CHAPTER 4 THE INFINITE BUSBAR

4.1       SELF-CONTAINED NETWORKS

In Chapter 1 of the manual ‘Electrical Protection’ the method of calculating the fault level at any point in an offshore installation network was described.

Each element of the network - mainly generators and transformers - was laid out in its relative position as shown on the left of Figure 4.1, and the percentage subtransient reactance of that element was noted against it.  For the short distances involved in an offshore installation cable runs were ignored and all resistances were neglected as compared with the reactances.  The corresponding reactance diagram was drawn, as shown on the right of Figure 4.1, and each reactance figure was adjusted (shown in red) with reference to some arbitrarily chosen ‘base MVA’. The network of adjusted reactances was then reduced by ordinary series and parallel methods to a single reactance (X%) between the fault point and all connected power sources.  From the value of this final percentage reactance the fault level at that point was calculated by the simple equation:

                                                    

From the figures in the example the fault level so calculated is shown in red at the foot of Figure 4.1.

FIGURE 4.1

OFFSHORE NETWORK

In the typical offshore case shown in Figure 4.1 the main sources of power were the platform’s own generators, assisted perhaps by auxiliary generators.  The main generators, though large and sometimes totalling up to 80MVA rating in a large installation, are nevertheless finite in capacity and do not even approach the vast concentration of generated power which lies behind a normal onshore installation fed from the National Grid.

4.2       GRID SUPPLIED NETWORKS

The National Grid consists of a number of very large interconnected power stations, between them capable of delivering continuously a total of many tens of thousands of MVA.

Figure 4.2 represents, on the left, a simple onshore installation fed from the grid.  It is similar to Figure 4.1 except that, instead of the few generator sources, it is fed through large incomer grid transformers from a source consisting of the whole grid system.  Suppose that this source were represented by a single, very large generator of, say, 10 000MVA rating and with a typical percentage subtransient reactance of 15%.  The various reactances are given on the right of the figure, the adjacent figures in red being these reactances adjusted to the same ‘base MVA’ as was used in the first example of Figure 4.1.

Whereas the reactances of most of the other elements will be somewhat increased by this adjustment, the reactance of the very large 10 000MVA generator will be enormously reduced - to only 0.027% - as shown in red on the right of Figure 4.2.

FIGURE 4.2

ONSHORE GRID SUPPLIED NETWORK

When calculating the overall reactance of the network by series and parallel methods as before, the very low reactance of the power source will be far too small to make any appreciable effect on the final answer - in fact it could be ignored.

A power source with negligible reactance - or more strictly with negligible source impedance - is said to be an ‘infinite busbar’.  It is regarded as a source of infinite capacity, such that it can deliver as much active power as is demanded without loss of frequency, and as much reactive power as is called for without loss of voltage.  Put another way it is the ‘perfect generator’, perfectly governed and perfectly regulated.

Figure 4.3 shows such an arrangement in much simplified form.  On the left are the main elements of the plant, including the main incomer grid transformer (which is the property of the CEGB).  Behind this transformer is the infinite busbar.  On the right is the equivalent network diagram, with adjusted reactances shown in red, on which the infinite busbar is shown as having a percentage reactance of zero and which therefore does not influence the calculations which would follow.

It is very convenient, when making network fault calculations on a limited onshore installation which is fed from the grid, to regard the grid power source as an infinite busbar and to concentrate the calculations on those elements within the installation itself.

Clearly the infinite busbar concept can apply only to onshore installations and never to offshore, since on offshore installations the generators, however large, are still limited and finite in size and have a reactance which cannot be ignored.

FIGURE 4.3

ONSHORE NETWORK WITH INFINITE BUSBAR

4.3       GRID CALCULATIONS

Although not strictly the subject of this manual, the calculation of faults within the grid system itself - that is, within the area of the infinite busbar - must be mentioned.  It is of great importance to the system engineer because the generating levels are so high that faults in this area of interconnected systems can give rise to enormously high MVA fault levels and call for switchgear of very great breaking capacity.

It is therefore necessary for the system engineer to make very accurate calculations, and when doing so he can no longer ignore cable runs and resistances as has been done previously in calculations for offshore or limited onshore installations.  It is customary (but not necessary) to adopt a base of 100 000kVA (100MVA).

With long overhead lines or long cable runs in cities the resistance is by no means negligible as compared with reactance, and it must be taken into account.  Indeed the long 275kV lines - and even more so the long 400kV lines - have appreciable capacitance which must also be included.  Similarly, the percentage resistance of generators and transformers, previously disregarded, must be brought into the calculation.

Such calculations, especially with interconnected networks as shown in Figure 3.1, are very complicated and are done by computer; they do not fall within the scope of this manual.

The presence of capacitance in long overhead lines - especially the 400kV and, to a lesser extent, the 275kV lines - gives rise to an unexpected effect.  Whereas reactive power flowing through inductive reactance causes a voltage drop, that same power combined with a large shunt capacitance can cause a voltage rise.  A typical capacitive loading on a long high-voltage line is 2 Mvar (leading) per mile.  Thus a 200 mile line has a standing capacitive loading of some 400 Mvar (leading); the inductive loading on the line in Mvar (lagging) must exceed this value before the capacitance ceases to swamp the inductance and the voltage rise becomes a voltage drop.

The problem for the grid controllers is to deal with voltage rises (that is, overvoltages) in the system during periods of low power demand - for example, at night.  Typically a long line will show a voltage rise with up to one-third power loading, and a voltage drop with a loading above that level.

Voltage control in such cases may be exercised by transformer on-load tap changing, by AVR control or by switching-in reactors to absorb the leading megavars.