EP1166422A1

EP1166422A1 - Improved method for supplying electric power to a network using an alternating current generator associated with a turbine

Info

Publication number: EP1166422A1
Application number: EP01907701A
Authority: EP
Inventors: Serge Mourier; Marc Vezinet
Original assignee: Air Liquide SA; LAir Liquide SA pour lEtude et lExploitation des Procedes Georges Claude
Current assignee: LAir Liquide SA pour lEtude et lExploitation des Procedes Georges Claude
Priority date: 2000-01-26
Filing date: 2001-01-25
Publication date: 2002-01-02
Also published as: BR0104247A; FR2804254A1; CA2365632A1; FR2804254B1; AR027930A1; KR20020004977A; AU3560401A; WO2001056131A1; US6351103B1; JP2003521209A

Abstract

The invention concerns a method for supplying electric power to a network using an alternating current generator associated with a turbine, via a transformer (reactance XT). The invention is characterised in that it consists in assessing the generator as precisely as possible based on the site requirements by lowering its nominal voltage to a maximum low while continuing to satisfy the capacity of the generator to export its active power delivered by the turbine and the reactive power corresponding to said active power demanded by a consumer or operator of the network.

Description

Improved method of supplying electrical power to the network using an alternating current generator associated with a turbine

The present invention relates to the conditions for connection to the network of electrical energy production sites and relates more particularly to the improvement of the conditions for connection to the network of high power alternators. The connection to the network of a generator of the aforementioned type is ensured by means of a transformer.

The generator delivers active power and reactive power to the network.

The active power Pg supplied by the alternator is given by the relation:

Pg = -Vg.l. cos φ _g (1) where

- Vg is the voltage across the generator, I is the current, cos φ _g is the generator power factor. The active power Pcp received by the network at the delivery point is given by the relationship:

Pcp = + Vcp.l.cosφ _cp (2) where

Vcp is the voltage across the network, at the transformer output, I is the current, φ _cp is the power factor seen by the network at the delivery point.

Pcp = - Pg (3)

It will be noted that taking into account the above: Vg. cosφ _g = Vcp.cosφ _cp

It can therefore be seen that the active power received by the network is equal to the active power supplied by the alternator, assuming that the transformer losses are negligible.

The reactive power Qg of the generator is given by the relation: Qg = -Vg.l.sinφ _g (4)

The reactive power of the network is given by the relation:

Qcp = + V _cp .l.sinφ _cp (5)

Qg-Xtl ² = Qcp (6) or

Xt is the resistance of the transformer.

Thus, the reactive power received by the network is equal to that supplied by the alternator, reduced by the consumption of the transformer in reactive power Xtl ² .

However, the active power and the reactive power of the network supplied by the generator are constant data which the electric energy supplier must take into account for the construction and operation of the generator. Specifications drawn up by the EDF / GS21 study committee define the conditions for connecting private energy production units to the public transport network.

The main rule concerning the electrical sizing of an energy production installation is as follows: - verification of the maximum short-circuit current supplied by the installation at the point of delivery on the network, this verification resulting in the most vast majority of cases of inducing an increase in the reactance of the transformer to reduce the short-circuit current of the installation, - verification of the capacity of the installation to supply to the delivery point, a certain amount of power reactive depending on the voltage conditions (value) at the delivery point.

This supply constraint is expressed in the document GS21 by a graph on which the reactive power Q exchanged at the delivery point is shown on the abscissa, expressed in proportion to the maximum active power of the installation and officially called Pmax and on the ordinate, the delivery point voltage expressed as a relative value compared to a reference voltage, officially called Vref, the value of which is given by EDF.

The above verification and the resulting optimizations have naturally contradictory effects. Respecting the initial constraint on the short-circuit current tends to very often increase the reactance of the transformer connecting the alternator to the network.

The increase in this reactance induces: - higher consumption of reactive power in the transformer, which is harmful when trying to export this reactive power to the network,

- an increasingly marked incapacity when the reactance increases, to:

- export reactive power when the voltage at the delivery point is a little high,

- import reactive power when the voltage at the delivery point is a bit low. The performance degradation linked to the reactance of the transformer is noted on the basis of the examination of two graphs.

The performances [q, u] at the terminals of the alternator which can be qualified as raw,

The corresponding net performance [q, u] at the delivery point. The raw performance graph [q, u] across an alternator is a rectangle.

As an indication, we can give:

- a height varying from umin = 0.9 Un to umax = 1, 1 Un

- a width of Qmin = - 0.2Pmax to + 0.6 Pmax The graph at the origin of the demonstration in the specifications is based on an alternator whose diagram [q, u]) at its terminals is a rectangle .

The graph of net performance [q, u] at the delivery point becomes a trapezoid. The initial rectangle containing all the points for which a functioning of the generator is guaranteed deforms naturally when one crosses the reactance of the transformer.

The horizontal upper and lower limits indicating the constant maximum or minimum voltage of the alternator tilt and at the transformer secondary remain almost straight.

They represent the voltage conditions and the reactive power conditions existing at the transformer secondary for the same operating state of the alternator. In practice, the upper and lower limits are naturally obtained when the alternator regulator is blocked in the maximum voltage and minimum voltage positions respectively and the network voltage at the delivery point is varied to maintain the same power reactive absorbed or supplied by the alternator.

There is then an inability of the “alternator + transformer” assembly to supply, or to consume reactive power for the same voltage conditions.

Much larger voltage differences at the point of delivery are required to exchange the same amounts of reactive power.

The greater the reactance of the transformer, the greater the deviations must be.

The increase in the reactance of the transformer is reflected on the graph by an increase in the slope of the trapezoid. The invention aims to remedy the aforementioned drawbacks of connecting large machines to the network and to improve the conditions for connecting such machines to the network.

It therefore relates to a method of supplying electrical energy to the network using an alternating current generator associated with a turbine, via a transformer, characterized in that it consists in evaluating the generator as fair as possible according to site needs by lowering its nominal voltage to the lowest level while continuing to satisfy the generator's capacity to export the active power delivered by the turbine and the reactive power corresponding to this active power required by a consumer or the network manager.

According to particular features of the invention:

- the voltage is lowered so that the generator can always transform the active power of the available turbine as closely as possible or sized according to site conditions and that the generator can simultaneously deliver reactive power as closely as possible associated

- the generator is adapted to a reduced voltage Ungf in order to reduce the short-circuit current and increase the dynamic range of the regulated voltages at least equal to that of its old range around its old nominal voltage Ungi; - the generator being oversized and initially designed for operation under a Ungi voltage, its operating voltage Ungf with a reduced short-circuit current is linked to the initial operating voltage Ungi by the relation:

in which Sg = Ps ² + Qs ²

Ps and Qs being respectively the active power and the reactive power of the site, and Png and Qng being respectively the nominal active power and the nominal reactive power of the generator;

- the reduction of the operating voltage of an existing generator is ensured by reducing its excitation current and adjusting it according to the new nominal voltage;

- re-evaluation of the generator is ensured by reducing its voltage dimensioning to the lowest possible nominal voltage, still allowing the desired active and reactive power values to be obtained.

The invention will be better understood on reading the description which follows, given solely by way of example and made with reference to the accompanying drawings, in which: - Fig.1 is an equivalent diagram of a current generator alternative connected to the network;

- Fig.2 is a vector representation of the voltage across the network delivered by the generator with a first reactance value of the link transformer; - Fig.3 is a vector representation of the voltage across the network, delivered by the generator with a second reactance value of the link transformer;

- Figs.4 and 5 are vector representations of the characteristic quantities of the generator and the network with two different values of the voltage at the delivery point;

- Fig.6 is a vector representation of the voltage at the delivery point and the generator voltage and their respective slopes illustrating an additional difficulty in connecting the generator to the network. We will first examine the operation of the diagram in Figure 1 at constant Pcp and Qcp, evolving as a function of the reactance Xt of the transformer interposed between the generator and the network (Fig1). This situation results in the relations: Vgcosφ. = Vcpcosφ _cp

Vg2cosφ _g2 = Vcpcosφ _cp

The diagrams represented in FIGS. 2 and 3 reproduce the changes in each of the characteristic quantities when the reactance of the transformer is varied by keeping the active power Pcp and the reactive power Qcp delivered to the network constant, which responds to a general constraint imposed by network managers to keep the Qcp / Pcp ratio constant.

The diagram in FIG. 1 is a vector representation of the different quantities involved in the operation of the system. The current I is directed towards the reference axis.

The voltage across the generator having a first value Vg1 has a phase shift cp _g1 relative to the current I.

The voltage Vcp at the delivery point has a phase shift φ _cp with the current I due to the presence of the reactance 1 having a first value X _t1 causing a voltage drop X _t1 .l.

The vector sum of the electromotive force of the generator Eg1 and the voltage Xpl due to the Potter reactance of the generator is equal to the voltage Vg1 across the terminals of the generator.

The vector sum of the voltage Vcp across the network and the voltage Xr.l due to the direct reactance of the network seen at the delivery point is equal to the electromotive force at the delivery point of the network.

If the reactance of the transformer increases, we are in the presence of the graph in Fig. 2.

The voltage Vcp at the terminals of the network and the electromotive force at the delivery point remaining unchanged, it is necessary to increase the voltage Vg2 at the terminals of the generator as well as its phase shift φ _g2 relative to the current I and of course the electromotive force Eg2 of the generator.

So to react to the increase in the reactance of the transformer and to keep Pcp and Qcp constant: - the alternator must be able to deliver a voltage Vg2> Vg1,

- the alternator must continue to supply the initial current I,

- the alternator must have a better power factor

Φg2> Φg-t = COSφ _g2 <COSφ _g1 Such requirements immediately imply for the alternator:

- a more extreme magnetic state resulting in higher losses in iron.

But losses in iron generally represent less than 15% of the total losses, this same total representing only about 1.5% of the nominal power of the machine.

- no change in losses by Joule effect and additional losses, because the stator current remains constant.

In fact, there is an increase in losses to the rotor because to obtain Eg2> Eg1, a higher excitation is required. The impact of the increase in φ _g is taken into account by the rotor current. It results from the increase in Eg.

It is a question of being able to excite more which induces losses by Joule effect to the rotor.

It can therefore be seen that if a standard machine is used: 1) it is necessary to be able to excite the machine more to increase its electromotive force or f.e.m. internal Eg and therefore its external voltage Vg.

The machine thus provides the same active power but more reactive power.

This additional reactive power is consumed in the reactance of the transformer.

2) it is necessary to have the capacity to evacuate a delta (difference, increase) of calories in addition to the increase in losses in iron and by Joule effect.

The increase in the reactance of the transformer connecting the alternator to the network to reduce the short-circuit current to the network leads to having an alternator with the same active power and a higher reactive power and able to operate at a nominal voltage more high. We now consider the operation at constant Pcp and Qcp and the evolution of this operation as a function of the voltage at the delivery point.

Vg1 and Vg2 being two different generator voltages, this condition results in the relations:

Vg1 cosφ _g1 = Vcp1 cosφ _cp Vg2cosφ _g2 = Vcp2cosφ _cp The diagrams in Figures 3 and 4 reproduce the changes in each of the characteristic quantities when the voltage at the delivery point is varied, keeping the active power Pcp constant as in the previous case and the reactive power Qcp delivered to the network.

The evolutions here are less obvious and cannot be deduced all from the graphics.

We must return to the basic equations of the balance sheets of powers (1) to (6) and look at the changes necessary for their conservation.

The apparent power Sep being kept at the delivery point (Pcp, Qcp unchanged): the power factor at the delivery point is unchanged and remains cosφ _cp ; the new current 12 is deduced from the old current 11 by the equation:

Scp1 = Scp2 = Vcp1.l1 = Vcpl2

12 = 11 (Vcp1Λ / cp2)

The new current is lower than the old; the reactive power consumption in the transformer becomes:

Qt2 = Xtl ₂ ² = Xtl, ² (Vcp1Λ / cp2) ² = Qt1 (Vcp1Λ / cp2) ²

It will be noted that in the event of an increase in voltage, Vcp1Λ / cp2 <1 and therefore Qt2 <Qt1.

The alternator must therefore supply the same active power; Pcp and a total reactive power Qg2 such that:

Qg2 = Qcp + Qt2 = Qcp + Qt1 (Vcp1Λ / cp2) ²

The alternator power factor is deduced from its new tgφ _g2 = Qg2 / Pcp.

Since Qg2 <Qg1, cosφ _g2 > cosφ _g1 . A better cosφ will therefore not be required from the alternator.

The new alternator voltage is deduced from equation (7) established in the paragraph which relates to the slope of evolution of the reactive power as a function of the voltage at the coupling point. Qcp + Qg = (Vg ² -Vcp ² ) / X ^* ^ Vg ² = Xt- (Qcp + Qg) + Vcp ² (7)

It is established above that Qg decreases when the voltage Vcp increases. It suffices therefore that the equation Vg ² = Xt. (Qcp + Qg) + Vcp ² be verified by Vcp = Vref = Uref / 3 ^{0 5} and that the alternator can raise its voltage Vg in a manner homogeneous to Vcp. To simplify, we can define a constant Vo:

Vo ² = Xt (Qcp + Qg) _{[Vcp = Vref]} = Xt (Qcp + Qg _{[Vcp = Vreη} )

Then Vg ² = Vcp ² + Vo ² and therefore dVg / Vg = dVcp / Vcp is the sufficient relationship so that the alternator can continue to "export" at least the same amount of reactive power seen from the point of delivery.

It is clear that to react to the increase in tension at the point of delivery and to keep Pcp and Qcp constant; the alternator must be able to raise the voltage Vg across its terminals so that the relative variation dVg / Vg is equal to the relative variation of the voltage Vcp at the delivery point (dVcp / Vcp).

The documents expressing the connection conditions show an additional connection condition.

It results directly from the inability of modern alternators to offer a voltage regulation dynamic across their terminals greater than ± 10% around their nominal voltage.

The resulting discomfort is especially noticeable for the high network voltages which prevent the alternator, itself unable to raise its voltage by more than 10%, to provide all the expected reactive power.

This usual limitation of alternators induces the management bodies of the transport networks to impose, rather than a maximum transformer reactance, a slope for the curve [q, u].

This slope is not mathematically equal to that of the parallelogram appearing in the usual diagrams of the specifications at delivery point, but still remains fairly well represented by the slope of the parallelogram.

The lower this slope, the less the loss of reactive energy in the transformer will be perceptible, especially for high network voltages. The slope is defined by the variation of reactive power exchanged at the coupling point as a function of the difference between the voltage at the coupling point and the alternator voltage.

By presenting the variations with the prefix "d" as "delta" or "difference": Vgcosφ _g = Vcpcosφ _cp (8)

Vgsinφ _g = Vcpsinφ _cp + Xt.l (9)

Vg.l.sinφ _g = Vcp.l.sinφ _cp + Xt I ² Qg = Qcp + Qt

Qg - Qcp = Qt (10) Vg ² = Vcp ² + 2Vcp.l.sinφ _cp Xt + Xt ² .l ²

Vg ² - Vcp ²

= 2Vcp.l.sin φcp + XTI ²

xt

Vg ² - Vcp ² = 2Qcp + Qt = Qcp + Qg Xt f Vg ² - Vcp Qg + Qcp = [* _χt j

2Qcp = (Vg ² -Vcp ² ) / Xt - Qt Qcp + Qg = (Vg ² -Vcp ² ) / Xt (7) dQcp = 2 (Vg / Xt) dVg (1 1)

These expressions represent the relationships sought. They only involve the variables directly incriminated in the control of the slope. Vcp: voltage at the delivery point, which will be assumed to be high in order to be able to assess the absolute voltage of the alternator Vg necessary for virtually maintaining Qcp.

Vg: absolute voltage across the generator, Xt constant: reactance of the transformer, Qcp: reactive power at the delivery point, the variation of which as a function of Vg (slope) must be controlled, Qg: reactive power supplied by the alternator, Qt: reactive power consumed by the reactance of the transformer.

It is from expression (7) that the network managers initially wanted to set a maximum reactance of the transformer Xt so that Qcp retains a sufficient value.

Fixing the slope of the above function is the same because the slope can be seen from the point of delivery without imposing constructive provisions on any particular machine (alternator or transformer). The usual sizing of industrial "turbine - alternator" groups is always carried out: a) on the basis of the maximum active power of the turbine reached for energetic and process fluid conditions not always available on the site where these groups will be installed; b) on the basis of a nominal reactive power Qmax = 0.6 Sn (and therefore Pmax above = 0.8 Sn), which corresponds to a power factor cosφ = 0.8 Sn; c) on the basis of a nominal voltage Vng (Ung between phases) optimized by the manufacturer of the alternator for reasons of stator insulation, overheating, etc.

The “turbine + generator” group is very often slightly oversized from the point of view of active power and almost always oversized from the point of view of reactive power.

The generator voltage is made by the manufacturer and not by the network manager.

It is therefore never imposed. This is based on the related implications:

- on the balance sheet when the reactance of the transformer increases in order to reduce the short-circuit current; - on the balance sheet when the voltage at the delivery point increases (increase in the Pcp, Qcp performance range);

This is based on the way in which the nominal characteristics of rotating machines (IEC34) are established and the way the alternators are usually standardized, measured and associated with the turbines that the present invention is constructed.

The verifications which follow are intended to allow the implementation of the invention. 1) Nominal active power on site of the turbine less than or equal to the nominal power of the alternator brought to the terminals of the alternator, therefore multiplied by the efficiency of the alternator.

2) Nominal reactive power at the delivery point (Q) lower than the nominal reactive power of the alternator minus the reactive power consumption of the transformer.

The reactive power consumed by the transformer Qt and its reactance X1 are linked by the relation Xt = ukr (Ure 7Snt) where ukr is the short-circuit voltage. Qt # Sut / uk2.

The balance sheet of the verification necessary for the implementation of the invention is as follows.

1) there is a generator perhaps oversized in active power;

2) this same generator is oversized in reactive power. The invention consists in evaluating the generator as closely as possible according to the need of the site, by lowering the nominal voltage to the lowest while continuing to satisfy the ability to export the active power delivered by the generator drive device, the power reactive corresponding to this active power required by the consumer or the network operator.

The lowering of the voltage is ensured so that the alternator can always transform the active power of the drive device available or dimensioned as closely as possible according to site conditions and that the alternator can simultaneously deliver, as accurately as possible, the power associated reactive, defined by the consumer.

Although many rotating machines operate over extended voltage ranges (outside the ranges defined by the standards in force, IEC 34 for rotating machines, if some generators supply variable voltages for specific, often temporary, applications, to the knowledge of the Applicant, there is no industrial application, for large machines (> 10MW) where the machine is voluntarily decommissioned, evaluated downward and patched with performance corresponding to a reduced voltage in order to:

- reduce the short-circuit current of the machine (subtransient, transient and permanent) and this calculated by the standards in force; - increase the dynamics of the regulated voltage range so that the alternator can continue to export its “new” reactive power (lower than the old one) for regulated voltages at least equal to those of its old range around its old voltage nominal.

We will now examine the evaluation of the minimum derating in tension.

The method used is pessimistic because as the losses to the rotor are reduced in a large proportion because they vary with the square of the excitation current and the f.e.m. demand is lower, as the overall iron losses are slightly reduced due to the f.e.m. reduced, in the caloric balance of the losses to dissipate, one could therefore accept losses of the stator a little higher than the original losses.

It will also be noted that the leakage fluxes decrease under excitation as well as the Potter's reactance of the generator, so that the losses to the rotor are greatly reduced compared to the relatively low derating at nominal voltage.

The original apparent power is expressed by the relation:

Sngi ² = Pngi ² + Qngi ² The final apparent power is expressed by the relation: Snfg ² = Pngf ² + Qngf ² In order to work at minimum at constant stator current, we must have:

Ungf Sngf Ungi Sngi

At a minimum, the new stator voltage is deduced from the old by the relation: Ungf = Ungi [Sngf / Sngi]

Ungi representing the original nominal voltage of the alternator.

Sngi representing the original alternator plate.

Sngf representing what the site actually asks the alternator. Of course, for optimization purposes, the manufacturer will have the task of obtaining for the generator a new nominal voltage as low as possible.

The short-circuit current evaluated by standards such as IEC 909 in Europe, for example, reveals two terms in the evaluation of the impedance involved in the short-circuit current.

The overall impedance is according to the international standard: Zkg = kg (0.05 or 0.07 + i) x'd.

0.05 for machines> 100MVA and 0.07 for machines with a power lower than 100MVA. x'd = transient reactance (originally subtransient x "d in IEC 909). kg is a corrective term for impedance, specific to synchronous machines whose value is: kg = cmax / (1 + x'd.sinφ _ng ) with: cmax = coefficient imposed by the standard and modifiable by the network operator. x'd = transient reactance (originally subtransient x "d) of the alternator.

Sinφ _ng is deduced from the nominal power factor cosφ _πg of the alternator. cmax varies from 1 to 1, 1 according to the type of minimum or maximum short-circuit current that one seeks to establish and according to the quantity of hazards of the network that one seeks to take into account in the result final.

The new cosφ having been increased resulting in a probable reduction in performance due to the reduction in the Qng / Png ratio, Sinφ _πg is therefore reduced.

As a result, kg increases slightly. The other terms are unchanged, except x'd which increases slightly due to the desaturation caused by a sub-excitation operation at lower fem. As a final result, the new short-circuit current evolves according to the standards as follows:

Initial value l'kgi = cmaxUni / (3 ⁰⁵ Zkgi) Final value l'kgf = cmaxUnf / (3 ⁵ Zkgf) With Unf <Uni and probably Zkgi <Zkgf, we therefore see that the short-circuit current across l The alternator is reduced in a ratio at least equal to the ratio of the initial and final nominal voltages. Red Icc <Ungf / Ungi We also obtain an increase in the voltage regulation dynamics towards high voltages.

Originally, for modern machines, the regulation dynamics is generally given by the relation: Umaxi = 1, 1 Ungi

With a new nominal voltage: Ungf = k Ungi with k <1, Umaxi remains unchanged in absolute value and is worth: Umaxi = (1, 1 / k) Ugnf The increase in dynamics, towards the maximum voltages is therefore in the ratio 1 / k. AugDyn = Ungi / Ungf

Claims

1. Method for supplying electrical energy to the network using an alternating current generator associated with a turbine, by means of a transformer, characterized in that it consists in evaluating the generator as accurately as possible according to the needs of the site by lowering its nominal voltage to the lowest while continuing to satisfy the generator's capacity to export the active power delivered by the turbine and the reactive power corresponding to this active power required by a consumer or the network manager .

2. Method according to claim 1, characterized in that the lowering of the voltage is carried out so that the generator can always transform as closely as possible the active power of the turbine available or dimensioned by site conditions and that the generator can simultaneously deliver as much as possible, the associated reactive power.

3. Method according to one of claims 1 and 2, characterized in that the generator is adapted to a reduced voltage (Ungf) in order to reduce the short-circuit current and increase the dynamic range of the regulated voltages at least equal to that of its former range around its former nominal voltage (Ungi).

4. Method according to one of claims 1 to 3, characterized in that the generator being oversized and initially designed for operation under a voltage (Ungi), its operating voltage (Ungf) with a reduced short-circuit current is linked to the initial operating voltage (Ungi) by the relation: / Png ² ₊ Qng ² in which Sg = VPs ² + Qs ²

5. Method according to one of claims 1 to 4, characterized in that the reduction in the operating voltage of an existing generator is ensured by reducing its excitation current and adjusting it as a function of the new nominal voltage;

6. Method according to one of claims 4 and 5, characterized in that the re-evaluation of the generator is ensured by reducing its dimensioning in voltage up to the lowest possible nominal voltage still allowing the obtaining of active power values and reactive sought.