Industrial Power Transformers -- Transformer enquiries and tenders [part 2]

Home | Articles | Forum | Glossary | Books

cont. from part 1

Other fittings

In addition to the foregoing fittings, the following are likely to be required:

(i) A thermometer pocket mounted in the top cover in vicinity of the hottest oil.

(ii) In the case of transformers having separate cooler banks, two further thermometer pockets, one in the outlet pipe to the cooler and one in the inlet pipe from the cooler. All thermometer pockets must be provided with captive weatherproof screwed caps.

(iii) An oil sampling device mounted 1 m above plinth level.

(iv) Jacking plates, haulage eyes and lifting lugs.

(v) A grounding connection point on the transformer tank and an additional such point on any separate cooler banks.

Terminations and accommodation for current transformers

This section of the specification should provide details of the terminations required. If air bushings are provided for either HV or LV terminations, any special requirements relating to terminal palms should be identified and any requirements concerning shed profile. If co-ordinating gaps are to be provided this should be made clear including details of type, and range of adjustment where appropriate and gap settings required on initial delivery to site.

If cable boxes are to be fitted then full details including number, type, size and rating of cables should be given, type of terminations to be used and whether cables will enter boxes from above or below. Requirements regarding gland plates should be identified. Ideally individual gland plates should be provided for each cable. The specification should also make clear who is to supply cable glands.

Any neutral termination arrangements should be described including details of physical arrangement of neutral connections and any neutral current transformers.

Full details of line current transformers should be provided. These should be accommodated in turrets mounted above the tank cover so that, should it be necessary for a CT to be removed at some time in the future, then this can be done without lowering the oil in the main tank below the top of the core.

Terminal connections for CT secondaries should be provided in weatherproof terminal boxes on the outside of each turret.

Marshalling cubicle

The tenderer will normally be asked to include in his tender for the supply of a marshalling cubicle to which he should extend connections from all equipment mounted on the transformer. The specification should identify all items to be included in this cubicle and indicate whether this is to be separate free standing or whether a tank-mounted cubicle is acceptable.

Whether tank mounted or free standing, the cubicle should be of outdoor weatherproof construction with a cover designed to shed water and a front opening door or doors. No equipment in the cubicle should be mounted more than 1.8 m above plinth level for ease of operator access.

The marshalling cubicle should provide accommodation for the equipment listed below. Equipment for each common function should be grouped together and each item should be labeled to identify its function in accordance with the appropriate circuit diagram:

(i) Temperature indicators, test links and ammeters for winding temperature circuits.

(ii) Interposing current transformers associated with the unit main protection system. Where these are not required to be installed initially, a space should be reserved for any future requirements.

(iii) Control equipment for forced cooling, where appropriate, including local/ remote and duty/standby selector switches, isolators, fuses, motor contactors and overload devices.

(iv) Any transformers required for the provision of 110 V AC control supplies.

(v) Terminal blocks to accommodate all interconnecting multicore cabling associated with alarms and tripping equipment and current transformer secondary circuits. The provision of some spare multicore terminal blocks to allow for future extensions/modifications is advisable. For most purposes about 20 of these will be adequate. It is also worthwhile reserving the right to approval of the terminal blocks to be used or, provided the requirement is qualified by the use of the description 'or equivalent,' a particular type reference may be quoted. It should be recognized, however, that the transformer manufacturer is unlikely to have a great deal of control over the type of terminal blocks used in proprietary equipment such as tap changer drive and control cubicles.

(vi) Sectionalized gland plates to accommodate all incoming and outgoing cables with sufficient allowance to meet any future additions.

All equipment must be mounted so that terminals are accessible for testing purposes but shrouded to prevent danger to operators.


It is important to recognize that the cubicle must withstand all weather conditions and provide protection for its contents against deterioration for many years. The cubicle must therefore be designed to shed water and should be free of features which may trap debris. The cover and sides of the cubicle should preferably not be pierced by fixings. All parts should have a non-corrodible

finish and ferrous parts should be covered by the requirements for painting and weather protection discussed above.

The cubicle needs to be adequately ventilated to ensure free air circulation over all equipment and a heater should be fitted to prevent condensation.

Doors should be provided with fastenings having integral handles and pad locking facilities. A sensible size of padlock shackle needs to be accommodated, say 9 mm diameter. Doors must be adequately weatherproofed.

Lighting and socket outlet

The cubicle should be provided with internal lighting arranged to illuminate all the internal equipment as evenly as possible so that an operator can work during darkness and consideration should be given to the provision of a 240 V socket outlet to provide a power supply for any portable test equipment which it might be required to operate on the transformer plinth.

Lamps should be to EN 60064 extended life (2500 hours), mounted in a heat resisting lampholder to EN 61184 and controlled by a door operated switch.

The lamp should be suitably protected to avoid accidental breakage or touch.

If a socket outlet is to be provided this should be of an appropriate pattern, protected by RCD, and mounted on the outside of the cubicle.

It will be necessary for the purchaser to provide an incoming 240 V sup ply to the cubicle to supply the lighting, heating and socket outlet as well as a power supply which may be required for any forced cooling equipment.

Interconnecting cabling

The type and standard of interconnecting cabling required between equipment mounted on the transformer and the marshalling cubicle may be specified unless the transformer manufacturer is to be given a free hand in this. Preferably 600/1000 V armored cable with stranded copper conductors and producing low smoke and corrosive gases in the event of fire in accordance with, for example BS 6724, or equivalent, should be used and it is desirable that the enquiry document should make clear who is to install, gland and terminate this. This is normally done by the transformer manufacturer. It is usual to specify a minimum nominal conductor cross-sectional area of 2.5 mm2 for the cable cores to ensure that these are mechanically adequate to withstand the duty to be imposed upon them in particular the vibration generated by the transformer, although nowadays smaller cores are considered acceptable for transmission of signals to SCADA systems.


It is rare for any transformer, other than a small distribution transformer, to be tested simply to EN 60076 and, even when this is considered adequate, there are frequently options which need to be identified such as whether the rated lightning impulse withstand voltage for windings of Um up to 36 kV should be the higher or lower values listed in Table 2 of EN 60076-3 or, for values of Um of 100 kV and over, which of the alternative power frequency and impulse withstand voltages, given in the same table, are to apply. Consequently it is desirable to carefully consider all the testing which will be required and to set this out clearly so that no misunderstandings or omissions will occur.

If any doubt exists as to the extent of testing which should be specified, reference should be made to Section 5.

It is also necessary to decide whether the option is to be retained to witness any or all of the tests. To provide for this option is likely to involve the manufacturer in some additional costs and might also limit his flexibility of operation at the testing stage, so it is likely that he will wish to include some allowance for this in the contract price. On the other hand, should any problem occur during testing, it can greatly assist constructive discussion and resolution of the problem or of any proposed remedial measures, to have been represented at its occurrence. This is particularly the case for testing which might to some extent be subjective in its interpretation, such as the measurement of partial discharges during an overpotential test, the examination of test records during a lightning impulse withstand test or the measurement of sound power.

Type testing

As indicated earlier, most transformers will be designed to meet a particular contract and are likely to differ from other designs previously manufactured and tested at least in respect of losses, impedance or tapping range, so that it is likely that some type testing will be required on any new contract. Type testing is considered at some length in Section 5, however, for the convenience of the reader the tests normally considered to be type tests, or special tests, are the following:

• Temperature rise test.

• Lightning impulse test including chopped waves and switching surge tests where appropriate.

• Impedance on all tap positions -- and may also include load loss on all tap positions.

• Zero-sequence impedance.

• Sound power level measurement.

• Short-circuit tests.

Consideration should also be given to the strategic importance of the transformer to be purchased and to whether this might justify any additional or enhanced testing as discussed in Section 5.3. It should also be considered as to whether the taking of oil samples for dissolved gas analysis as part of works testing will be required. This procedure can greatly increase the confidence which can be obtained from works testing and is also discussed in Section 6.7.

The appropriate times for the taking of samples are as follows:

• Before the commencement of final works testing.

• On completion of temperature rise tests.

• On completion of impulse testing.

• On completion of power frequency dielectric testing.

Routine tests

Routine tests can be simply specified as being in accordance with EN 60076 or other appropriate standard. If it is required that measurement of partial discharge is to be carried out during the short-duration power frequency with stand test, this requirement should be identified and the method of carrying out the test indicated, including the acceptance criteria (see Section 5).

Tank vacuum and leakage tests

Most manufacturers will carry out a leakage test on a transformer tank since it will be inconvenient for them if the tank should leak either during works tests or on site during the maintenance period. For transformers which are to be vacuum filled on site it will be necessary for the tank to be designed to withstand full vacuum so that in order to ensure that there are no problems on site, a manufacturer must satisfy himself in the factory with regard to the tank vacuum capability. It is, however, preferable that these aspects are not left to chance and that the technical specification includes tests for tank vacuum capability, where appropriate, and for freedom from leakage. Tests for tank vacuum capability and leakage are detailed in Section 5.1.


Following the issue of the transformer enquiry, at a date nominated by the purchaser, a number of tenders will be received. These should be left unopened until the declared time has passed and then formally opened. The pricing information for each tenderer should be extracted and logged together with the relevant information for any options and prices for work deemed extra by the tenderer. The tenders should be kept by one person who is responsible, until the contract is placed, for keeping them confidential and ensuring they are kept locked away when not being worked on. If each of the tenders is technically fully compliant with the specification then deciding which tender to accept is simply a matter of deciding which has the lowest cost. It is rare, however, for the tender selection process to be such a simple one, so that a careful assessment will need to be carried out to determine which, if any, of the tenders should be accepted.

Tenders may have aspects for which they are not technically compliant.

It may be that some of the tenderers wish to apply commercial conditions which could possibly result in additional costs. It is fairly certain that every tender will be for a different combination of no-load and load losses. If the program timescale is short it is possible that a tenderer might not be able to meet the required delivery date. If all other aspects of this tender make it attractive it might be appropriate to consider the cost implications of delaying completion of the project or of rescheduling construction.

This section however is restricted to a description of the procedure for making a technical assessment of the received tenders, although clearly the final decision concerning the placing of a contract will involve selecting the most acceptable combination of commercial, technical and program aspects.

Initial selection process

It can be quite common nowadays to receive as many as seven or eight tenders for even a fairly modest project. Making a detailed study of eight tenders can be quite a time-consuming process. The first step therefore is to reduce the number of tenders to be considered in detail to a short-list of three or four.

This will normally be done by an examination of the costs. For each tender the total cost can be calculated; this is the sum of the tender price plus the cost of the losses plus the cost of any special commercial aspects associated with the offer. Possibly, despite specifying a requirement for a 5 year guarantee, say, one of the tenderers might only be prepared to offer a 1 year guarantee period. Possibly some of the tenderers will be from overseas so that monitoring of the contract will be more costly, requiring some extra allowance to be made. Another might require a different schedule of stage payments, making the financing costs greater. Table 8.2 shows a typical initial tabulated series of costs for six tenders taking account of such factors. It will be seen how the order of preference can be significantly affected by carrying out this exercise.

Although on price alone tender A is the lowest, the extra cost of supplying the significantly higher losses during the operating life of the transformer make this less attractive overall and tender D, which on initial examination might appear considerably higher, appears to be the most attractive. The method of assessing the cost of the losses will be considered in the following section.

Having carried out this initial examination of the tenders to arrive at the position shown in Table 8.2, the next stage is to look at those which appear more attractive in a little more detail. Of those examples listed in the table, tenders A, C and D are worthy of consideration in greater depth.

Table 2: Typical price and loss summary for transformer tenders

Tenderers will frequently submit a tender letter in which they will highlight those aspects of their bid which they feel might require clarification. This letter might also identify aspects of the enquiry document which they did not consider to be entirely clear in its requirements and it will explain any assumptions which they felt it necessary to make. They will also probably include a detailed description of the transformer offered, including those aspects of their design and manufacturing processes, as well as their QA procedures, which they feel renders their bid worthy of extra commendation, and, of course, they should have completed the tender schedules included in the enquiry document.

All of these materials provide a great deal of information about the transformer offered and must be studied in detail.

In making this study the objective must be to obtain answers to the following questions:

• Are there any statements made in the covering letter, descriptive material or tender schedules which suggest that the equipment supplied will not be in accordance with the specification?

• Are the impedance values given in the tender schedules in accordance with those specified flare the impedances on extreme tap positions, including any possible tolerances, within acceptable limits? If zero-sequence impedance is important, is the value offered acceptable?

• Has all the specified testing, in particular type testing, been included in the offer?

• Has the waiving of any type testing been claimed? If so, is the supporting evidence included and is it acceptable?

• Has the tenderer taken due account of any special requirements included in the specification, for example special overloading capability?

• Will the transformer fit in the site?

• Has the tenderer included for all the specified fittings, marshalling cubicle, valves, anti-vibration mountings, etc.?

• Does the pattern of terminations offered comply with the specified requirements with regard to, for example, bushing shed profile, palm configuration, type of cable boxes?

• Has the tenderer included all the special descriptive information requested in the enquiry document, for example the measures incorporated to allow for a high level of harmonics in the load, or to cater for frequent severe overloads?

• Does the offer meet any specified noise level requirements, including the effect of a noise attenuation enclosure where appropriate? If a noise enclosure will be required, has it been included in the tender price?

• Has the tenderer included for all the necessary site work, including delivery and site erection?

Occasionally the descriptive material provided by one tenderer can raise questions in relation to the other tenders, for example some of the tenderers might comment that a specified overload duty at 10ºC will require an increase in the rating at normal EN 60076 ambient. This then raises the question as to whether a tenderer who has made no comments at all in relation to the specified over load duty has taken this into account in preparing his design. Similarly, it is sometimes the case that setting out the information provided by the tenderers in their completed schedules of technical particulars will highlight an anomaly in some of the data provided by one of the tenderers and raise the question as to whether his offer is in compliance with the specification.

In the example of the embedded generator mentioned in the previous section of this section, the transformer rating was specified at an ambient temperature of 10ºC but it was proposed that tenderers should be asked to quote the rating of the transformers offered at normal IEC ambients. It would be expected that the reduction in rating resulting from the increase in ambient temperature from 10ºC to an annual average of 20ºC and a daily average of 30ºC would be quite modest and very nearly the same for all tenders, but the one for which the reduction is least might be taken as an indication that this is the design which is least stressed thermally, and, as indicated elsewhere in this work, lowest thermal stress is likely to lead to longest life. Such considerations would only, of course, be relevant in differentiating between tenders which were very similar in other respects.

This careful scrutiny of the short-listed tenders will probably result in the need to make some adjustments to the initial assessment of costs as given in the example of Table 8.2. It is quite common for tenderers not to include for type testing in their total tender sum, even though they will probably indicate the price of the tests themselves. It might be the case that one tenderer can meet the specified noise level without the use of a noise enclosure, whilst Table 3 Amended price and loss summary after study of tender descriptive material others cannot. The figures comprising Table 2 for the three short-listed tenderers can thus be amended as shown in Table 3 so that a preferred tenderer will be identified.

Tender questionnaire

Often despite all the descriptive material provided as well as the information in the tender schedules, it will be the case that the purchaser does not have the confidence to place a contract with the preferred tenderer without some further investigation. In these circumstances it may be appropriate to issue a questionnaire to one, or more, preferred tenderers.

Sometimes, particularly in the case of tenders for smaller transformers of ratings up to perhaps a few tens of MVA where manufacturers are keen to limit their costs for the preparation of tenders, the extent of descriptive material may be very limited. In this case, unless there is definite evidence that some aspect of the specification will not be met, or description which has been specifically requested in the specification is not provided, it must be assumed that the tender is compliant. There is no need for questions to be raised simply because the tenderer has not written at great length about every one of the design features.

Frequently questions can arise because a manufacturer provides too much descriptive material and in the relatively short time that he has for the preparation of his tender he has not had chance to thoroughly check to ensure that no conflicting statements are included.

Although the response to questions can occasionally result in additional costs, questions should be phrased in such a way as to avoid inviting these, for example the tenderer should generally be asked to confirm that his offer includes the specified feature which is in doubt. After consideration of the response to the questionnaire the effect on the price comparison should be finally assessed before placing the contract. If two or more of the tenderers assessed prices are sufficiently close that the response to questions might change the order of preference, then questionnaires should be issued to all of these.


When the purchase of a transformer is considered, as with most other items of plant or equipment, there are two aspects to be taken into account:

(1) The initial capital cost.

(2) The running cost -- which in the case of a transformer is the cost of supplying the losses.

In the typical tender assessment exercise discussed in the preceding section, notional values were placed on each kilowatt of the guaranteed losses for the transformers tendered as a means of comparison of the tenders on a common basis. That is, a cost was assigned to the value of 1 kW of no-load loss and also to the value of a kilowatt of load loss during the operating life of the transformer. In the example no-load loss was costed at a considerably higher value, £3000/kW, than load loss at £650/kW, and although there was no mention in the example of the type of transformer being considered, this could be the case for a typical transmission transformer operating in a multiple transformer substation in the part-loaded condition, so that in the event of losing one transformer in the substation, the remainder must be capable of carrying the total substation load without becoming overloaded. In addition, the daily load cycle has a daytime peak and a very much lower value at night. As a result the transformer spends much of the time at less than half load, and the average load losses are less than one-quarter of their magnitude at nominal rated power, which is the rating for which the load losses are quoted and guaranteed by the manufacturer.

Whilst this illustration explains why the cost of the load losses is so much less than the no-load losses, it does not explain how their actual value is derived. To do this it is necessary to examine the subject a little further and assess the likely cost of a kilowatt of loss over the lifetime of the transformer.

The simplest method would be to calculate the total energy consumed in losses over, say, a 25 or 30 year life and cost this at today’s energy price. This calculation can be worth carrying out if for no other reason than the fact that the answer can be quite surprising.

Typical cost of losses to industry

Even when taking such a simple approach, it is worthwhile attempting to carry out the calculation as carefully as possible, that is the load factor should be estimated as closely as possible and factors such as time of day rates and maximum demand charges need to be taken into account.

Example 1. Consider a typical small factory which has two 11/0.415 kV transformers. The factory operates for 50 weeks of the year and during this time the plant is running 10 hours/day, weekdays only. The transformers are energized 24 hours/day, 50 weeks of the year, to provide power for lighting but their only significant load is whilst the plant is running.

It is often the case that such a factory will be considering the purchase of an additional transformer at the time of extending the electrical system. Perhaps it is planned to supplement two existing transformers because the factory load has grown to considerably more than could be handled by one alone with the other out of service. The purchase might have been initiated by the installation of new plant which will mean that on completion the new installation will have three transformers normally carrying the equivalent of full load for two, that is, each transformer will normally carry two-thirds full load.

The factory operates on a Seasonal Time of Day Tariff, supplied from the local distribution network operator (DNO).

The cost per year (of 351 days) for 1 kW of iron loss is typically thus:

Supply capacity charge, say £18 Maximum demand charge - winter p.m., say 40 Night units: 23.30-06.30, 7 hours daily, 7 _ 351 _ 2457 units at, say, 2.5p/hour 61

Weekend units: 06.30-23.30, 17 hours/day for 49 weekends,

17 _ 98 _ 1666 units at, say, 4.5p/hour 75 Evenings Monday to Friday: 20.00-23.30, 3.5 hours/day

3.5 _ 253 _ 885 units at, say, 4.5p/hour 40 Days November to February: 06.30-20.00, 13.5 hours/day

13.5 _ 86 _ 1161 units at, say, 6.5p/hour 75 Days rest of year: 06.30-20.00, 13.5 hours/day

13.5 _ 167 _ 2255 units at, say 5p/hour 113 Total, per year £422

Over a 25 year lifetime this would amount to a quite surprising £10 550/kW.

The cost of 1 kW of copper loss can be calculated in a similar manner:

Supply capacity charge (transformer at 66.6% load) _

0.443 _ £18 £7.98 Maximum demand charge _ 0.443 _ £40 17.72

Days November to February, 10 hours/day during period 06.30 to 20.00

0.443 _ 10 _ 86 _ 381 units at 6.5p 24.76

Days rest of year, 10 hours/day, same time of day

0.443 _ 10 _ 167 _ 740 units at 5p 37.00

Total, per year £87.46 or approximately £2186/kW over a 25 year lifetime of the transformer.

Most accountants would not accept the above method of assessing lifetime cost, probably rightly so when a life of 25 years or more is expected, since costs incurred a long time ahead can be expected to have been eroded by inflation, or, alternatively to meet a commitment some years ahead cash can be set aside now which will accrue interest by the time the payment is due. An alternative viewpoint is that these losses will continue to have the same magnitude and the cost of energy will probably have increased roughly in line with inflation.

Generally the accountants view prevails so that the cost of making provision for the lifetime cost of losses is expressed in terms of the sum which must be set aside now to pay for these. This can be calculated from the following expression:

C _ {a(1 _ b) n

_ b _ a}/{(1 _ b) n

_ 1} (eqn 1)

where C is the cost per £ annual cost of losses.

a is the rate of interest payable for loans at the date of purchase (expressed on a per unit basis).

b is the rate of interest obtainable on sinking funds (expressed on a per unit basis).

n is the estimated lifetime of the transformer in years.

Typically, and for the purpose of illustrating this example, 'a' might be taken as 9 percent for a large organization seeking a long term loan and 'b' as 7 percent. For a value of 'n' equal to 25 years 'C' is then 0.1058 Hence the capitalized value of no-load loss is 422/ 0.1058 _ £3988/kW and the capitalized value of load loss is 87.46/0.1058 _ £827/kW.

These are the values for losses which it would be reasonable for an organization such as the one described to use in its assessment of tenders for an additional transformer. The tenders and the assessment of them might typically be as in the following example.

Example 2. The following tenders have been received for a 1000 kVA 11/0.415 kV transformer:

Capitalizing at £3988/kW for no-load loss and £827/kW for load loss gives the following assessed costs:

From this assessment, it can be seen that the lowest loss, highest priced option, offered by manufacturer B, provides the factory with the lowest lifetime cost.

It should be noted, however, that the lifetime cost of a transformer from manufacturer C is only just over £340 greater than one from manufacturer B and buying this would save nearly £3000 now. It is therefore worthwhile carrying out a sensitivity check on the assumptions made. If the new transformer were only loaded to 60 percent of its capacity and not 66.6 percent as assumed, what effect would this have on the most economic option? To check this is a fairly simple matter. Returning to the calculation, above, of the cost of load loss, substitution of a load factor of 0.60 instead of 0.666 would give a load loss factor of 0.36 instead of the figure of 0.443 used in the calculation. This would reduce the annual cost of 1 kW of load loss at name plate rating to (0.36 _ 87.46)/0.443 _ £71.07/kW. This in turn reduces the load loss capitalization value to 71.07/0.1058 _ £672/kW.

Repeating the loss evaluation with this revised load loss capitalization value gives the following figures: and it can be seen that manufacturer B remains the most economic option.

Provided the factory management is confident that the load factor on the new transformer is not likely to fall below about 0.60, they can justify the expenditure of the additional £3000 initially.

Test discount rate

In the above example the lifetime cost of losses has been converted to an equivalent capital sum per kilowatt which will meet the lifetime costs. The purchaser may therefore either spend up to that additional sum at the outset for each kilowatt reduction in losses or alternatively set it aside to pay for the losses during the transformer lifetime. Both alternatives have the same weighting and there is therefore no constraint on the spending of extra capital initially, provided it will produce at least an equivalent saving in losses. The concept of test discount rates (t.d.r.) was originally applied to publicly owned utilities in the UK some years ago to control capital spending so as to ensure that extra expenditure was only incurred if it could produce real returns. The practice is still widely used by the now privatized utilities and the t.d.r. applied has varied between 5 and 10 percent over the years since its inception. The figure of 5 percent being normally used but this can be increased to 10 percent at times when cash is particularly tight. A 5 percent t.d.r. requires that any additional capital spent over and above that necessary for the basic scheme should show a return of 5 percent.

Applied to a capitalizing rate of C per £ as derived in the example above, a t.d.r. of r percent has the effect of multiplying the cost of losses by a factor k, where

k _ C/(C _ r/100) (eqn. 2)

The effect of a 5 percent t.d.r. on the capitalized cost of losses in the above example is thus

k _ 0.1058/(0.1058 _ 0.05)

Hence k _ 0.679

Cost of no-load loss thus becomes £2707/kW and cost of load loss £553/kW and it is now the case that greater initial expenditure to obtain lower losses will only be allowed if a real return can be obtained from that extra expenditure. Use of a t.d.r. greater than 5 percent reduces k and hence reduces the value placed on energy savings still further. If the management of the factory in the above example were to stipulate the use of a 5 percent t.d.r., the decision would clearly be in favor of saving money at the outset, as can be seen by repeating the above assessment process using these new loss values:

... and there is now no doubt that the additional £3000 of expenditure over the cost of manufacturer C's tender cannot be justified.

The above example illustrates how the cost comparison process can be rationalized. It must be recognized however that the process is very greatly influenced by basic policy decisions such as the level of t.d.r. applied or whether any t.d.r. is applied at all. Thus, if a purchaser wishes to sway his assessment towards low first cost, he can do so and conversely, if he wishes to invest more initially to provide energy efficient plant he can ensure that his procedures make this more likely.

Transformers for an electricity supply network

Although it is important to recognize that any system for assessment of losses such as that used in the above example can never be regarded as absolute, since no decision made today can take account of long term changes in energy costs or of availability of investment capital, the industrial user of the example has the benefit of a known tariff structure and a fairly constant works system loading to enable him to make his estimates fairly easily of the magnitude and cost of losses. The operator of an electricity supply network is faced with a slightly more complex assessment process. For any new transformer installed on this network there will almost certainly be a daily cyclic variation in load as well as an annual summer/winter variation. In addition, there is the possibility of a load growth cycle resulting in the loading on a transformer after some years in service being greater than that applied on initial commissioning. All of these factors need to be taken into account in the loss capitalization process.

In the UK, since privatization, the distribution companies have been faced with the added complication of having to buy most of their energy from the energy market, thus encountering the added unpredictability of this system.

The situation which existed previously was that they were simply required to pay the CEGB's Bulk Supply Tariff (BST) and were, in addition, provided with long term predictions as to its likely magnitude. Whatever the actual method of payment, however, the cost of energy to the distribution companies has two main components and these remain, thus appearing as costs to be assigned to any source of losses within their network. These components represent a capacity charge and an energy charge and both are required to cover the marginal costs in meeting incremental increases in the demands of the system. In the days when the BST was in operation, these charges were identified as specific components of the BST. Now the distribution companies must rely on their own experience of energy trading and their own predictions of future trends in deciding the values to be placed upon them.

The capacity charge is a reflection of the long term cost of providing the additional power (as distinct from simply energy) required to supply the additional losses. This will involve the cost of increasing the capacity of the distribution companies own network as well as the cost of the additional generation and transmission plant which must be installed. The magnitude of the generation capacity charge is likely to be dependant on the requirement for the capacity at some specified critical time or times of the annual and daily load cycle, in the UK usually a weekday in December or January between the hours of 17.00 and 17.30 when the system loading is likely to be near to its highest.

The energy charge represents primarily the cost of the consumable element of supplying the losses, namely the additional fuel cost.

No-load losses are generally assumed to be constant and incur capacity and energy charges on this basis.

Load losses are, of course, variable according to the magnitude of the load squared. The daily and annual load loss factors are nowadays usually calculated by means of a computer program which can accurately reproduce the daily and annual patterns of load variation, making due allowance for the 'load squared' relationship.

Figure 2 Calculation of r.m.s. load on a substation

It is also possible to allow some diversity for load losses. The diversity factor is defined as the sum of the load losses for all substations divided by the effective load losses taken from the supply. The value of the diversity factor will not differ greatly from unity. It is usually of the order 1.1-1.2.

All of these factors are then applicable to the value of load loss energy charge and the computer program can be arranged so that it has the necessary information concerning the critical times to enable it to also compute the load loss capacity charge.

The load growth cycle can be relatively easily accommodated and is best illustrated by means of an example.

Example 3. Typically a distribution network may consist of a number of 'two-transformer' substations in which two similar units share the load equally. The maximum peak load for this substation is limited by the capacity of the switchgear. Consider an 11 kV two-transformer substation having 1200 A switchgear, giving it a peak capacity of 23 MVA. In the event of a failure of one of the transformers, the other will be limited to 23 MVA.

Now suppose the annual rate of growth of load to be 7 percent. After 6 years the substation total load will have risen to 1.5 times the initial load, since 1.076 _ 1.5, but, if the load is not to exceed 23 MVA, then this must be the peak load at the time for reinforcement, and the initial load must thus be 23/1.5 _ 15.3 MVA, or 7.65 MVA per unit.

The load loss growth factor for the substation will, however, be equivalent to the RMS load over the 6 year period which can be calculated by reference to Fig. 8.2. The initial load on the substation is S and the load after n years is

This is the r.m.s. loading on the substation and the RMS load on each transformer over the 6 year period is 9.5 MVA.

In the UK at the present time it is relatively uncommon to uprate a two transformer substation with a third transformer due to the existing extent and interlinkage of the network. It is more usual to reinforce two two-transformer substations with a third substation or three two-transformer substations with a fourth substation. Still limiting the peak rating of one transformer to 23 MVA and based on a 7 percent annual growth rate, the r.m.s. loadings on the sub stations are:

two to three two-transformer substations _0.828

three to four two-transformer substations _0.865

In this situation it would probably be convenient to capitalize load losses for all purchases of transformers for this duty at an r.m.s. load factor of 0.85 times the appropriate factors resulting from daily and annual load cycles.

Capitalization values

From the foregoing it is possible to calculate typical capitalization values which might apply when purchasing transformers for the reinforcement of a distribution network. It must be stressed that the calculations are intended to illustrate how the process might be carried out and are not representative of the method or values used by any particular DNO.


It can be seen that the expensive low loss option is not the most attractive to the distribution company as it was for the factory, despite the fact that the loss savings will accrue for 40 years, and although the tenders from manufacturer B and C are very close in overall assessed costs, there is less incentive to repeat the sensitivity study as in the case of the factory transformer, since the option having the lowest assessed cost is also very attractive on first cost.

Top of Page PREV.   NEXT   Guide Index HOME