Industrial Power Transformers -- Special features of transformers for particular purposes [part 6]

Home | Articles | Forum | Glossary | Books


In the UK the term system transformer is normally used to describe that class of transformer which provides the first stage of distribution beyond the step ping down to 33 kV, or occasionally 66 kV, of the bulk supply from the transmission system operating at 132 kV or above. That is, it is the transformer used to make the transformation from 33 or 66 to 11 kV.

These transformers are unique in that they are not strictly designed to EN 60076 temperature rises but are tailored to meet a particular duty. They were widely introduced in the early 1960s, although the concept had been around for somewhat longer, and were designed with the intention of minimizing use of material and manufacturing costs as well as more precisely matching the operational requirements of what were at that time the area electricity boards (now distribution network operators). For this reason, at the time of their introduction, they were known as 'integrated system transformers' usually abbreviated to ISTs. Now they are generally termed CERs or 'continuous emergency rated' transformers, referring to the manner in which their rating is derived.

It was the practice of the distribution authorities in the 1960s, and it generally still is, to operate primary distribution transformers in pairs connected in parallel so that, should one of these fail, the remaining unit will carry the substation load until the failed transformer can be repaired or replaced so that supplies to the consumer will not be interrupted. Standard ratings for these transformers at that time were 10/14, 15/21 and 20/28 MVA. In each case the lower rating is achieved with ONAN cooling and the higher value by means of pumps and fans to provide an OFAF or ODAF rating. A 10 MVA transformer has an LV current, at, say, 11 500 V of 502 A. At 14 MVA this is increased to 703 A. But the available 11 000 V switchgear had a standard current rating of 800 A, so when operating singly at its forced cooled rating the transformer is not able to match the full current capability of the switchgear and at its ONAN rating, as one of two transformers sharing the substation load equally, it would be quite sufficient for it to be able to carry 400 A, not 502 A. The transformer is considerably over designed at the ONAN condition and switchgear capacity is being wasted at the forced cooled condition which was intended for the emergency, 'one transformer carrying the total substation load,' situation.

The requirement for the IST was thus that it should have an emergency rating which was a close match to that of the substation 11 kV incoming switch gear when carrying the full load of the two-transformer substation. This rating need only be sustainable for possibly a week or two and may be achieved by the operation of pumps and fans. At all other times the rating should be half this emergency duty and must be achieved without the operation of any forced cooling equipment. The following ratings are those normally used at the present time:

• 7.5/15 MVA corresponding to 800 A switchgear,

• 12/24 MVA corresponding to 1200 A switchgear,

• 15/30 MVA corresponding to 1600 A switchgear,

• 20/40 MVA corresponding to 2000 A switchgear.

The LV currents for these ratings, for the forced cooled condition and assuming an LV voltage of 11 500 V are, respectively 753, 1205, 1506 and 2008 A, from which it will be seen that the 12/24 MVA and 20/40 MVA match the switchgear ratings better than do those of 7.5/15 MVA and 15/30 MVA. In fact 8/16 and 16/32 MVA are a better match for 800 and 1600 A, respectively. The reasons for this are not clear but it is probably the case that at sometime some one's desire to use round numbers got the better of simple logic. It remains to be seen whether a new series of CER transformer ratings will be introduced having values which match the ISO 3 preferred R10 series of switchgear cur rent ratings of 630, 800, 1000, 1250, 1600, 2000, 2500 A, etc. currently in use. When the IST was first introduced the intention was that there should be a high degree of standardization to enable a flow-line type of production to be used thereby assisting the objective of minimizing costs. In the event, this aspect of the original concept has been somewhat lost, so that designs tend to be tailored to suit the requirements of each particular application, that is tap ping range, voltage ratio, impedance values, terminal connections, etc. are varied as each installation demands. Logically there is therefore no reason why a new series of transformer ratings having ODAF rated powers of 12.5, 16, 20, 25, 32, 40 MVA, etc., which have LV current ratings to match present-day switchgear, should not be introduced.

In line with the fact that the CER transformer ODAF rating is regarded as an emergency rating, the permitted ODAF temperature rises are related to an ambient temperature of 5ºC, and a hot spot temperature of 115ºC is permitted.

In fact, some specifications now allow a hot spot temperature of 125ºC at the emergency ODAF rating. By reference to Table 21 (Section 6.8) it will be seen that the first of these two values corresponds approximately to rate of use of life of 8 times normal and the second, using some interpolation, to about 23 times normal. Although it will appear that the latter rate, in particular, is exceedingly high even if occurring for only 1 or 2 weeks in the transformer lifetime, it should be recognized that substation loading is likely to be cyclic of the form shown in FIG. 19 where the time at maximum load is unlikely to be more than 10 hours per day and may possibly be only 6-8 hours. Ten hours daily for 14 days at 125ºC will thus use up (10/24 _ 14 _ 23) _ 134 extra days of life, that is about 5 months of life. This is not going to be noticed in a lifetime of 30 years plus, and is probably a fairly small price to pay for the economies gained from rating the transformer in this way.

FIG. 19 Typical daily load cycle

In reality, there is another factor which will reduce the degree of ageing that the CER transformer will be likely to suffer in such an emergency. Because system loading is always tending to increase, substations must be periodically reinforced by the installation of additional transformers. The loss of one transformer will only result in the remaining unit being required to carry the full rated load of two should this emergency occur at about the time the substation is due for reinforcement. At any other time the load per transformer will be less.

However there is also one note of caution necessary. In the UK maximum system loading tends to occur during the winter months when ambients are generally lowest. This is the reason why the permitted hot spot temperature is quoted at an ambient temperature of 5ºC. If a CER transformer is called upon to perform an emergency duty during the summer months, the hot spot temperature must be carefully monitored to ensure that this does not exceed the design figure of 115ºC or 125ºC as appropriate. For each 10ºC that the ambient temperature exceeds 5ºC it will be necessary to reduce the maximum rated load by about 7 percent. It must be remembered that a CER transformer has no overload capability beyond its emergency rating and it is designed for operation at one ambient of 5ºC maximum and not a variable one as set out in EN 60076. FIG. 20 shows the core and windings of a 12/24 MVA, 33/11.5 kV CER transformer and FIG. 21 shows a typical CER installation.

FIG. 20 Core and windings of 12/24 MVA, 66/11 kV, three-phase, 50 Hz CER system transformer (ABB Power T&D Ltd)

FIG. 21 Two three-phase 12/24 MVA, 33/11 kV, 50 Hz continuously emergency rated system transformers fitted with single compartment high-speed tapchangers, installed on site (ABB Power T&D Ltd.)

Testing of CER transformers

It will have been noted that in the foregoing description of the CER transformer that the rating is based on achieving compliance with a specified hot spot temperature rise, this despite the fact that, as explained in Section 4.5, hot spot temperature cannot be measured. It is thus difficult to ensure by testing that the manufacturer has complied with the specification as regards achieving the specified hot spot temperature.

CER specifications usually approach this problem in two alternative ways.

The first involves very extensive monitoring on temperature rise test by the use of fiber-optic probes or similar devices placed in what are adjudged by the designer as being the critical locations. In discussing this as an option it must be recognized that, as stated above, when the concept of CER transformers was first developed it was intended that a high degree of standardization of CER transformers should be adopted with possibly several hundred identical units being built. In these circumstances elaborate monitoring of a prototype on temperature rise test is an economic proposition. (Even though, at that time, fiber-optic equipment was not available as a means of dealing with the voltage problems.) When designs are far from standardized then more conventional methods have to be adopted, which means the second alternative approach.

It is not too difficult to take measurements of top oil temperature where it emerges from the windings rather than at the tank outlet to the coolers where it is likely to have mixed with cooler oil which has not passed through the windings.

Once again, nowadays fiber-optic probes can be of assistance, the merit is that only a small number of locations, not actually within the windings require to be monitored. If this is done it is possible to obtain a reasonably accurate estimate of mean oil temperature in each winding (because the inlet oil temperature to the windings is fairly easy to establish) and from this and a measurement of the temperature rise by resistance of the individual windings, it is possible to obtain a fairly accurate average gradient for each. It is still necessary, having established the average gradient, to make an estimate of the maximum gradient. This needs the designer's knowledge of the design, but, since the hot spot factor derived in this manner, that is based on a gradient derived from a fairly accurate knowledge of the oil temperature within the windings, is only likely to be two or three degrees, even a large percentage error in this will not greatly affect the accuracy of the estimate of hot spot temperature. Hence, by one means or another, compliance with guarantees can be fairly clearly demonstrated.

For the ONAN condition top oil temperature rise and winding temperature rise should be measured in the conventional manner and these should comply with EN 60076.

Top of Page

PREV.   NEXT   Guide Index HOME