T ) is then according to ( 7 ):
  The polytropic exponent k _i is not taken constant for the entire control range of the turbine power, but is determined for each turbine power from the power control range according to the relationship ( 4 ), so that k _i = k _i (P) applies (see FIG. 9, " Flow chart for variant 3 a "). The load-dependent polytropic exponent k _i (P) is delivered with the aid of a function generator with the entry signal P _{S, act} , so that k _i = k _i (P _{S, act} ). Variant 3 a already fulfills the control technology task of eliminating the effect of the malfunctions inherent in the KW block on electrical power. After connecting the tubo set generator in parallel, the entire turbine power range is covered by the signal P _T ^* . When the Turbosat zes starts up, the power controller arranged in parallel with the speed controller is switched off anyway.
• b) With this procedure, the effective enthalpy gradient Δh _{i is determined} directly, ie from the steam temperatures and steam pressures measured at the boundaries of the sections and not by means of the document on the thermodynamic turbine efficiency η _{th, i} at time zero. The advantage ge genüber variant 3 a is that the enthalpy to each beliebi gen later stage of the turbine operation with the same accuracy in the range of the superheated steam can be determined, that is, even after egg nem wear in the turbine flow area. In the area of wet steam, however, proceed according to variant 3 a, since the degree of moisture cannot yet be measured in an operationally acceptable manner. For the calculation of the enthalpy h _ei and h _ai - based on measurement data and the relationship ( 8 ) - a building block in a control system, z. B. the system module "enthalpy calculation" can be provided in the ABB control system PROCONTROL P.
  h = h (p, T) (8)
  The effective enthalpy gradient Δh _{i of} the i-th turbine section is then calculated according to "( 9 ) at the respective instantaneous turbine output.
  Δh _i = h _ei - h _ai (9)

The section outputs are calculated according to ( 6 ) and the controlled variable P _T ^* (≅ P _T ) according to ( 7 ).

The enthalpy gradient Δh _i in the area of wet steam (cf. FIG. 10, flow chart for variant 3 b) can be determined on the basis of the specific steam wetness (1 - x _i ) instead of the polytropic exponent k _i . From the documents, the efficiency η _{th, i of} the respective turbine section is not determined off-line in the wet steam area, but the coefficient x _i , which approaches the value one. Using the load-dependent x _i , the enthalpy of entry and exit in ( 9 ) are then calculated as follows:

h _{x i} = (1 - x _i ) h '| i + x _i h''| i (10)

Which of the four variants for determining the approximate inner turbines power is used, depending on the design of the turbine and de to decide their use.

Fig. 8 shows a further possibility for direct determination of the effective internal turbine power P _{T, eff} , which is used as a controlled variable for the power controller and can also be used for monitoring the turbine with restrictions. A turbogenerator set is shown in the usual way, with generator G, LP, MD and HP part of the steam turbine, condenser K and steam generation K _e . At the dome point between the turbine and generator, the turbine torque M is detected by means of a transducer MG. In addition, the speed n is measured by means of a speed sensor or tachogenerator TG. The product from M. n is the effective internal turbine power P _{T, eff} formed by the multiplication point 80 .

Claims (5)

1. Controller circuit for forming a turbine control valve control signal (Y) in a steam power plant block by means of a turbine speed controller ( 1 ) and a power controller ( 2 ), wherein

• a) the speed controller ( 1 ) is designed as a P controller and the power controller ( 2 ) as a PI controller,
• b) a speed deviation (Δn) is supplied to the speed controller ( 1 ) as an input signal, and
• c) the power controller ( 2 ) is supplied with an input signal which is formed by sum mation ( 3 ) of the speed controller output signal (K. Δn) and the power deviation (ΔP), which is based on an approximate inner turbine power (P * | T ) is related.

2. Control circuit according to claim 1, characterized in that the internal turbine power (P * | T) in a module ( 6 ) as a function of the measured vapor pressure (P), and the steam temperature (T) is determined. 3. Control circuit according to claim 2, characterized in that in the construction stone ( 6 ) in addition the measured condensate flow ( _K ) is used to determine the internal turbine power (P * | T). 4. Control circuit according to one of claims 2 or 3, characterized records that the steam pressure and the steam temperature at several points of the Steam turbine is measured. 5. Control circuit according to claim 1, characterized in that for loading tuning the internal turbine performance with measuring means (TG, MG) Turbine speed (n) and the turbine torque (M) are recorded and from them the internal turbine power is calculated.