– Page 1 of 57 – 40130017-TDC 02-24226 ENERGY SAVING IN INDUSTRIAL DISTRIBUTION TRANSFORMERS May 2002 Authors: W.T.J. Hulshorst J.F. Groeman KEMA Copyright © KEMA Nederland B.V. Arnhem, the Netherlands. All rights reserved. This document contains proprietary information that shall not be transmitted to any third party without written consent by or on behalf of KEMA Nederland B.V. The same applies to file copying, wholly or partially. ECI and KEMA Nederland B.V. and/or its associated companies disclaim liability for any direct, indirect, consequential or incidental damages that may result from the use of the information or data, or form the inability to use the information or data. KEMA report reference 40130017-TDC 02-24226A – Page 2 of 57 – 40130017-TDC 02-24226 TABLE OF CONTENTS MANAGEMENT SUMMARY ......................................................................................................4 1. 1.1 1.2 1.3 INTRODUCTION .................................................................................................................5 Background .......................................................................................................................5 Project objective and scope ............................................................................................6 Methodology ......................................................................................................................6 2. 2.1 2.2 2.3 2.4 2.5 2.6 2.7 INDUSTRIAL ELECTRICITY SUPPLY IN EUROPE..........................................................7 Energy consumption by industry sector ........................................................................7 Transformer population in industry ................................................................................8 Large office buildings .................................................................................................... 10 Data hotels ...................................................................................................................... 10 Failure statistics of transformers in the field .............................................................. 11 Costs of electricity for industry in Europe .................................................................. 11 Environmental issues .................................................................................................... 12 3. 3.1 3.2 3.3 LOSSES IN TRANSFORMERS ....................................................................................... 14 No Load losses .............................................................................................................. 14 Load losses .................................................................................................................... 15 Extra losses due to harmonics ..................................................................................... 16 3.3.1 3.3.2 3.3.3 3.3.4 Harmonic components of the transformer load .................................................................................... 16 Extra losses due to harmonics ............................................................................................................ 17 Characterisation of non-linear transformer load: K-factor (US practice) ............................................... 18 HD538 Transformer derating due to harmonics: Factor K (Europe) ..................................................... 19 3.4 Thermal ageing of transformers ................................................................................... 21 3.5 Loss evaluation .............................................................................................................. 22 3.6 CO2 reduction and emission trading............................................................................ 24 4. 4.1 4.2 4.3 TYPICAL INDUSTRIAL TRANSFORMER DATA ........................................................... 25 HD standard data for distribution transformers up to 2500 kVA .............................. 25 Data for oil-immersed transformers between 100 and 1600 kVA ............................. 28 Data for transformers between 1000 and 4000 kVA ................................................... 29 5. CASE STUDIES ............................................................................................................... 31 5.1 Case study 1: a large company in the iron-steel sector ............................................ 31 5.1.1 5.1.2 Iron-steel: a big consumer of electricity ............................................................................................... 31 Conclusion .......................................................................................................................................... 33 5.2 Case study 2: A large company in the non-ferrous metals sector ........................... 34 5.2.1 5.2.2 Non-ferrous metal sector: replacing older transformers ....................................................................... 34 Conclusion .......................................................................................................................................... 35 5.3 Case study 3: a paper and pulp company ................................................................... 36 5.3.1 5.3.2 Paper industry: energy savings based on practice .............................................................................. 36 Conclusion .......................................................................................................................................... 37 5.4 Case study 4: a chemical industrial enterprise .......................................................... 38 5.4.1 5.4.2 Chemical industry: energy savings by placing energy efficient transformers ........................................ 38 Conclusion .......................................................................................................................................... 39 5.5 Case study 5: a large data hotel start-up .................................................................... 40 5.5.1 5.5.2 5.5.3 5.5.4 5.6 5.7 5.8 5.9 Data hotels: a quickly growing sector .................................................................................................. 40 Evaluation of more efficient transformers with the same rating ............................................................ 41 Evaluation of smaller transformers ...................................................................................................... 42 Conclusion .......................................................................................................................................... 43 Case study 6: An office building .................................................................................. 43 Case study 7: Reliability and availability..................................................................... 44 CO2 emission trading .................................................................................................... 46 Sensitivity of the input parameters .............................................................................. 46 – Page 3 of 57 – 40130017-TDC 02-24226 5.10 Total energy saving potential by efficient transformers in the industrial sector .... 50 6. CONCLUSIONS AND RECOMMENDATIONS ............................................................... 52 6.1 Conclusions ................................................................................................................... 52 6.1.1 6.1.2 6.1.3 6.1.4 Transformer population ....................................................................................................................... 52 Transformer purchasing policies ......................................................................................................... 53 Loss evaluation, including the effects of harmonic pollution ................................................................. 53 Energy saving potential ....................................................................................................................... 53 6.2 Recommendations ......................................................................................................... 54 6.2.1 6.2.2 6.2.3 6.2.4 Purchasing policies and promotion...................................................................................................... 54 Standardisation: HD428 and 538 ........................................................................................................ 54 Savings potential of special transformers and high-voltage motors...................................................... 55 Optimised system voltage in industrial networks ................................................................................. 55 REFERENCES ........................................................................................................................ 57 – Page 4 of 57 – 40130017-TDC 02-24226 MANAGEMENT SUMMARY Industrial companies account for approximately 1000 TWh/a or half the total electricity consumption in Europe. The total European population of industrial distribution transformers is estimated to be 100.000-150.000 units, totalling to 100-150 GVA installed, total losses being almost 10 TWh/a. A study has been performed to explore the potential for energy savings and CO 2 emission reduction in industries and offices by using energy-efficient distribution transformers. Based on interviews with representatives from several large Dutch industrial companies, the population was roughly characterised and case studies have been carried out. Comparing distribution transformers in industry to those in (public) utilities, some salient differences are present: industrial transformers sizes are typically 1000-4000 kVA, contrary to the public transformers which are in the range 50 – 1000 kVA the average load of an industrial transformer is relatively high (30 – 100% of the rated loading) the newest transformers in industry are often dry-type instead of oil-immersed. Losses in these dry-type transformers are relatively high high levels of harmonic pollution of the load current frequently occur (this causes extra losses and a higher risk of extra ageing) the transformer population is relatively young (up to 30 years). Despite several favourable exceptions, industrial purchasing policies are relatively unfavourable for application of energy-efficient transformers, the purchase price being a dominant factor over loss evaluation. The case studies carried out show that there is a considerable potential for energy saving in industrial distribution transformers.. When ordering a new transformer, industries should pay attention to the loss evaluation, since they can save a lot of money, energy and CO 2 emissions during the lifetime: the extra investment costs of energy-efficient transformers can be earned back. This savings potential is fully economically feasible. On the other hand, neglecting the loss evaluation criteria will almost certainly lead to a wrong investment (the cost of transformer losses will more than offset the lower purchase price of an inefficient transformer). If, as is often the case in industries and office buildings, the transformer load current contains harmonic distortion, e.g. by computers or power electronic drives, transformer losses increase as does the savings potential of energy-efficient transformers. Two activities are proposed to improve the use of energy-efficient distribution transformers in industry: promoting the importance to use loss evaluation, i.e. to use the (simple) loss evaluation formula when specifying and evaluating transformers adapting the international distribution transformer standards to include the transformer population in industry and to reduce the allowed loss levels. – Page 5 of 57 – 1. INTRODUCTION 1.1 Background 40130017-TDC 02-24226 A study carried out in 1999 by the European Copper Institute has revealed that the scope for energy savings and CO2 reduction through the use of energy-efficient distribution transformers in the EU is substantial [1]. The savings potential is estimated at approximately 22 TWh/year, worth EUR 1,2 bn at 1999 prices, mainly for oil-filled distribution transformers in “public” electricity networks. The savings potential can be achieved by application of energyefficient transformers, the extra investment cost of which is earned back by the energy savings. In the European project mentioned, the focus was not directed to large industries and large office buildings, which obtain their electricity from the public medium and high voltage networks. However, much electricity used by those industries is consumed at low-voltage level – the conversion to low-voltage is performed by privately owned distribution transformers. Some differences between these privately owned distribution transformers and their public counterparts are: The load pattern, often constant load The insulation medium, in many cases transformers are of the dry type for fire safety Harmonic components in the loading current Especially related to the latter, these transformers are close to the end-use, and could suffer from significant harmonic loads in IT-intensive facilities. Harmonic loads cause extra heating in conductors, but especially in transformers. This heating effect causes circuit breakers to trip, neutral and phase conductors to heat up to critical flash over temperatures, and premature failure of motors and transformers. This is costly in terms of downtime, loss of production, repair and possible replacement. In many cases, in the private sector, little attention is paid to energy losses in e.g. transformers. Energy-efficient transformers may be more robust against harmonics, and provide more flexibility for future load-growth over the 30 years of transformer lifetime operation. Medium and large industrial electricity user account for over half the Dutch electricity use. In order to estimate the savings potential, it has been proposed to investigate the potential for applying energy-efficient transformers for following industrial sectors: Metal Non-metallic minerals Chemical & petro-chemical Paper Data hotels Large offices (above 10 000 m 2). – Page 6 of 57 – 1.2 40130017-TDC 02-24226 Project objective and scope The project objective was to carry out six technical/economic case studies to estimate the energy saving potential for different industrial sectors by economic application of energy-efficient distribution transformers (copper-wound, dry type or oil-cooled). Description and quantification, where possible, of the soft benefits resulting from efficient transformers (increased reliability, possibility for future load growth, robustness against harmonics) was another objective of the study. 1.3 Methodology The study was based on interviews with several Dutch enterprises to collect data on the energy use and generation pattern, the transformers applied, energy costs and investment evaluation criteria. The aim was to collect sufficient data for at least one representative transformer type for each industrial sector. From these data, KEMA has estimated loss evaluation values and assessed the benefits of changing from a standard dry type distribution transformer (based on CENELEC HD538) to an energy-efficient transformer. A dry transformer was designated energy-efficient at a loss level 20% below HD538 levels, an oil-cooled transformer at HD428 – CC’ loss level. For users which already use oil-cooled transformers, a DD’ level with further reduced losses have been evaluated. Pauwels Trafo Belgium has contributed typical transformer data for the transformers mentioned, based on HD538 required and reduced loss values. For HD428 range transformers, data from the previously mentioned study [1] have been used. The benefits have been expressed in terms of net present value, internal rate of return, energy savings and CO2 emission reduction. Based on extrapolation, the European potential for energy savings and CO2 reduction has been roughly estimated. For CO 2 emissions, each kWh saving was associated with 0,4 kg CO2 reduction, which is the average European value. Two other cases to be considered have been low emissions (0,3 kg CO 2 / kWh) and high emissions (0,6 kg CO2 / kWh). The impact of emission trading on the transformer investment decision has been evaluated at € 10/tonne and € 33/tonne. Finally, a simple spreadsheet model has been developed for dissemination, allowing users to evaluate individual transformer investment decisions by input of: load loss, no-load loss, load pattern, transformer price, electricity tariff, and to evaluate their energy & emission savings potential. – Page 7 of 57 – 40130017-TDC 02-24226 2. INDUSTRIAL ELECTRICITY SUPPLY IN EUROPE 2.1 Energy consumption by industry sector Industry forms a sector that accounts for almost half the final electricity consumption in Europe. The figure below shows OECD electricity consumption data for 1997. Misc 2% Network losses 8% Commercial/public 22% Machinery 4% Food/tobacco 4% Non-metallic minerals 3% Non-ferrous metals Industry 3% 44% Chemical, petrochemical 7% Iron/steel 5% Paper, pulp & print 5% Misc. Ind. 9% Residential 28% Figure 2.1 Final electricity consumption (and network losses) for the European OECD countries 19971 Industry sector INDUSTRY, of which: Iron and steel Chemical/petrochemical Non-ferrous metals Non-metallic minerals Machinery Food/tobacco Paper, pulp and print Other industry TRANSPORT COMMERCIAL AND PUBLIC SERVICES RESIDENTIAL OTHER SECTORS FINAL CONSUMPTION Electricity consumption (TWh) 1044 131 195 88 76 99 95 127 235 70 570 719 62 2464 Source: Energy statistics of OECD countries 1996-1997, IEA, Paris, 1999 Includes following countries: Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Luxembourg, Netherlands, Norway, Poland, Portugal, Spain, Sweden, Switzerland, Turkey and the UK. NON-OECD Europe: Albania, Bosnia-Herzegovina, Bulgaria, Croatia, Cyprus, Former Yugoslav Republic of Macedonia, Gibraltar, Malta, Romania, Slovak Republic, Slovenia, Federal Republic of Yugoslavia. 1 – Page 8 of 57 – 40130017-TDC 02-24226 Although large industries and large office buildings obtain their electricity from the public medium- and high-voltage networks, most electricity used by those industries is consumed at low-voltage level – the conversion to low-voltage level is performed by privately owned distribution transformers. Often the transformers placed at industry or commercial offices are similar transformers as the distribution transformers in the public electricity supply system. Distribution transformer losses in industry are estimated to account for 1-2% of total final electricity consumption in Europe. As these losses occur in privately owned networks, these losses are not shown in the statistics as network losses! Private generation represents less than 10% of total capacity in the EU. However on-site generation of electricity for non-utility systems is growing rapidly, frequently using gas as fuel. Overall, it is estimated that private generation could reach 20% of total capacity in the near future. Growth is being assisted by a number of special factors, including the development of renewable and combined heat and power technology, improved economics for gas-based generation, the liberation of tariff controls, and deregulation of electricity supply. 2.2 Transformer population in industry According to [1], electricity utilities are estimated to own and operate about 70% of the total population of distribution transformers in the EU. Transformer ownership outside the utility sector is shared between the non-utility electricity supply systems and the medium sized customers for electricity. The population of distribution transformers installed in European electricity utility and private sector networks is estimated to be about four million units. Statistical records are poor, however, particularly for privately owned installations. Non-utility distribution transformers account for about 30% of the total population, but a much higher proportion, possibly almost 50%, of the total installed capacity. This is because nonutility transformers have a higher rating than the transformers at the utility (see figure 2.2). Some differences between these privately owned distribution transformers and their public counterparts are2: larger transformers: as the load relatively concentrated on a small area, the transformers are often larger in size than distribution transformers in residential areas, even urban areas: the latter normally range from 50 to 1000 kVA, industrial transformers often range from 1000 to 4000 kVA, see figure 2.2. These differences are deducted from KEMA’s experience and information from the interviews. 2 40130017-TDC 02-24226 70% 60% 50% 40% 30% 20% 10% 0% 00 40 50 31 00 25 00 16 00 10 0 63 0 40 0 25 16 10 0 Utility Industry 0 Relative population – Page 9 of 57 – Size Note: transformers with intermediate sizes are attributed to the next higher size Figure 2.2 Relative transformer population in the Netherlands (estimate KEMA) the low-voltage system: the low-voltage side of transformers up to 1600 kVA is usually at 420 V (sometimes 690 V). For transformers above 1600 kVA the low-voltage side is at 420 V or 690 V. Some industries use 3 or 6 kV, but they find it hard or expensive to get new components at this voltage, and new installations will be built for a voltage of 10 kV or 690 V the average load. As the load can be predicted better in an industrial process than in a public supply system, the transformer rating often is well tailored to the peak load. The interviewed industries have an average transformer load between 30 and 65% of the installed ratings, peak loads up to 100% for several hours sometimes occurring. Compared to the average loading of a transformer in the public network, typically 10-30%, the loading of ‘industrial’ transformers is relatively high the large loads, which are often equipped with their own transformer, e.g. drives the load pattern. The electricity consumed by industry keeps pace with economic activities. Almost each industry is working 24 hours a day and 7 days a week. This means the load pattern is often constant. the insulating medium: contrary to transformers in public networks, which usually apply oil-filled transformers, industrial transformers are often of the dry type for fire safety or due to bad experiences with oil-filled transformers3. In some cases, the higher civil engineering costs for civil works regarding fire safety of oil-filled transformers are said to offset the higher price of the dry-type transformer harmonics: industrial transformers are close to the end-use, and could suffer from significant harmonic loads in IT-intensive facilities. See section 3.3 industrial transformers with a range up to 1600 kVA are the same as those used by the utilities. Since 1970, there is an increasing use of variable speed drives (AC or DC) in the industry. Therefore the industries have some special transformers (3 windings) or transformers with different impedance or phase sequences. These special transformers are outside the scope of this report. Transformers in industry seem to be relatively new compared to those at utilities. In the indusThe oil in some older transformers contained very toxic components (e.g. PCBs) and had to be replaced) 3 – Page 10 of 57 – 40130017-TDC 02-24226 try and some large older office buildings only few transformers are older than 30 years. Since 1980 some industrial companies started to use dry-type transformers instead of oil transformers. While utilities generally rely on their own engineering staff to set standards for performance, including energy efficiency, private sector electricity supply systems are often designed with outside assistance. The pattern in Europe varies widely. In some countries, this work is undertaken mainly by firms of management contractors, or the design staff of a major electrical contractor. Elsewhere, independent professional consulting engineers are responsible for design and project management. In many cases, little or no attention is paid to energy losses in transformers. This may be due to the fact that the electrical infrastructure is not a core business of the owner and the electrical infrastructure is often realised on a turnkey, lumped-sum basis. This may make the specification and costs of transformers invisible to the future owner. Because the industry itself is not the customer ordering the transformer, the properties of the transformer are often unknown. Industries are reported to show less interest in longer-term problems, and demand more rapid paybacks, than the public sector network. A great difference between the industry and utilities is the payback period. The required payback period in industry may be between 1 and 5 years. 2.3 Large office buildings The load pattern for an office building is low (< 20%) in the weekend, and is higher for the working days (between 20% for the night hours up to 80% at the office times). The transformers used at an office building are between 250 kVA and 1000 kVA (almost the same range as the utilities). Until the 60s, in a typical office few electrical devices were present, typically an electric typewriter, a calculator and of course incandescent lights. These devices are known to be linear loads: the current waveform is the same as the voltage waveform. Later, the fluorescent lamps were introduced. Since the mid-80s, there has been a proliferation of electronic equipment in the offices. Today, in a typical office there is a minimum of personal computers, telephone systems, copier machines, fax machines and laser printers. All of these new loads are non-linear. 2.4 Data hotels Power supply designs for data processing rooms tend to be much more oversized compared to designs for general office areas or typical building wire systems where harmonic currents from electronic loads (like personal computers and terminals) have caused problems or public networks. Data hotels have a very large specific load (W/m2); they also have a high share of non-linear loads. Dimensioning is done with large margins, to be reliable and to anticipate (high) growth expectations. The consequence is that the actual loading is low compared to the nameplate rating. A survey [2] of computer power systems showed that the majority is op- – Page 11 of 57 – 40130017-TDC 02-24226 erated at less than half of their rated capacities. The transformers used for the data hotels are between 250 kVA and 1600 kVA. 2.5 Failure statistics of transformers in the field Only limited information is available about the transformer failure pattern in Europe. Several studies have been undertaken, but the results are rather inconclusive. A 1983 survey [1] based on 47,000 transformer-years of service in 13 European countries estimated the meanlifetime-between-failures (MLBF) of installed transformers to be 50 years, and showed design defects, manufacturing problems and material defects to be the main causes of failure. The same project identified windings and terminals to be the components most likely to cause failure in service. Failures in coils using jointed conductors, built in earlier years, have caused some problems. Next to MLBF values, some information about the mean-time-between-failures (MTBF) was found. The table below shows some data that were collected during the present study. Table 2.2 Reliability of distribution transformers in Europe Situation, source Number Total units in Annual transformer of failed operation faults units Distribution transformers, 0,2% per year, i.e. source [1] one in 500 transformers Average figure in one Dutch 0,25% industrial company interviewed Distribution transformers in 87 117297 0,0742%, i.e. Dutch networks, year 2000 1 in 1350 trans[12] formers Remark Oil transformers Dry-type transformers, high load, harmonic distortion Oil transformers From the table it can be seen, that distribution transformers are very reliable components. The industrial company cited suggests a somewhat lower reliability (although, strictly spoken, the population is too small to draw statistically significant conclusions). This may be explained by differences of transformer types (dry vs. oil), the higher load, harmonic distortion or a combination of these factors. It should be noted that these figures are not suitable to draw conclusions on which type of transformer (dry or oil) is better. 2.6 Costs of electricity for industry in Europe The rates for electricity depends on the industry sector and the country. In general, a company with a higher electricity consumption benefits a lower electricity rate. Table 2.3 shows the average electricity price for captive customers (i.e. smaller customers that may not yet choose their electricity supplier) in the Netherlands. – Page 12 of 57 – 40130017-TDC 02-24226 Table 2.3 Average electricity prices in the Netherlands (exc. regulatory energy tax and vat) Customer category Small consumers Of which households Large consumers (offices, datahotels) Of which industry Source: EnergieNed Electricity price (Euro/kWh) in 2000 0,095 0,104 0,082 0,064 The values given above are just indicative of the average price. There are differences between the electricity prices between the industry sectors. Figure 2.3 gives a comparison of the average electricity prices for some European countries. Comparison of european electricity prices for industry Euro/MWh 100 80 60 40 20 Ita ly G re ec Fi e nl a Sw nd ed e Ire n la nd Un it e Sp d Ki ain n Lu gd xe om m bo u Po rg rtu g Th Fr al e Ne an th ce er la nd Au s s G tria er m a Be ny lg i De um nm ar k 0 Figure 2.3 Electricity prices in Europe (source EnergieNed [6]) 2.7 Environmental issues Carbon dioxide (CO2) is believed to significantly reinforce to the greenhouse effect, which may lead to a change in the global climate. The annual emission of CO 2 in the Netherlands is about 181 million tonnes, almost half of which is due to energy consumption. This includes combustion of gas by energy consumers and combustion of fuels in electric power stations. Figure 2.4 gives the emission of CO2 per sector. – Page 13 of 57 – 40130017-TDC 02-24226 Emission of carbon dioxide in the Netherlands (181.009 thousand tonnes per year) 12% 15% road traffic 3% 14% 6% other transport oil refineries industry energy sector 26% 24% households other Figure 2.4 CO2 emissions in the Netherlands Source EnergieNed In the Netherlands the government and industry have adopted an Environmental Action Plan in 1991 with the objective of reducing by the year 2000 CO 2 emissions. On one hand the Environmental Action Plan stimulates energy conservation in homes, commerce and industry; while on the other hand the measures focus on more efficient energy production techniques (cogeneration) and the use of sustainable sources of energy. In Europe, each kWh end-use of electricity is associated with 0,4 kg CO2 emission, on the average. The emission rates vary from one country to another, in fact even from one power station to another, depending on the plant efficiency, and the fuel mix. The country average emission rates range from 0,3 to 0,6 kg per kWh in Europe. – Page 14 of 57 – 3. 40130017-TDC 02-24226 LOSSES IN TRANSFORMERS A power transformer normally consists of a pair of windings, primary (HV) and secondary (LV), linked by a magnetic circuit or core. When an alternating voltage is applied to one of these windings, generally the HV winding, a small current will flow which sets up an alternating magnetic flux in the core. This alternating flux, linked to both windings, induces a voltage in each of them. The current which is flowing is the situation that both windings are not loaded, is the magnetising current. This chapter describes the losses in transformers as provided by the J&P transformer book [11]. 3.1 No Load losses An unloaded transformer experiences losses. The magnetising current is required to take the core through the alternating cycles of flux at a rate determined by system frequency (50 Hz). In doing so energy is dissipated. This loss is known as the core loss, no load loss or iron loss. The core loss is present whenever the transformer is energised. Thus they represent a constant and therefore significant energy drain on any electrical system. In addition, the alternating fluxes generate also alternating forces in the iron core and hence noise. The core loss is made up of two components: the first one, the hysteresis loss, is proportional to the frequency and dependent on the area of the hysteresis loop in the B-H diagram, and therefore characteristic of the material and a function of the peak flux density. The second component is the eddy current loss that is dependent on the square of the frequency, the square of the thickness of the material and the resistivity. Minimising hysteresis losses therefore implies application of a material having a minimum area of hysteresis loops, while minimising eddy current loss is achieved by building up the core from a laminate of thin strips and high resistivity. – Page 15 of 57 – 3.2 40130017-TDC 02-24226 Load losses The load loss of a transformer is that part of the losses generated by the load current and which varies with the square of the load current. This falls into three categories: Resistive loss within the winding conductors and leads Eddy current loss in the winding conductors Eddy current loss in the tanks and structural steelwork. The latter two categories are also referred to as “extra losses”. Resistive loss follows Ohm’s law and can be decreased by reducing the number of winding turns, by increasing the cross-sectional area of the turn conductor, or by a combination of both. However, reducing the number of turns requires an increase of the flux i.e. an increase in the core cross-section, which increases the iron weight and iron loss. So a trade-off has to between made between the load loss and the no-load loss. Eddy currents arise from the fact that not all the flux produced by one winding links to the other winding. This flux leakage also leads to the short-circuit reactance or impedance of a transformer. In the past, this reactance was simply considered an imperfection arising from the unavoidable existence of leakage flux. Nowadays, the transformer impedance is a valuable tool for the system designer to determine system fault levels to meet economic limitations of the connected plant. The path of eddy currents in winding conductors is complex. The magnitude of this leakage flux depends on the geometry and construction of the transformer. The effect of leakage flux within the transformer windings results in the presence of radial and axial flux changes at any given point in space and any moment in time. These induce voltages which cause currents to flow perpendicular to the fluxes which lead to losses. The magnitude of these currents can be reduced by increasing the resistance of the path through which they flow, and this can be effected by reducing the total cross-sectional area of the winding conductor, or by subdividing this conductor into a large number of strands insulated from each other (in the same way as laminating the core steel reduces eddy-current losses in the core). However, the former alternative increases the overall winding resistance and thereby the resistive losses. Conversely, if the overall conductor cross-section is increased with the object of reducing resistive losses, one of the results is an increase of the eddy current losses. This can only be offset by a reduction in strand cross-section and an increase in the total number of strands. It is costly to wind a large number of conductors in parallel and so a manufacturer will wish to limit the total number of strands in parallel. Also, the extra insulation resulting from the increased number of strands results into a poorer winding space factor. It will be evident that in a transformer having a low reactance, winding eddy currents are less of a problem than one with high reactance. On very high currents (>1000 A) fluxes generated at the main leads can give rise of eddy current losses in the tank adjacent to these. Due to the leakage flux there are also eddy-current losses in tanks and internal structural steelwork. – Page 16 of 57 – 3.3 Extra losses due to harmonics 3.3.1 Harmonic components of the transformer load 40130017-TDC 02-24226 In the last few years, there has been a growing concern about harmonic distortion and the effects of harmonics in the power systems. The distribution grid is designed to carry the fundamental 50 Hz frequency. Almost all industries have non-linear loads. Non-linear loads generate high levels of harmonic currents. Typical non-linear loads include: computers UPS systems variable speed drives inverters. The non-linear load at the industries is a major difference in circumstances between transformers placed at the industries and utilities. The harmonic currents cause higher losses in the transformer and therefore a higher temperature, which will affect the lifetime of the transformers. The extra losses depend on the harmonic spectrum of the load current. The triple harmonics (3rd, 9th, 15th, etc.) are the major cause of heat because the phase currents add in the neutral conductor. The magnitude of the harmonic current produced by the triples can approach twice the phase current. This causes the neutral conductor to overheat because neutral conductors were historically designed for the same current as the phase conductors. Transformers are configured with a delta-wye connection to reduce the effects of harmonics. The triple harmonics are trapped and circulate in the delta primary of the transformer. Thus the harmonic content reflected back to the source (the medium-voltage network) is reduced. The circulating harmonics in the delta connected winding of the transformer create heat because of their higher frequencies. Next to problems with the triple harmonics a transformer feeding a converter or inverter can also have problems with other harmonics. The typical harmonics (h) in a semiconductor bridge can be calculated by using the formula: h=pxk 1 Where p is the pulse number of the bridge (6, 12) and k is an integer equal to 1, 2 …, n. For a 6 pulse bridges the typical harmonics are: 5, 7, 11, 13, 17, 19, 23, 25 etc. For a 12 pulse bridges the typical harmonics are: 11, 13, 23, 25 etc. Due to the harmonic currents in the transformer load at industries and the higher loading profile it is reasonable to expect that these transformers will have a lower lifetime. Harmonics are responsible for a higher load loss in the transformer and therefore a higher hot spot temperature in the transformer, which reduces the lifetime of a transformer. According to IEC 61378-1: “it is necessary that the harmonics spectrum of current at rated load be specified by the purchaser prior to the time of placing the order. In the absence of specific information, a harmonic spectrum can be derived according to 3.6.2 of IEC 60146-1-2 [10]”. – Page 17 of 57 – 40130017-TDC 02-24226 IEC 60146-1-2 says: “at an early stage of the design the following formulae may be used when only the convertor size is known by its transformer rating”. The harmonic current can be calculated as: I hl I 1L 1, 2 5 h h Where: IhL is the harmonic current h is the harmonic number I1L is the fundamental current The fundamental current is given by: I 1l S U 3 Where: S is the power of the transformer U is the line voltage Notes: 1 The formula for IhL is approximate, especially for large values of the angle of delay of control in rectifier operation for converters. 2 Depending on the DC load circuit inductance, the direct current ripple may increase the 5th harmonic current by 0,3 p.u. or more 3 Other uncharacteristic harmonics may remain due to supply voltage unbalance which prevents the expected cancellation. For this study it’s presumed that the harmonics in the current can be calculated according the given formula. For the 5th harmonic it’s presumed that this current is increased by 0,1 p.u. 3.3.2 Extra losses due to harmonics The extra losses arise as follows. Higher frequency components in the load current (harmonics) cause extra losses because harmonics do not fully penetrate the conductor. They travel on the outer edge of the conductor. This is called skin effect. When skin effect occurs, the effective cross sectional area of the conductor decreases; increasing the resistance and the I2R losses, which in turn heats up the conductors and anything connected to them. Harmonic currents increase also the eddy current losses in transformers as described in [3]. The eddy current losses are of most concern when harmonics are present, because they increase approximately with the square of the frequency. The no-load losses are transformer core losses, which are affected by harmonics only in relation to voltage distortion, not current distortion. Consequently, the increase in no-load losses due to harmonics is usually negligible. Harmonic currents however, very significantly affect load losses. – Page 18 of 57 – 40130017-TDC 02-24226 As described earlier, load losses consist primarily of resistive (or I 2R or ‘copper’) losses and extra (eddy current) losses. Due to harmonics, not only the eddy current losses but also the resistive losses increase. By definition, however, the loss increase due to the presence of harmonics is usually also designated as “extra losses”, see figure 3.1. Extra loss Load loss Resisitive loss Extra loss due to harmonics Conventional load loss excl harmonics No load loss Unloaded Rated load Actual load Actual load (excl harmonics) (incl harmonics) Figure 3.1 Extra losses due to harmonics There are several approaches to account for the increased losses caused by harmonics in selecting a transformer. The first one, devised by transformer manufacturers in conjunction with Underwriters Laboratories in the United States, is to calculate a factor for the increase in eddy current loss; this is known as ‘K-Factor’. The second method is to estimate by how much a standard dry-transformer should be de-rated so that the total loss on harmonic load does not exceed the fundamental design loss; this is known as ‘factor K’. The ‘factor K’ method (used in Europe) is described in the Harmonisation document HD 538.3.S1. A third way to calculate the influence of harmonics is described in the IEC 61378-1 “Transformers for industrial applications”. 3.3.3 Characterisation of non-linear transformer load: K-factor (US practice) In US practice, where dry-type transformers are often used, the K-factor is the ratio of eddy current losses when driving non-linear and linear loads: h K I h2 h 2 1 The higher the K-factor the greater the harmonics present, hence the more harmonic current the transformer must be designed to survive. K=1 K=4 K = 13 K = 20 Resistance heating, motors, distribution transformers Welders, induction heaters, Fluorescent lighting Telecommunications equipment Mainframe computers, variable speed drives, desktop computers – Page 19 of 57 – 3.3.4 40130017-TDC 02-24226 HD538 Transformer derating due to harmonics: Factor K (Europe) In Europe, the transformer de-rating factor is calculated according the formula in HD 538.3.S1. The factor K is given by: 2 e I h n N q I n K 1 n 1 e I n 2 I 1 2 0,5 Where e= the eddy current loss as the fundamental frequency divided by the loss due to a DC current equal to the RMS value of the sinusoidal current, both at reference temperature. n= the harmonic order I= the RMS value of the sinusoidal current including all harmonics given by: n N 2 I I n n 1 0,5 n N I I 1 n n 1 I 1 2 0,5 In = I1 = q= the magnitude of the nth harmonic the magnitude of the fundamental current exponential constant that is dependent on the type of winding and frequency. Typical values are 1,7 for transformers with round rectangular cross section conductors in both windings and 1,5 for those with foil low voltage windings. 3.3.5 Method according to IEC 61378-1 IEC 61378-1 deals with the specification, design and testing of power transformers and reactors, which are intended for integration within semiconductor convertor plants; it is not designed for industrial or public distribution of AC power in general. The scope of this standard is limited to applications of power converters, of any power rating, for local distribution, at moderate rated convertor voltage, generally for industrial applications and typically with a highest voltage for equipment not exceeding 36 kV. The convertor transformers covered by this standard may be of the oil immersed or dry-type design. The oil-immersed transformers are required to comply with IEC 60076, and with IEC 60726 for dry-type transformers. As described, the load losses in transformers are subdivided into loss as measured with DC (I2R loss) and, in addition, eddy current loss in windings and connections, and stray losses in conductive structural parts of the transformer. For this study its presumed that for the dry transformer there are only extra losses due to eddy losses in the windings. For oil transformers it is presumed there are extra losses due to eddy losses in the windings, connections and stray losses in construction parts (e.g. tank wall). – Page 20 of 57 – 40130017-TDC 02-24226 In normal service, the convertor transformer load current is non-sinusoidal hence, when transformed into a Fourier series; it shows a considerable amount of harmonics. This nonsinusoidal current raises the eddy loss and stray flux loss, and significantly increases the total loss calculated or measured with purely sinusoidal current. A correction to the higher loss value at rated; non-sinusoidal convertor load is required for the thermal rating of the transformer, and for the correct calculation of the loss and efficiency of the transformer. The total RMS value of the current IL with harmonic content can be calculated according to: n I L2 I h2 1 Where IL is the RMS value of the non-sinusoidal current of the transformer and h is the harmonic order number and therefore I h is the RMS value of the harmonic current, having order number h. The eddy current losses PWE in the windings are equal to: I PW Eh PW E1 h 1 1 I1 n PW E n 2 h 2 With PWE is the winding eddy loss at current IL. I1 represents the RMS value of the fundamental current, at rated load. For oil transformers where the low-voltage windings have high rated currents, in the range of a few kA, the internal high current connections require a separate analysis of the additional eddy current loss. The calculation of these losses (P CE) is almost identical to the eddy current losses in the winding, except that the enhancement factor for the harmonic number now is not equal to 2 but 0,8. n PCE PCEh 1 I PCE1 h 1 I1 n 2 h 0,8 The stray losses in construction parts can be calculated the same way: n PSE PSEh 1 I PSE1 h 1 I1 n 2 h 0,8 The total load loss of a transformer without harmonics (fundamental) equals: PT1 = PDC1 + PEXTRA1 = PDC1 + PWE1 + PCE1+ PSE1 If the same transformer is used in a grid with harmonics the total load loss equals: I PT PDC1 L I1 2 n I PW E1 h 1 I1 2 n I 2 h PCE1 PSE1 h 1 I1 2 0,8 h The above formula can be used for oil- and dry type transformers. As mentioned before for the dry type transformer PCE1 and PSE1 are both assumed to be zero. – Page 21 of 57 – 40130017-TDC 02-24226 During testing of the transformer at the manufacturer, PT1 and PEXTRA1 are measured at 50 Hz. Establishing of the PWE1, PCE1 and PSE1 can be done by the manufacturer, since they exactly know the design and construction of the transformer. It is also possible by measuring transformer losses both at DC, at the standard sinusoidal 50 Hz current and at a frequency other than 50 Hz. Combining these three measurements, it is possible to establish the eddy current losses in the windings and in the structural parts. 3.4 Thermal ageing of transformers Since harmonics introduce extra losses and therefore extra heat dissipation in transformers, harmonics may have a great influence on the lifetime of the transformer. The effect is usually expressed in terms of loss of lifetime or lifetime consumption. Based on IEC 60354 (oilimmersed transformers) and IEC 60905 (dry type transformers) it is possible to estimate the loss of lifetime due to harmonics. The calculation models provided in these standards are based on the hottest part in the winding. This hot-spot temperature is used for evaluation of a relative value for the rate of thermal ageing. The parameters and formulas as given in the both standards are used in this study to establish the hot spot in the transformers and the lifetime consumption when the transformers are loaded with harmonic current. The actual lifetime of a transformer depends to a high degree on extraordinary events, such as overvoltages, short-circuits in the system and emergency overloading. In this study we only take into account the hot-spot temperature in the winding, caused by harmonics in the current. The consequence of a higher hot-spot temperature in the winding, caused by harmonics will be a risk of premature failure. This risk may be of an immediate short-term character or come from the cumulative deterioration of the transformer over many years. Oil –immersed transformers For oil-immersed transformers designed in accordance with IEC 60076, the relative rate of thermal ageing is taken to be equal to unity for a hot-spot temperature of 98 oC, which corresponds to operation at an ambient temperature of 20 oC and a hot-spot temperature rise of 78 K. The relative ageing rate as a function of the hot-spot temperature θh is defined as: V = (ageing rate at θh / ageing rate at 98 oC) = 2 ( h 98) / 6 So for oil-immersed transformers the relative ageing doubles every increment of approximately 6 K. If the load and ambient temperature are constant during a period, the relative loss of life (L) is equal to L = V x t, t being the period under consideration. A value equal above 1 means that the loss of life greater is than “normal” expectation. In this report the maximum allowable hot spot temperature for oil immersed transformers is taken as 140 oC. Dry-type transformers For dry-type transformers designed in accordance with IEC 60726 the daily use of life due to thermal effects is calculated at an ambient temperature of 20 oC. The parameter θc (Hot spot temperature in the winding) is used to calculate normal lifetime consumption. For dry-type transformers, the relative ageing doubles every increment of approximately 10 K. – Page 22 of 57 – 40130017-TDC 02-24226 In this report the maximum allowable hot spot temperature for dry type transformers is taken as 190 oC. 3.5 Loss evaluation The total owning cost of a transformer consists of several components, including the purchase price, the value of energy losses, maintenance and repair costs over the lifetime, and decommissioning cost. The purchase price and the energy losses are the two key factors for comparison of the different transformers. In the industry it is very common that transformers are part of a turn key project. The contractor is often interested in a transformer with a low purchase price. However the user/owner of the transformer aims at buying the cheapest transformer, i.e. with the lowest total owning cost, which complies with the requirements for a given application. Losses, installation, maintenance, repair and decommissioning costs are seldom taken into account by the contractor when choosing between transformers. When comparing two transformers with different purchase prices and/or different losses, one must take into account that the purchase price is paid at the moment of purchase, while the cost of losses come into effect during the lifetime of the transformer. Usually the costs are converted to the moment of purchase by assigning capital values. When transformers are compared with respect to energy losses, the process is called loss evaluation. In the basic loss evaluation process, three transformer figures are needed: purchase price load loss no-load loss. For the specified load loss of a transformer, the purchaser can assign a cost figure per kW of loss representing the capitalised value (net present value) of the load losses over the lifetime of the transformer or a shorter time scale e.g. 5 or 10 years. This cost figure is based on the expected transformer load over time and the average cost per kWh. Similarly, for the no-load loss of a transformer, the purchaser can assign a cost figure per kW of no-load loss representing the capitalised value of the no-load losses. This cost figure is also based on the average cost per kWh and the interest rate chosen by the purchaser. As nearly all transformers are connected to the grid for 100% of the time, and the no-load losses are independent on the load, the load curve is not relevant. The average cost per kWh will tend to be lower than for the load losses, as the latter will tend to coincide with peak loads, at which time energy is very expensive. If high capitalisation values for losses are chosen, transformers with low losses but with higher investment cost tend to be favoured. If however capitalisation values are set to zero, a purchaser effectively eliminates energy loss evaluation from the purchase decision, which favours the cheapest transformer. Thus, the capitalised cost (CC) of a transformer can be expressed as the sum of the purchase price (Ct), the cost of no-load losses and the cost of the load losses, or as a formula: – Page 23 of 57 – 40130017-TDC 02-24226 CC = Ct + A x Po + B x Pk where A represents the assigned cost of no-load losses per watt, Po the value of the no-load losses per watt, B the assigned cost of load losses per watt and Pk the value of the load losses per watt. This formula can also be found in HD428 and HD538. Po and Pk are transformer properties. A and B are properties that depend on the expected loading of the transformer and energy prices. The choice of the factors A and B is difficult, as often the future loading of the transformer is not known. If the load grows over time, the growth rate must be known. Also, the applicable kWh price over the lifetime must be forecast. A tricky task! Finally, the interest rate and the economic lifetime may be difficult to choose. In practice, therefore, there is a lot of guesstimation in determining A and B factors, A ending up between less than 1 and 6 EUR/W and B ending up between 0,2 and 5 EUR/W. For this report, we provide a relatively simple method for determining the A and B factor for small transformers. The total cost over the lifetime of a transformer depends on a lot of figures. We only look at the costs of purchaser price and the price of the losses during the lifetime. This method is not entirely correct, but gives an indication of the factor A and B for the industry and is therefore better than disregarding the costs of loss during lifetime at all. A and B are calculated as follows: A (1 i) n 1 8760 (no-load loss capitalisation) i (1 i) n C kW h and I (1 i ) n 1 B C kW h 8760 l n i (1 i ) Ir 2 (load loss capitalisation) where: i n CkWh 8760 IL Ir = interest rate [%/year] = lifetime [years] = kWh price [EUR/kWh] = number of hours in a year [h/year] = loading current [A] = rated current [A] These formulae assume that energy prices and the loading are constant over the transformer life. Usually, the loss evaluation figures A and B are submitted to the transformer manufacturers in the request for quotation. They can in turn start the complicated process of transformer design, to obtain a transformer design which performs best using the same formula. The result of this open process should be the cheapest transformer, i.e. with the lowest total owning cost, optimised for a given application. Drawbacks of this process are, as mentioned, its extreme complexity and the uncertainty of the purchaser with the exact load profiles of the transformers and energy prices in the future. Tariff structures are very complex. For large transformers, above a few MVA, the cost of losses are so high, that transformers – Page 24 of 57 – 40130017-TDC 02-24226 are custom-built, tailored to the loss evaluation figures specified in the request for quotation for a specific project. For distribution transformers, often bought by large batches, the process is undertaken infrequently, e.g. once every 5 years. This yields an optimum transformer design, which is then kept for several years until energy prices or load profiles have changed dramatically. In fact the loss levels established in HD428, HD538 and national standards reflect established practice of preferred designs with respect to loss evaluation values. 3.6 CO2 reduction and emission trading Benefits of energy saving are not only the avoidance of energy cost, but also the avoidance of CO2 emission and a contribution to the reduction of global warming. Usually, the CO 2 emission associated to the use of electricity is assumed to be 0,4 kg CO 2 per kWh of electricity used. The CO2 emission depends on the average share of fossil fuels in the fuel mix of power plants, and values between 0,3 and 0,6 kg/kWh are used for different countries in Europe. Presently, CO2 emission is free, but emission limits are underway. To allow for economic optimisation, emission-trading schemes are under discussion, and by that time, CO2 emission rights will have a price. The price will depend on actual market conditions. The price for CO 2 emissions will add to the cost of energy. For this study, prices of EUR 10 and EUR 33 per tonne of CO2 will be considered4. The extra cost of energy are equal to the CO2 emission per kWh times the price of the emission. For the variants for CO2 emission and emission cost, following variants arise for the extra cost of energy due to the emission cost: Table 3.1 Extra cost of energy due to emission cost, expressed as EUR/MWh CO2 emission (depending on fuel mix of power plants) Emission cost EUR 10/tonne EUR 33/tonne 0,3 kg/kWh 3 EUR/MWh 9,9 EUR/MWh 0,4 kg/kWh 4 EUR/MWh 13,2 EUR/MWh 0,6 kg/kWh 6 EUR/MWh 19,8 EUR/MWh The extra cost of energy can vary between 3 and about 20 EUR/MWh. These prices may seem high, but could become reality in the near future. If the legislation becomes implemented, the price ceiling will be the penalty for producing too much emission. As described in [13] the European Parliament has proposed a penalty of about EUR 50,= per tonne CO2 in 2005, rising to even EUR 100,= in 2008. 4 – Page 25 of 57 – 4. 40130017-TDC 02-24226 TYPICAL INDUSTRIAL TRANSFORMER DATA Distribution transformers as used in the industry, are normally considered to be the same transformers which provide the transformation from medium-voltage to low-voltage in public distribution networks. In Europe this is 420 V phase to phase, but for the industry 690 V phase to phase is also a common value. Distribution transformers are usually made in a different factory from larger transformers. There are many more manufacturers that build small transformers than those at the larger end of the scale. The industry is very competitive, and as a result the main consideration in the design of the active part is to achieve the best use of materials and to minimise costs. 4.1 HD standard data for distribution transformers up to 2500 kVA Most of the characteristics of industrial transformers are specified in national or international product standards for distribution transformers. Generally, the purpose of standards is to facilitate the exchange of products in both home and overseas markets, and to improve the product quality, health, safety and the environment. International standards are also of importance in reduction trade barriers. The application of standards can be legally required, or by specific reference in the purchase contract. For distribution transformers purchased in the European Union, three levels of standards are applicable: World-wide standards (ISO, IEC) European standards and regulations (EN, HD) National standards (e.g. NBN, BSI, NF, DIN, NEN, UNE, OTEL). European Harmonisation Documents (HD) are initiated if there is a need for a European standard. The draft HD is a compilation of the different national standards on the subject. The HD is finalised by eliminating as many national differences as possible. When a harmonisation Document has been issued, conflicting national standards have to be withdrawn within a specified period of time, or modified to be compatible with the HD. Usually, the HD is the predecessor of an European standard (EN), which must be adopted as a national standard in the EU member countries. Thus, purchase orders which refer to national standards are compatible with European standards (EN) and/or harmonisation documents (HD). Among the many international standards for distribution transformers, two main European Harmonisation Documents specify energy efficiency levels: HD 428: Three phase oil-immersed distribution transformers 50 Hz, from 50 to 2500 kVA with highest voltage for equipment not exceeding 36 kV. HD 538: Three phase dry-type distribution transformers 50 Hz, from 100 to 2500 kVA, with highest voltage for equipment not exceeding 36 kV. For the industrial transformers there are also other world-wide standards. These standards are IEC 61378-1: Transformers for industrial applications, and if the transformer is loaded with a non-linear (converter) load IEC 60146-1-2: Semiconductor convertors, general requirements and line commutated converters. – Page 26 of 57 – 40130017-TDC 02-24226 Figure 4.4 gives the limits for load losses for some important types of oil-filled and dry-type distribution transformers according to HD428.1 and HD538.1 for the preferred rated power range of the transformers. For oil-filled distribution transformers, the HD allows a choice of energy efficiency levels, A, B and C. The no-load losses (iron losses) for the same range of transformers are given below. For oil-filled distribution transformers, the HD offers a choice between three efficiency levels, A’, B’ and C’. (Figure 4.1) Figure 4.1 Distribution Transformer Loss Standards RATED POWER kVA 50 100 160 250 400 630 /4%1) 630 /6% 1000 1600 2500 4000 Notes Load Losses OIL-FILLED (HD428) DRY TYPE UP TO 24kV2) (HD538) LIST A LIST B LIST C 12kV PRIMARY 3) W W W W 1100 1350 875 N/A 1750 2150 1475 2000 2350 3100 2000 2700 3250 4200 2750 3500 4600 6000 3850 4900 6500 8400 5400 7300 6750 10500 17000 26500 N/A 1. 2. 3. 8700 13000 20000 32000 N/A 5600 9500 14000 22000 N/A 7600 10000 14000 21000 N/A No-Load Losses OIL-FILLED (HD428) DRY TYPE UP TO 24kV2) (HD538) LIST A’ LIST B’ LIST C’ 12kV PRIMARY 3) W W W W 190 145 125 N/A 320 260 210 440 460 375 300 610 650 530 425 820 930 750 610 1150 1300 1030 860 1500 1200 1700 2600 3800 N/A 940 1400 2200 3200 N/A 800 1100 1700 2500 N/A 1370 2000 2800 4300 N/A The short-circuit impedance of the transformers is 4% or 6%, in most cases. This technical parameter is of importance to a utility for designing and dimensioning the low-voltage network fed by the transformer. Transformers with the same rated power but with different short-circuit impedance have a different construction and therefore slightly different losses. For HD428 / HD538 compliant distribution transformers, the preferred values for the short-circuit impedance are 4% for transformers up to and including 630kVA, and 6% for transformers of 630kVA and above. For 36kV transformers, different values apply. For 24 and 36kV transformers, different values apply. Source: European Harmonisation Documents HD428 therefore allows customers to choose between three levels of no-load losses and three levels of load losses. In principle, there are a total of 9 possible combinations, ranging from the lowest efficiency, (B-A’), to the highest, (C-C’), which may be regarded as providing a high practical standard of energy efficiency for a distribution transformer. HD428 defines five preferred combinations of these losses. These combinations are shown below in Table 4.2, where the combination A-A’ is chosen as the base case (shown as a bold line – the percentages refer to this combination). – Page 27 of 57 – 40130017-TDC 02-24226 Table 4.2 lower losses Load losses B (+20..30%) A (base) C (-15..20%) No-load losses A’ (base) B’ (-15..25%) C’ (-35%) There is a significant difference in total no-load and load losses between A-A’ and C-C’ distribution transformers, approximately 1.5 kW for a 630 kVA unit. HD428.1 (part 1: general requirements and requirements for transformers with highest voltage for equipment not exceeding 24 kV) as well as other HD sections also contain phrases such as “(…) in the case of established practice in the market (…) the transformers can be requested and, by consequence, offered, with losses differing from the tabled losses”, which indicates some freedom to national or local deviations. As stated before, HD428 and HD538 represent a compilation and/or compromise on the various old standards which were used in European countries. It appears to be rather unambitious in terms of the loss levels set, and by allowing capitalisation formulas to be used. The standard HD 538 specifies only one level of losses. Both standards seem to be tailored to public utility applications, but they could easily become much more applicable for industry by: extending the rated power range to 4000 kVA transformers adding 690 V as a secondary voltage for the higher power range (1600-4000 kVA) differentiating preferred loss levels for HD538 transformers. Distribution transformers built according to HD428 and HD538 have a limited number of preferred values for rated power (50, 100, 160, 250, 400, 630, 1000, 1600 and 2500kVA). Intermediate values are also allowed. The two key figures for energy efficiency, the load losses and the no-load losses, are specified for each rated power. The interviews do not provide enough data for concise statements on the population in a specific sector in the industry. In the companies interviewed, often the preferred ratings from the HD documents are found, but also many transformers with ratings of 850, 1250, 3000 or 3150 kVA. This suggests that the range of preferred values for rated power according the HD 428 and HD 538 could be extended. The extension should be up to 4000 kVA rated power, and should follow a geometric series. In order to reduce losses due to mismatch between transformer – Page 28 of 57 – 40130017-TDC 02-24226 and load, the rating for transformers >630 kVA should become more fine (R10 geometric series instead of R5 series), all the more since the load can often be predicted more exactly, which allows a better match between transformer and load. Table 4.3 gives the proposed (extended) range of preferred transformer ratings. Table 4.3 proposed (extended) range of preferred transformer ratings Present range in Proposed range Proposed secondHD428/538 HD 428/538 ary voltages 50 50 420 V 100 100 420 V 160 160 420 V 250 250 420 V 400 400 420 V (500) 420 V 630 630 420 V 800 420 V 1000 1000 420 V 1250 420 V 1600 1600 420 V 2000 420 V 2500 2500 420 V, 690 V 3150 420 V, 690 V 4000 420 V, 690 V 4.2 Data for oil-immersed transformers between 100 and 1600 kVA Table 4.4 shows the data for oil-immersed transformers as given in [1]. Since the industries use transformers with higher ratings this table should be extensive with higher transformer ratings as given in chapter 4.3. Table 4.4 data for transformers used in the utilities RATING kVA 100 400 1600 HV kV 20 10 20 LV V 400 400 690 LOSS-LEVEL HD428 A-A' C-C' ACA-A' A-A' C-C' C-C' ACA-A' A-A' C-C' C-C' ACAMDT AMDT AMDT AMDT AMDT AMDT 210 60 60 930 930 610 610 150 160 2.600 2.600 1.700 1.700 380 420 NO-LOAD LOSSES W 320 LOAD LOSSES W 1.750 1.475 1.750 1.475 4.600 4.600 3.850 3.850 4.600 3.850 14.000 14.000 17.000 17.000 17.000 14.000 TOTAL MASS kg 520 650 740 770 1.190 1.200 1.300 1.400 1.590 1.750 3.300 3.240 3.370 3.680 4.310 4.550 CORE MASS kg 150 220 220 225 435 440 450 540 570 600 1.100 1.210 1.200 1.460 1.400 1.550 FLUX DENSITY T 1,83 1,45 1,35 1,35 1,83 1,84 1,65 1,6 1,35 1,35 1,84 1,84 1,7 1,6 1,35 1,35 CONDUCTOR MA- Cu/Al TERIAL WINDING MASS kg CURRENT DENSITY A/mm2 Cu Cu Cu Cu Cu Al Cu Al Cu Cu Cu Al Cu Al Cu Cu 85 115 130 155 203 145 350 220 360 450 505 295 725 465 1.120 2,9 2,3 2,35 2 2,9 1,55 2,1 1,1 2,3 1,85 3,65 2 2,75 1,4 2,45 1.225 2,1 HEIGHT mm 1.300 1.300 1.300 1.300 1.330 1.420 1.350 1.550 1.400 1.400 1.890 1.820 1.860 2.000 1.870 1.900 LENGTH mm 890 830 1.050 1.100 1.320 1.100 1.010 1.130 1.340 1.240 1.820 2.000 1.710 1.850 1.770 1.770 WIDTH mm 600 560 620 620 800 840 800 780 770 800 1.180 1.280 1.100 1.020 1.320 1.200 EFFICIENCY (*) % 97,94 98,32 98,19 98,46 98,62 98,62 98,89 98,89 98,81 99,00 98,78 98,78 99,02 99,02 98,91 99,10 – Page 29 of 57 – SOUND POWER dB(A) 57 36 UNIT COST UNIT COST Euro 2538 2800 % 90,7 100 59 59 61 68 3456 3567 4385 123,5 127,5 93,2 40130017-TDC 02-24226 56 58 4286 4881 4705 91,1 103,7 100 68 68 68 72 63 63 6373 6797 9692 9251 10307 10119 15050 15531 135,5 144,5 95,8 91,4 148,7 101,9 100 76 76 153,5 Source: [1] 4.3 Data for transformers between 1000 and 4000 kVA As mentioned before, the industries have transformers with higher ratings than those used in the distribution grid of the utilities. Pauwels Trafo calculated different types of transformers with power between 1000 and 4000 kVA for industrial use. Table 4.5 shows the calculated values for 16 transformers. As can seen there are 4 different transformers chosen with power of 1000, 1600, 2500 and 4000 kVA. The information is based on oil-immersed transformers and dry-type transformers. For each type of transformer there is made a calculation with the losses according (or in range) with the HD 428 or HD 538 and losses with 15% reduction. This is reduction is done to investigate the transformers if they are loaded typical industrial loading (i.e. loading with harmonics). It should be noted that the values given in table 4.5, are rough indications and can only be used for indication! Table 4.5 calculated losses for industry transformers Typical Industry Transformer Parameters rating kVA 1000 1600 2500 HV kV 10 10 10 10 LV V 420 420 420 420 Uk % LOSS-LEVEL 6 Oil CC' Oil DD' Dry base 6 Dry Low Oil CC' Oil DD' Dry base 8 Dry Low Oil CC' Oil DD' Dry base 4000 8 Dry Low Oil CC' Oil DD' Dry base Dry Low NO-LOAD LOSSES W 1100 935 2000 1735 1700 1445 2800 2670 2500 2125 4300 4130 3800 3230 7000 5540 LOAD LOSSES 75 ºC W 9500 8075 8600 7270 14000 11900 10000 9350 22000 18700 18000 14930 34000 28900 27000 26630 TOTAL MASS kg 2715 3157 2530 2800 3900 4210 3840 3900 4925 6065 5350 5410 8885 10108 7660 7710 HEIGHT mm 1890 1800 1560 1620 2090 2090 1830 1820 1925 1915 2040 2130 2485 2415 2470 2410 LENGTE mm 1500 1540 1710 1690 1875 1795 1920 1840 2360 2370 2160 1980 2545 2545 2310 2360 WIDTH mm 950 1800 940 940 1155 2090 940 940 1235 2370 1230 1230 1375 2545 1230 1230 T HS (F) K 65 65 100 100 65 65 100 100 65 65 100 100 65 65 100 100 T LS (H) K 65 65 100 100 65 65 100 100 65 65 100 100 65 65 100 100 SOUND POWER dB(A) 56 51 68 61 68 57 70 67 69 59 74 73 72 60 80 77 EFFICIENCY (*) % 98,94 99,10 98,94 99,10 99,02 99,17 99,20 99,25 99,02 99,17 99,11 99,24 99,06 99,20 99,15 99,20 UNIT COST Euro 8007 10353 10074 11108 10865 12832 14451 14990 13670 17887 17951 19073 24987 29402 25527 27494 UNIT COST % 100 129 126 139 133 138 131 140 102 110 100 118 100 131 100 118 (*) at full load and cos phi = 1 Source Pauwels Based on the provided information one can conclude that in general dry-type transformers are having a higher purchase price than oil-immersed transformers. There is also a big difference in the losses and dimensioning. Since harmonics do have a great influence on the extra losses, Pauwels also calculated the following values for the extra losses (as a percentage of the total load losses at 50 Hz). These losses have been subdivided in: extra losses in the winding (expressed as percentage of the total load losses at 50 Hz). These losses are designated as PWE extra losses in structural parts, tank walls etc. and in the internal high current connections (expressed as percentage of the total load losses at 50 Hz). For dry type transformers, these losses are negligible. These losses are designated as PSE and PCE Table 4.6 Extra losses in the winding and structural parts 1000 kVA 1600 kVA In wind- Other In wind- Other 2500 kVA In wind- Other 4000 kVA In wind- Other – Page 30 of 57 – Oil CC’ HD 428 Oil DD’ HD 428 Dry type HD538 Dry type low losses Source Pauwels ing PWE 6% 6% 6% 6% PSE + PCE 5% 5% – – ing PWE 9% 9% 9% 9% PSE + PCE 13 % 13 % – – 40130017-TDC 02-24226 ing PWE 11% 11% 11% 11% PSE + PCE 14 % 14 % – – ing PWE 13% 13% 13% 13% As can seen from the data in table 4.6, there is a large difference between the extra losses of a oil- or dry type transformer in the range from 1 MVA up to 4 MVA. It should be noted that, in order to make the difference clear between a dry and oil transformer, Pauwels based the calculations on a foil LV winding for both transformer types and for all transformer ratings. Therefore the extra losses of the 2500 kVA and the 4000 kVA transformers are relatively high. To lower these extra losses, it is possible to use another (not a foil) LV winding configuration. However this will increase the purchase price. Next to this, the transformer supplier should take care of the axial forces during short circuit. PSE + PCE 28 % 28 % – – – Page 31 of 57 – 5. 40130017-TDC 02-24226 CASE STUDIES This chapter describes some typical examples (based on the practice) about using energyefficient transformers. For these case studies, typical data collected during interviews were used, completed with KEMA estimates. All cases refer to Dutch industrial companies. 5.1 Case study 1: a large company in the iron-steel sector As shown in chapter 2 the iron and steel industry is one of the industry sectors with the highest electricity consumption in Europe. This case describes the possible energy saving for a steel plant in the Netherlands by using energy efficient transformers. 5.1.1 Iron-steel: a big consumer of electricity In this case study we are considering a company in the iron-steel industry with average electricity loading of 400 MW. About 60 MW of the loading is used at higher voltages (mainly highvoltage motors) and are therefore not distributed by distribution transformers. The electricity consumption is relatively constant during 24 hours a day, 7 days a week. The transformer rating is between 800 kVA and 4800 kVA. There are about 400 transformers. 200 transformers (50%) are 1250 kVA; 25% of the transformers 1600 kVA and 25% other ratings. Almost all transformers are dry-type transformers because of problems in the past with PCB in oil. Most of the transformers have been (re)placed between 1982 and 1990. The losses of the existing transformer are almost identical as the losses given in the HD538. The company usually buys transformers separately (i.e. not in large turnkey contracts). The loss evaluation used for small transformers is given as EUR 2,27/W (no load losses) and EUR 1,63/W (load losses). These values are regarded as outdated now. In the following cases we evaluate the energy saving between the HD538 dry-transformers and the low loss dry transformers for the ratings 1250 and 1600 kVA compared with the actual present transformers. 1250 kVA transformer The table below shows the input data for the case study. Also shown are the economic loss evaluation factors resulting from the input data. Table 5.1 input data 1250 kV transformer Transformer size 1250 kVA dry-type Transformer load 65% (constant load, 24/24h) with 6 pulse harmonics Economic lifetime 10 years Interest rate 7% Energy price EUR 40/MWh Harmonic spectrum 1 3 5 7 9 11 13 15 17 19 % 100 0 29 11 0 6 5 0 3 3 A (no-load loss evalua- EUR 2,46 /W tion) B (load loss evaluation) EUR 1,04 /W 21 0 23 2 25 2 – Page 32 of 57 – 40130017-TDC 02-24226 As can be seen, the actual no-load loss evaluation (A) and load loss evaluation (B) are different from the calculated loss evaluation factors. The load loss evaluation used by the company seems rather high, the actual no-load loss evaluation rather low. Because of the loss evaluation factors in the past, the actually placed transformers have relatively low losses and it is not cost-effective to replace the existing transformers with new transformers with lower losses. However if an existing transformers failed or a new transformer would be needed, it would be very attractive to use transformers with low losses. The key output data are given in table 5.2 for the situation with harmonic loading. Table 5.2 outcome 1250 kVA transformer Unit Transformer rating Rated no-load loss Rated load loss Total annual losses CO2 emission @ 0,4 kg/kWh Purchase price Present value no-load loss Present value load loss Capitalised costs Pay back (years) Internal rate of return kVA W W kWh/a ton/a EUR EUR EUR EUR Dry transformer 1250 2400 13568 71241 28,5 12250 5907 14108 32265 Dry transformer, low losses 1250 2200 11712 62618 25,0 13000 5414 12178 30592 Difference -200 -1856 -8623 -3,5 750 -493 -1930 -1673 2,2 45% Although the dry transformer with low losses has a price that is about 6% higher than the dry transformer according the HD 538, it is clear that the low loss transformer in fact is the most economical transformer. The payback time (2,2 years) is far smaller than the economic lifetime of 10 years. Even without evaluating the CO2 emission values, the dry transformer with the lowest loss is the cheapest dry transformer during the lifetime. 1600 kVA transformer Table 5.3 shows the input data for the case study. Also shown are the economic loss evaluation factors resulting from the input data. Table 5.3 input data 1600 kV transformer Transformer size 1600 kVA dry-type Transformer load 65% (constant load, 24/24h) with 6 pulse harmonics Economic lifetime 10 year Interest rate 7% Energy price EUR 40/MWh Harmonic spectrum 1 3 5 7 9 11 13 15 17 19 % 100 0 29 11 0 6 5 0 3 3 A (no-load loss evaluation) EUR 2,46 /W B (load loss evaluation) EUR 1,04 /W 21 0 23 2 As seen at the 1250 kVA transformer, the actual no-load loss evaluation (A) and load loss evaluation (B) are different than the calculated loss evaluation factors. Because of the loss evaluation factors in the past, the actually installed transformers have relatively low losses 25 2 – Page 33 of 57 – 40130017-TDC 02-24226 and it is not cost-effective to replace the existing transformers with new transformers with lower losses. However if an existing transformer failed or the industry were growing, it would be very attractive to use transformers with low losses. The key output data are given in table 5.4 for the situation with harmonic loading. Table 5.4 outcome 1600 kVA transformer Unit Dry transformer Transformer rating kVA 1600 Rated no-load loss W 2800 Rated load loss W 15207 Total annual losses kWh/a 80809 CO2 emission @ 0,4 ton/a 32,3 kg/kWh Purchase price EUR 14451 Present value no-load EUR 6891 loss Present value load loss EUR 15812 Capitalised costs EUR 37154 Pay Back (years) Internal rate of return Dry transformer, low losses 1600 2670 14218 76012 30,4 Difference -130 -989 -4797 -1,9 14990 6571 539 -320 14784 36345 -1028 -809 2,8 34% Although the dry transformer with low losses has a price that is 4% higher than the dry transformer according the HD 538, it is clear that the low loss transformer in fact is the most economical transformer. Even without evaluating the CO 2 emission values, the dry transformer with the lowest loss is the cheapest dry transformer during the lifetime (pay back period 2,8 years). It is clear that this transformer is more attractive if the emissions levels are taken in account. 5.1.2 Conclusion Since this company is familiar with (and uses) loss evaluation for small transformers, the transformers are already rather efficient and replacing all existing dry type transformers with a transformer with low losses is not economical. However if a transformer fails, or if there are new transformers needed, it is economical to buy dry-type transformers with low losses compared with the existing transformers and HD 538 transformers. The payback period for energy efficient transformers is even smaller than 3 years. Based on a European average value of 0,4 kg/kWh CO 2 emission, the annual energy saving for the 1250 kVA transformer is about 3,5 tonne. The annual energy saving for the 1600 kVA transformer is about 1,9 tonne. For all other small transformers it is presumed that the energy and CO2 emission saving can be taken as the average saving of the 1250 and 1600 kVA transformer. Table 5.5 gives the potential annual energy and CO2 emission saving for this company, if all transformers were replaced. – Page 34 of 57 – Table 5.5 annual savings potential Transformer Total Energy saving size number [MWh/a] 1250 kVA 200 1724,6 1600 kVA 100 479,7 Other 100 734,8 Total 400 40130017-TDC 02-24226 CO2 emission saving [tonnes/a] 700 190 300 2939,1 1190 The average annual energy consumption of this company is about 3,5 TWh. This means a reduction of 0,084% which equals an annual cost saving of EUR 117564,= (exclusive CO 2 emission saving). 5.2 Case study 2: A large company in the non-ferrous metals sector Almost 3% of the electricity consumption in Europe is used by the non-ferrous metals sector. However, only a small share of this electricity passes a distribution transformer, as much electricity is used for e.g. electrolysis. 5.2.1 Non-ferrous metal sector: replacing older transformers For this case study we are considering a company in the Netherlands, with an electricity loading of about 190 MW. Almost 180 MW of the loading is used at higher voltages (electrolysis) and therefore not distributed by distribution transformers. This load is present during 24 hours a day and 7 days a week. The remaining electricity consumption varies between 10 MW between 07.00 and 17.00 h, and 8 MW during the night hours. There are 25 transformers installed, with ratings between 500 kVA and 1250 kVA. Half of the transformers have a size of 1000 kVA. Three new dry type transformers were installed in 1999, and 1 oil transformer in 1987. All other 1000 kVA transformers are old transformers and installed between 1965 and 1970. The losses for these older transformers are: No load loss : 1900 W Load loss : 10250 W. During working hours the average loading of these transformers is about 45%, during the night hours the average loading is about 35%. In the following cases we evaluate the energy saving between the existing 1000 kVA transformers and the 1000 kVA low loss transformers. Since the latest transformers are dry type transformers and the older transformers are oilimmersed transformers, the differences between these two transformers are calculated. 1000 kVA transformer The table below shows the input data for the case study. Also shown are the economic loss evaluation factors resulting from the input data. Table 5.6 input data 1000 kVA transformer Transformer size 1000 kVA oil-type Transformer load 45% during 10/24 hours; 35% during 14/24 hours – Page 35 of 57 – Economic lifetime Interest rate Energy price Harmonic spectrum A (no-load loss evaluation) B (load loss evaluation) 40130017-TDC 02-24226 5 years 7% EUR 40/MWh No harmonic load is considered EUR 1,44 /W EUR 0,24 /W Since the evaluation of the load loss (B) is relative low, it is not economical to replace the older transformers with a transformer with lower losses. Even when the CO2 emission values are taken into account, it is not economical to replace. When this plant buys new transformers, the key output data are given in the table below for the situation with harmonic loading and the lowest CO2 emission values. Table 5.7 outcome 1000 kVA transformer Unit Dry HD 538 transformer Transformer rating kVA 1000 Rated no-load loss W 2000 Rated load loss W 8600 Total annual losses kWh/a 30336 CO2 emission @ 0,4 ton/a 12,1 kg/kWh Purchase price EUR 10074 Present value no-load EUR 2873 loss Present value load loss EUR 2102 Capitalised costs EUR 15049 Pay back (years) Internal rate of return Oil C-C’ transformer 1000 1100 9500 23793 9,5 Difference -900 900 -6543 -2,6 8007 1580 -2067 -1293 2322 11909 220 -3140 N/A N/A The internal rate of return can not be calculated since the difference between the purchase price and the differences between the annual losses are both negative. This makes clear that the oil transformer should be preferred, since the purchase price is lower and annual losses are also lower. The difference between these transformers will even be more, if the CO 2 emission values and/or losses due to harmonics are taken into account 5.2.2 Conclusion Although there are some older transformers (1965) in this plant, it is not recommended to replace the existing transformers. As long as these transformers do their job properly, no action should be taken. The latest transformers placed at this plant are all dry-type transformers. The given calculation shows however that if new transformers should be placed the oil-immersed transformers should preferred (looking at the transformer types used in the case study). Based at a European average value of 0,4 kg/kWh CO 2 emission, the annual energy saving – Page 36 of 57 – 40130017-TDC 02-24226 between the 1000 kVA oil and dry type transformer is about 2,6 tonne. For all other small transformers it is presumed that the energy and CO 2 emission saving can be taken as the average saving of the 1000 kVA transformer. Table 5.7 gives the annual energy and CO 2 emission saving potential for this company by placing oil immersed transformers instead of dry type transformers Table 5.8 annual savings potential Transformer Total Energy saving [MWh] size number 1000 kVA 12 78,5 Other 13 85,1 Total 25 CO2 emission saving [tonnes] 164 31,2 33,8 65 The average annual electricity consumption of this company is about 1,66 million kWh. This means a reduction of 0,0099% which equals an annual cost saving of EUR 6560,= (exclusive CO2 emission saving). The reason for the relative small reduction is because the most of the electricity consumption does not pass a distribution transformers but is used by the electrolysis equipment. 5.3 Case study 3: a paper and pulp company This case study shows an example (based on practice) of a paper and pulp plant which in the past (1978) bought transformers with normal losses (at that time) and for another process (1986) bought transformers with low losses (even nowadays these losses are very low). 5.3.1 Paper industry: energy savings based on practice This paper and pulp plant has two paper mills. The first mill was build in 1978, the second mill in 1986. The peak electricity loading is about 110 MW. The loading at higher voltage is about 72 MW (High voltage motors). About 12 MW of this loading is consumed during 24 hours a day, 7 days a week (Paper mill), and presumed is that the other 60 MW loading is consumed between 21.00 u and 06.00 u for the making of pulp. There are 52 transformers, with ratings between 1000 kVA and 3150 kVA. The latest 6 transformers (placed in 1997) are dry-type transformers, all other transformers are oil transformers. Most of the transformers (28) are 3150 kVA transformers with a LV of 690 V. The average loading of these transformers is about 65%, some of them with 6 pulse drives. Since the lifetime of a paper mill is several decades, this company applies since 1986 an evaluation of the losses of transformers and the purchase price over a lifetime of 20 years. However, this method is not based on the given loss evaluation method. The transformer losses for the transformers placed in 1986 are even lower than the D-D’ transformers as given in this report. The losses of the transformers placed in 1978 are almost identical to the losses of the C-C’ transformers. To evaluate the differences between the two transformers, it is presumed that the purchase price of the 1978 transformers nowadays equals the C-C’ transformer and the 1986 transformer equals a D-D’ 4000 kVA transformer, – Page 37 of 57 – 40130017-TDC 02-24226 however actual losses of this transformer are even lower than the given D-D’ transformer. 3150 kVA transformer This calculation will show the differences in losses between the 1978 and 1986 transformers (3150 kVA). Table 5.9 gives the input data as used for this evaluation: Table 5.9 input data 3150 kVA transformer Transformer size 3150 kVA oil-type Transformer load 65% during 24/24 hours with 6 pulse harmonics Economic lifetime 20 years Interest rate 7% Energy price EUR 40/MWh Harmonic spectrum 6 pulse according to IEC 146-1-1 A (no-load loss evaluation) EUR 3,71 /W B (load loss evaluation) EUR 1,57 /W Since the economic lifetime is long, the no-load and load loss evaluation are relative high. Obviously, it is not cost-effective to replace the existing 3150 kVA transformers (build in 1978) with D-D’ transformers. The key output data between the 1978 and 1986 transformers are given in the table below, including harmonics and with the lowest CO2 emission values. Table 5.10 Outcome 3150 kVA transformer Unit Oil 1978 transformer Transformer rating kVA 3150 Rated no-load loss W 2870 Rated load loss W 24500 Total annual losses kWh/a 181908 CO2 emission @ 0,4 ton/a 72,8 kg/kWh Purchase price EUR 19329 Present value no-load loss EUR 10654 Present value load loss EUR 66432 Capitalised costs EUR 96415 Pay back (years) Internal rate of return Oil 1986 Transformer 3150 3150 16800 135092 54,0 24987 11693 45553 82233 Difference -280 -7700 -46816 -18,8 5658 1039 -20879 -14182 3,0 33% Although the purchase price of the 1978 transformer is about 30% more expensive, over a life time of 20 years, the 1986 transformer is most cost effective (even without evaluating the CO 2 emission values). Thanks to the lower annual losses of the 1986 transformer, this company saves about 46816 kWh/year per 3150 kVA transformer. 5.3.2 Conclusion Although there are some older transformers (1978) in this plant, it is not recommended to re- – Page 38 of 57 – 40130017-TDC 02-24226 place the existing transformers. As long as these transformers do their job properly, no action should be taken. There are 18 transformers (3150 kVA) from about 1986 with very low losses. By calculation of the difference between the 1978 and 1986 transformers it is possible to give an estimation of the energy and emission savings by using energy efficient transformers. Based at an European average value of 0,4 kg/kWh CO2 emission the annual energy saving between the 3150 kVA oil type transformer is about 18,8 tonne. Table 5.11 gives the annual energy and CO2 emission saving for this company by placing oil immersed 1986 transformers instead of 1978 transformers (C-C’). Table 5.11 annual savings potential Transformer Total Energy saving size number [MWh/a] 3150 kVA 18 842,7 CO2 emission [tonnes/a] 338,4 saving Due to the use of energy efficient transformers (3150 kVA) this company already saves about 843 MWh per year. If we are considering the other transformers placed at this plant (16 x 2500 kVA; 5 x 1600 kVA; 7 x 1000 kVA and some other small transformers) the total potential of energy saving could be about 1250 MWh, which equals a CO2 emission of 500 tonnes. The average annual electricity consumption of this company is about 600 Million kWh. This means a reduction of 0,14% which equals an annual cost saving of EUR 33.700,= (exclusive CO2 emission saving). Due to the fact that most electricity consumption is caused by the HV motors, this reduction looks relative small compared with the total electricity use. 5.4 Case study 4: a chemical industrial enterprise Differences between plants in this sector are enormous. However, despite differences in the processes, the electrical installations are often similar. Often this industry uses electrolysis for the process and/or HV motors. Depending on the process, high reliability requirements cause a lot of redundancy in the electrical installation. Because of this redundancy the average loading of transformers is often not higher than 40%. Since there are some differences between the industrial plants, we have taken a fictitious case based on several chemical plants. The average loading is assumed to be 110 MW. About 40 MW is used by HV motors or electrolysis. The loading is 24 hours a day, 7 days a week. The loading is not linear (AC 12 pulse drives). A typical transformer for this kind of industry is a 1250 kVA transformer (60 out of 71 transformers). The other transformers have a ratings of 630, 1000 and 1600 kVA. 5.4.1 Chemical industry: energy savings by placing energy efficient transformers In this case we compare an 1250 kVA dry type transformer with losses in the range of the HD 538 and a 1250 kVA dry type transformer with low losses. 1250 kVA – Page 39 of 57 – 40130017-TDC 02-24226 This calculation is based on a loading of 40%. Table 5.12 gives the input data as used for this evaluation: Table 5.12 input data transformers Transformer size 1250 kVA dry-type Transformer load 40% 24/24 hours Economic lifetime 5 year Interest rate 7% Energy price EUR 50/MWh Harmonic spectrum No harmonics A (no-load loss evaluation) EUR 1,80 / W B (load loss evaluation) EUR 0,29 / W The key output data between the two transformers are given in table 5.13 (without harmonics and without the CO2 emission values). Table 5.13 Outcome 1250 kVA transformer Unit HD 538 transformer Transformer rating kVA 1250 Rated no-load loss W 2400 Rated load loss W 9500 Total annual losses kWh/a 34339 CO2 emission @ 0,4 ton/a 13,8 kg/kWh Purchase price EUR 12250 Present value no-load loss EUR 4310 Present value load loss EUR 2730 Capitalised costs EUR 19712 Pay back (years) Internal rate of return Low loss Transformer 1250 2200 8200 30765 12,3 13000 3951 2356 19686 Difference -200 -1300 -3574 -1,5 750 -359 -374 -17 4,2 6% 5.4.2 Conclusion As can be seen from table 5.13, the differences between the two 1250 kVA transformers are small. Even a small variation of one of the input variables can influence the decision which transformer is more economical. This shows that the input parameters are very sensitive for the decision to buy low loss transformers or regular transformers. However estimating these parameters can save a lot of energy loss and money (see also chapter 5.10). Still the pay back period for the transformer with the low losses is lower than the economic life time (4,2 respectively 5,0 years). Based at a European average value of 0,4 kg/kWh CO 2 emission the annual energy saving between the 1250 kVA dry type transformer is about 1,5 tonne. Table 5.14 gives the annual energy and CO2 emission saving for this company due to placing dry type transformers with lower losses. Table 5.14 annual savings potential Transformer Total Energy saving [MWh] CO2 emission saving [tonnes] – Page 40 of 57 – size 1250 kVA number 60 214,4 40130017-TDC 02-24226 85,8 Due to the use of energy efficient transformers this company saves about 214 MWh per year. If we are considering the other 11 transformers placed at this site, the total potential of energy saving could be about 254 MWh, which equals a CO 2 emission reduction of 101 tonnes/year. The average annual electricity consumption of this company is about 964 Million kWh. This means a reduction of 0,026% which equals an annual cost saving of EUR 12700,= (exclusive CO2 emission saving). Due to the fact that most electricity consumption is caused by the HV motors or electrolysis, this reduction is relatively small compared with the total electricity use. 5.5 Case study 5: a large data hotel start-up 5.5.1 Data hotels: a quickly growing sector Within the time scope of the project, it was not possible to arrange an interview with a representative of this sector. Based on KEMA’s experience, however, sufficient data was available to set up a typical case study. Within a short time, data hotels have become notorious for applying for huge connections at network companies, sometimes in the order of 100 MW or more, i.e. comparable to heavy industry. The rated power of the network connection is, as with any customer, based on growth expectations, the annual growth expectation for data hotels often being in the double or triple figures. The cited power ratings are extrapolations of the growth rates for a few years. The result of this approach often is an electrical power supply installation that is overdimensioned to a very large extent. The loading of transformers may be as low as 5-10%. Caused by the power supply of computers, datahotels have a lot of harmonics in their loading. The spectrum used for this case study is similar to the harmonic spectrum of the office building (case 6). However it contains more third harmonics. The economic lifetime will be very short, e.g. one year only. Table 5.15 shows the input data for the case study. Also shown are the economic loss evaluation factors resulting from the input data. Table 5.15 input data transformers Transformer size 1600 kVA dry-type Transformer load 25% (constant load, 24/24h), initial loading 70% (constant load, 24/24h), end of year 1 Economic lifetime 1 year Interest rate 7% (not very relevant in view of the economic lifetime) Energy price EUR 60/MWh Harmonic spectrum 1 3 5 7 9 11 13 15 17 19 21 % 100 59 41 23 0 6 5 0 3 2 0 A (no-load loss evaluation) EUR 0,52 /W (initial loading) EUR 0,52 /W (end of year 1) 23 1 25 1 – Page 41 of 57 – B (load loss evaluation) 40130017-TDC 02-24226 EUR 0,03 /W (initial loading) EUR 0,24 /W (end of year 1) Since the loading is low and the economic lifetime is short, one could ask if the relatively large transformers are an optimal choice and if it is cost effective to have a transformer with a lower power rating. Based on the following figures it is possible to estimate if a 1600 kVA transformer is economically the best chosen transformer. Also, it is estimated if energy-efficient transformers are also economically efficient. 5.5.2 Evaluation of more efficient transformers with the same rating Table 5.16 shows the results for a more efficient transformer. Table 5.16 Outcome 1600 kVA transformer (initial loading) Unit Dry trans- Dry transformer former Low losses Transformer rating kVA 1600 1600 Rated no-load loss W 2800 2670 Rated load loss W 10000 9350 Total annual losses kWh/a 38428 36385 CO2 emission @ 0,4 ton/a 15,3 14,5 kg/kWh Purchase price EUR 14451 14990 Present value no-load loss EUR 1444 1377 Present value load loss EUR 818 765 Capitalised costs EUR 16714 17132 Pay back (years) Internal rate of return Difference -130 -650 -2043 -0,8 539 -67 -53 418 N/A. -76% Since the economical lifetime is very short (1 year) and the loading is low, it is clear that the purchase price of the transformer is the most dominant part in the capitalisation of the costs. In fact, the cheapest transformer is not a dry type transformer but an oil-immersed C-C’ one, having total capitalised costs of about Euro 12951,= (including harmonics). At the end of year 1, when the loading has increased from 25% to 70%, following results are obtained (see table 5.17). Table 5.17 Outcome 1600 kVA transformer (end of year 1) Unit Dry trans- Dry transformer former Low losses Transformer rating kVA 1600 1600 Rated no-load loss W 2800 2670 Rated load loss W 10000 9350 Total annual losses kWh/a 133501 125279 CO2 emission @ 0,4 ton/a 53,5 50,1 kg/kWh Purchase price EUR 14451 14990 Present value no-load loss EUR 1444 1378 Difference -130 -650 -8222 -3,4 539 -66 – Page 42 of 57 – Present value load loss Capitalised costs Pay back (years) Internal rate of return EUR EUR 6416 22311 40130017-TDC 02-24226 5999 22366 -417 55 N/A. -4% Even at this relatively high load and much higher losses, the purchase price is still dominant. Energy efficiency does not pay off in this case either. Again, the cheapest transformer is an oil-immersed C-C’ one, having total capitalised costs of Euro 21225,= (including harmonics). Although this transformer is only loaded with 70%, due to the harmonics the loss of life at the end of the year is about 5. If we presume a lifetime about 30 years, this means the actual life time is shorted to 30 / 5 = 6 years. Since the economical lifetime is taken as one year, the reduction in technical life time is no problem as long as the hot spot temperature is below the maximum levels. 5.5.3 Evaluation of smaller transformers Since the loading is low and the economic lifetime is short, one could ask, why one would choose for a 1600 kVA transformer. Instead of the 1600 kVA transformer it is possible to use a 1000 kVA transformer. The loading of the 1000 kVA transformer will initially be 40%. Table 5.18 gives the calculation for the 1000 kVA transformer. Table 5.18 Outcome 1000 kVA transformer (initial loading) Unit Dry trans- Dry transformer former smaller size Transformer rating kVA 1600 1000 Rated no-load loss W 2800 2000 Rated load loss W 10000 8600 Total annual losses kWh/a 38428 48121 CO2 emission @ 0,4 ton/a 15,3 19,2 kg/kWh Purchase price EUR 14451 10074 Present value no-load loss EUR 1444 1032 Present value load loss EUR 818 1802 Capitalised costs EUR 16714 12907 Pay back (years) Internal rate of return Difference -800 -1400 9693 3,9 -4377 -412 984 -3807 N/A. -86% In fact, changing to a smaller transformer seems very attractive: the purchase costs are much lower, and the much higher (+25%) energy losses will not offset the lower purchase costs: capitalised cost saving is approximately 25%. The internal rate of returns has turned negative, which is due to the initial cost advantage and the later periodic extra cost. Financially speaking, the smaller transformer is a good investment! The influence of harmonics will reduce the technical lifetime of the 1000 kVA transformer, but this should not be critical, as the economical lifetime is only 1 year. Obviously, if this company would exist more than one year, it would regret the short sight of the initial investment. Also, if the expected load growth would occur, the transformer would – Page 43 of 57 – 40130017-TDC 02-24226 become overloaded very quickly. 5.5.4 Conclusion For this company, the predominant condition is the short economic scope. Hence, all investments can be viewed in light of initial cost only. In this case, only the setting of minimum energy loss levels in standards forms the limit to buying cheap and inefficient equipment. This somewhat unsatisfactory outcome of the case study is mitigated in another business model, in which the transformer is owned by e.g. the utility. Utilities are used to long planning scopes of their infrastructure and, as they may move equipment along their infrastructure as needed, will probably optimise the transformers with respect to losses as they can be sure the transformer will be employed anywhere during its lifetime. A scope for annual energy saving and CO2 reduction can not be given in this case. 5.6 Case study 6: An office building In a lot of older office buildings there are transformers that are as old as the building. Due to the enormous growth of non-linear loading in office buildings, it is known that in the past there were serious problems with harmonics. One could ask, if the used transformers are capable to take care of these harmonics, since they were not always designed for a non-linear loading. This case study is based on an actual failure caused by non-linear loading in an office building. The given office building was placed in the sixties and on each floor there are two transformers of 400 kVA, each feeding half a floor. The last decennium there is a certain growth of computers. The power supplies of each computer and monitor are causing harmonics. Based on measurements it is found that the harmonics in the current are according to table 5.19. Table 5.19 input data transformers Transformer size 400 kVA oil-immersed transformer (1965) Transformer load 40% (constant load, 10/24h), during office time 20% (constant load, 14/24h) Economic lifetime 10 year Interest rate 6% Energy price EUR 70/MWh Harmonic spectrum 1 3 5 7 9 11 13 15 17 % 100 59 41 23 0 6 5 0 3 A (no-load loss evaluation) EUR 4,51 /W B (load loss evaluation) EUR 0,34 /W 19 2 21 0 23 1 The no-load loss of the older 400 kVA oil-immersed transformer is Pn = 710 W and the load loss Pk = 3925 W (based on figures in 1966/1967). The eddy current loss for the windings is presumed to be 4%, while the other eddy current losses are also presumed to be 4%. Although these losses are relatively high, it is not economical to replace these older transform- 25 1 – Page 44 of 57 – 40130017-TDC 02-24226 ers with new transformers. Since the no-load loss evaluation (A) is high, one could ask if it is cost effective to place new transformers with amorphous cores if one of the existing transformers failed. Table 5.20 shows the outcome for an amorphous core compared with a C-C’ oil transformer. Table 5.20 Outcome 400 kVA transformer (initial loading) Unit C-C’ Oil Amorphous transformer transformer Transformer rating kVA 400 400 Rated no-load loss W 610 160 Rated load loss W 3850 3850 Total annual losses kWh/a 10528 6586 CO2 emission @ 0,4 ton/a 4,2 2,6 kg/kWh Purchase price EUR 4874 6787 Present value no-load loss EUR 2753 722 Present value load loss EUR 2671 2671 Capitalised costs EUR 10298 10180 Pay back (years) Internal rate of return Difference -450 0 -3942 -1,6 1913 -2031 0 -118 6,9 6% As can be seen, the transformer with the amorphous core reduces the losses and can be cost effective. The total energy and emission savings between the existing and amorphous transformer are given in table 5.21. Table 5.21 Annual savings Transformer Total size number 400 kVA 24 Energy saving [MWh] CO2 emission saving [tonnes] 94,6 37,8 Since the average use of electricity per year equals almost 12,5 Million kWh, this means a energy saving of about 0,76%. 5.7 Case study 7: Reliability and availability A lot of industries (e.g. chemical) are interested in a reliable power supply, as an unforeseen interruption of the power supply may have severe consequences. First there is the economical damage if there is a shutdown, which leads to loss of production until the process has restarted. But also there can be damage of the installation. Next to the direct damage, long outages may cause pollution and human safety problems. To reduce the risks of an outage, it is possible to have two transformers in redundancy. This means if one transformer fails, the other transformer will carry the full load, and there will be no interruption of the power and/or shutdown of the factory. For this case study, the economical cost when an unforeseen outage occurs is a necessary input figure. The costs of outages are very hard to determine, for this given situation it is pre- – Page 45 of 57 – 40130017-TDC 02-24226 sumed that outage of the electricity will cause a shutdown of the factory, whereby each hour outage is equal to Euro 10.000,=. The MTBF (mean time between failure) for a transformer is presumed to be 40 years. The MTTR (mean time to repair) or time to replace a failed transformer is 8 hours. This means that the average outage frequency equals 0,025 per year and the average outage duration equals 12 minutes per year. If there is redundancy the average outage frequency equals 1,14*10 -6 per year, while the average outage duration equals 2,7*10 -4 minutes per year. This means the chance that both parallel transformers having a failure at the same time is very small compared to the outage of one transformer. The choice for the designer to use one transformer (2500 kVA) or two transformers (1600 kVA) can now be quantified. Presuming the load is 1500 kVA; the 2500 kVA transformer loading is 60%, while the loading of the 1600 kVA transformer is 47% when both transformers are in parallel. Based on the transformer data shown in chapter 4, the total annual losses for both options are given in table 5.22. Table 5.22 electricity losses over a year 2500 kVA Transformer Oil C-C’ No load kWh/yr. 21900 Load kWh/yr. 69379 Total kWh/yr. 91279 2x 1600 kVA Transformer Oil C-C’ No load kWh/yr. 29784 Load kWh/yr. 54182 Total kWh/yr. 83966 Oil D-D’ 18615 58972 77587 Oil D-D’ 25316 46054 71370 Dry HD 538 37668 56765 94433 Dry HD 538 49056 38702 87758 Dry Low loss 36179 47083 83262 Dry Low loss 46778 36186 82964 If the economic life time is estimated at 10 years, and the electricity price Euro 70,= per MWh, the following costs are expected in these 10 years (see table 5.23). Table 5.23 Costs over 10 years. 2500 kVA Transformer Purchase price [Euro] Cost of no load [Euro] Cost of load [Euro] Cost of outage [Euro] Total cost [Euro] 2x 1600 kVA Transformer Purchase price [Euro] Cost of no load [Euro] Cost of load [Euro] Cost of outage [Euro] Total cost [Euro] Oil C-C’ 24897 15330 48565 20000 108792 Oil C-C’ 27340 20849 37927 <1 86117 Oil D-D’ 29402 13030 41280 20000 103712 Oil D-D’ 35774 17721 32238 <1 85734 Dry HD 538 25527 26368 39736 20000 111631 Dry HD 538 35902 34339 27091 <1 97333 Dry Low loss 27494 25325 32958 20000 105777 Dry Low loss 38146 32745 25330 <1 96222 From table 5.23 it can be seen that the costs of outage are having influence on the total costs over 10 years. If the costs of outage are neglected, the designer would probably have chosen for a single 2500 kVA oil-transformer type D-D’ or a dry transformer with lower losses than given in the HD 428. – Page 46 of 57 – 40130017-TDC 02-24226 However if the average costs of an outage are taken in account, the designer will probably order two 1600 kVA transformers. However, the two 1600 kVA transformers will also need an installation more than the 2500 kVA transformer, which is not taken in account in this case study. Nevertheless, it is clear that redundancy of transformers is preferred anyway for situations were shutdown of a process causes pollution or safety risks. 5.8 CO2 emission trading In all above given cases the influence of CO 2 emissions was not evaluated. Throughout the world it is clear that CO2 emissions and trading can not be omitted in the future. To allow for economic optimisation, emission trading schemes are under discussion, and by that time, CO2 emission rights will have a price. The price for CO 2 emission will add to the cost of electricity. The extra costs are shown in table 3.1 (chapter 3.6). Considering these extra costs for case study 4 table 5.24 gives the payback time, IRR and capitalised costs for different CO2 emissions and costs. Table 5.24 Evaluation with cost for CO2 emissions Costs Eur 0/tonne Eur 10/tonne Eur 33/tonne Eur 50/tonne 0,3 kg/kWh Pay back IRR 4,2 6% 4 8% 3,5 13% 3,2 17% CC -17 26 128 203 CO2 emission 0,4 kg/kWh Pay back IRR 4,2 6% 3,9 9% 3,3 15% 3 20% CC -17 41 176 276 Pay back 4,2 3,7 3 2,6 0,6 kg/kWh IRR 6% 10% 20% 26% CC -17 250 273 423 This example shows that if the cost per tonne CO2 emission is high, it is cost effective to choose the transformer with the low losses. If there is a cost evaluation considered for the CO2 emission, a higher CO2 emission (kg/kWh) makes it more attractive to choose a transformer with low losses. 5.9 Sensitivity of the input parameters As discussed in chapter 5.4, the input parameter can have a great influence on the results. Establishing the right values is very difficult, since a lot of circumstances during the lifetime of the transformer can (and will) change. By variation of the input parameters it is possible to establish which parameters are the most important for making a right decision. This chapter shows an example for making an accurate decision between two 1250 kVA dry type transformers (one with losses according the HD 538; one with lower losses). Table 5.25 shows the parameters that are changed and how much they are changed. The middle values shown form the base case. Table 5.25 Parameter variation Harmonic loading Electricity price (Eur/MWh) CO2 emissions (kg/kWh) Low Medium (base case) High No harmonic loading 40 0,3 12 pulse loading 60 0,4 6 pulse loading 80 0,6 – Page 47 of 57 – CO2 costs (Eur/tonne) Loading profile (%) Economic lifetime (years) Interest (%) Purchase price (%) 40130017-TDC 02-24226 0 20 1 5 80 10 40 5 7 100 33 60 10 9 120 For each situation, the payback period, IRR and capitalised costs are calculated. Table 5.26 and figure 5.1 shows the results for the pay back period in years. Table 5.26 Parameter sensitivity on the payback period Parameter Parameter variation L M H Unit Harmonic spectrum Electricity price CO2 emissions CO2 costs Loading profile Economic lifetime Interest Purchase price None 12 pulse 6 pulse EUR/MWh 40 60 80 kg/kWh 0,3 0,4 0,6 EUR/tonne 0 10 33 % 20 40 60 years 1 5 10 % 5 7 9 % 80 100 120 Payback time (years) L M H 3,3 4,5 3,2 3,3 5,2 3,1 3,1 2,5 3,1 3,1 3,1 3,1 3,1 3,1 3,1 3,1 2,7 2,4 3,0 2,7 1,9 3,1 3,1 3,7 Purchase price Interest Economic life time Loading CO2 value CO2 emission Electricity price Harmonics 0 1 2 3 4 5 6 years Figure 5.1 Parameter sensitivity on the payback period At the average values, the payback period is 3,1 year. As can be seen, the loading profile and electricity price have the most influence on the pay back period. The influence of all other parameters is smaller. By definition, interest and economic lifetime do not have an influence on the payback period at all. This example clearly shows that it makes more sense to determine the expected loading of the transformer than just looking at the purchase price of the trans- – Page 48 of 57 – 40130017-TDC 02-24226 former. In this example it also makes more sense to focus on the electricity price than on the purchase price of the transformers. It can also been seen that in this example the transformer with the low losses always has an acceptable pay back period. This means that, even when a parameter is not correctly specified, the transformer with the low loss is the best choice, however the pay back time will change. Figure 5.2 and table 5.27 show the results for the IRR parameters, where the parameters are changed as shown in table 5.25. Table 5.27 Parameter sensitivity on the IRR Parameter Parameter variation L M H Unit Harmonic spectrum Electricity price CO2 emissions CO2 costs Loading profile Economic lifetime Interest Purchase price None 12 pulse 6 pulse EUR/MWh 40 60 80 kg/kWh 0,3 0,4 0,6 EUR/tonne 0 10 33 % 20 40 60 years 1 5 10 % 5 7 9 % 80 100 120 IRR (%) L M H 16% 3% 17% 15% -1% -68% 18% 29% 18% 18% 18% 18% 18% 18% 18% 18% 25% 31% 20% 24% 45% 30% 18% 11% 60 80 100 Purchase price Interest Economic life time Loading CO2 value CO2 emission Electricity price Harmonics -100 -80 -60 -40 -20 0 20 40 Percentage Fig. 5.2 Parameter sensitivity on the IRR As can be seen in figure 5.2 the most important parameter based on the IRR method, is the expected economic lifetime. Contrary to the method based on the payback time, a different economic lifetime or loading pattern can mean a different choice of the transformer (negative – Page 49 of 57 – 40130017-TDC 02-24226 IRR). Although its influence is smaller, a low electricity price can give a low IRR value (3%). This means that for an interest rate higher than 3%, the transformer with the low losses is not economical. The purchase price is only the fourth dominant factor. This example shows that determining a (too) short economic lifetime of the transformer will cost the industry money. Although the numbers from this example are not applicable for every situation, the method itself is and it is strange that a lot of industries do not take in account the economic life time when deciding which transformer should be used, and are only interested in the lowest purchase price of the transformer. Figure 5.3 and table 5.28 show the results when the capitalisation formula is used. A negative value means it is useful to choose for the transformer with the low losses. When the result is positive, the transformer with the losses according the HD 538 should be used. Table 5.28 Parameter sensitivity on the capitalised cost Parameter Parameter variation L M H Unit Harmonic spectrum Electricity price CO2 emissions CO2 costs Loading profile Economic lifetime Interest Purchase price None 12 pulse 6 pulse EUR/MWh 40 60 80 kg/kWh 0,3 0,4 0,6 EUR/tonne 0 10 33 % 20 40 60 years 1 5 10 % 5 7 9 % 80 100 120 Capitalised cost (Euro) L M H -188 71 -222 -176 158 525 -292 -387 -237 -237 -237 -237 -237 -237 -237 -237 -392 -546 -268 -379 -896 -941 -186 -87 500 750 Purchase price Interest Economic life time Loading CO2 value CO2 emission Electricity price Harmonics -1500 -1250 -1000 -750 -500 -250 0 EURO Figure 5.3 Parameter sensitivity on the capitalisation formula 250 – Page 50 of 57 – 40130017-TDC 02-24226 Again, it is clear that most attention should be given to the used economic lifetime, loading pattern and electricity price. Disregarding these values will certainly have a great influence on deciding which transformer should be used and could give a wrong optimised transformer. Again the influence of the purchase price of the transformer is relatively small. 5.10 Total energy saving potential by efficient transformers in the industrial sector In this section, a rough estimate is made for the total European energy savings potential by application of energy-efficient distribution transformers in industry. Since the load conditions of distribution transformers and ordering procedures are different even within an industry sector, it is not possible to determine one or more transformers which could be typical for the transformer population in a given industry sector. The estimate for Europe therefore will have to be very global. A complicating factor is that distribution transformers do not distribute all electricity consumed by the industries. Depending on the sector of industry, there are HV motors, large drives with dedicated transformers and/or electrolysis processes which are having a large share in the total electricity consumption. Therefore it is very hard to estimate the exact share of electricity consumption that is distributed by distribution transformers. Even between two companies in the same industry sector, there are differences. If we presume the above given cases per sector to be representative for the sector, the annual saving of electricity in Europe can be estimated based on the values given in chapter 2. Table 5.30 estimated electricity saving, based on the case studies Industry sector INDUSTRY, of which: Iron and steel Chemical/petrochemical Non-ferrous metals Non-metallic minerals Machinery Food/tobacco Paper, pulp and print Other industry COMMERCIAL AND PUBLIC SERVICES FINAL SAVINGS approximately Electricity saving (GWh) 650,6 0,084% * 131G = 110,3 0,026% * 195G = 50,7 0,01% * 88G = 8,8 0,06% * 76G = 45,6 0,06% * 99G = 59,4 0,06% * 95G = 57,0 0,14% * 127G = 177,8 0,06% * 235G = 141,0 0,76% * 570G = 4332 5000 This means a total electricity saving of about 0,3%. This equals a saving of 2 million tonnes CO2, which is about 0,6% of the total European aim for reduction of CO 2 until 2012. Of course, the extrapolation of six cases gives rather unreliable results. Therefore, another approach is chosen to estimate the order of magnitude of the energy savings. The industries interviewed for this project employ distribution transformers with relatively low losses. It seems reasonable that the average energy efficiency of distribution transformers – Page 51 of 57 – 40130017-TDC 02-24226 across Europe is lower than in the companies considered. The average losses are estimated at 1,5% of the energy transmitted. For privately owned transformers, an energy savings potential of 0,5% is assumed, i.e. approximately one-third of the losses in these distribution transformers. This results in table 5.31. Table 5.31 estimated annual electricity saving in private distribution transformers Economical sector INDUSTRY, of which: via distribution transformers not via distribution transformers COMMERCIAL AND PUBLIC SERVICES TOTAL Electrici- Losses in disty con- tribution transsumption formers (TWh) (TWh) Savings potential (TWh) 1000 500 500 600 7,5 7,5 2,5 2,5 9 3 1700 16,5 5,5 Privately owned transformer offer a savings potential of approximately 5,5 TWh/year, if all distribution transformers would be replaced by energy-efficient transformers. The associated CO2 emission reduction would amount to 2,2 million tonnes/year, or 0,65% of the 340 Mton emission reduction target of the European Union for 2012. – Page 52 of 57 – 40130017-TDC 02-24226 6. CONCLUSIONS AND RECOMMENDATIONS 6.1 Conclusions The industrial sector is a large energy user, covering nearly half of the electricity use or 1000 TWh annually in Europe. Although these enterprises obtain their electricity from the public medium and high voltage networks, much electricity is consumed at low-voltage level – the conversion to low-voltage is performed by privately owned distribution transformers. A preliminary investigation by interviews with representatives several large Dutch industrial companies and case studies has been carried out to obtain an insight into the energy saving potential in industry through the use of energy-efficient distribution transformers. 6.1.1 Transformer population The population of distribution transformers in industrial enterprises and large office buildings shows some salient differences with their counterparts in the public electricity networks: industrial transformer sizes are typically 1000–4000 kVA, contrary to the public transformers which are in the range 50–1000 kVA the secondary voltage for larger transformers is regularly 690 V instead of 420 V the average load of industrial transformers is relatively high (30-100% or rated load) the newest transformers are often dry-type transformers instead of oil-cooled ones high levels of harmonic pollution of the load current frequently occur (extra losses, risk of extra ageing) a relatively young transformer population (up to 30 years). The total European population of industrial transformers is estimated to be 100.000-150.000 units, with total installed power of approximately 100-150 GVA. Based on the HD 538, HD 428 and the interviews the following differences are observed when comparing a dry- to an oil-cooled transformers: The purchase price of dry transformers is higher than the purchase price of oil-immersed transformers. The no-load losses of a dry transformer are higher, due to their bigger dimensions The load losses however, are at full load lower compared to oil-immersed transformers. Harmonic pollution of the load causes less heating and ageing the dry transformer less than the oil-immersed transformer. However, due to epoxy the heat emission of the drytransformer is less than the oil-immersed transformer. Dry-type transformers are considered better provided against fire5 Dry-type transformers do not need an oil-spilling container. As a rule of thumb, for a lower loading profile, the oil-immersed transformers are cost effective, sometimes even with an amorphous core, however if the load is growing and/or significant harmonic pollution is present, the dry-type transformers are more cost effective. If the losses given in the HD 538 and HD428 are not taken in account, but are based on the based on the loss evaluation ( A and B factors), it is possible to buy transformers with even a It should be noted that there are special ranges of oils (O, K and L) for better fire protection of oil-immersed transformers. 5 – Page 53 of 57 – 40130017-TDC 02-24226 higher energy efficiency than the D-D’ range. This is irrespective of the type of transformer. 6.1.2 Transformer purchasing policies In case of turnkey contracting of entire installations, little control is exerted on the loss levels of transformers purchased. If loss levels are considered, payback times often are very short. Also, energy prices are often very low due to large purchase volumes. These factors are relatively unfavourable for the application of energy-efficient transformers. Some industrial companies are aware of the energy savings by using transformers with low losses. These industries typically do not buy their transformers in a large turnkey contract but separately. An excellent illustration of this practice is one of the interviewed companies which has been applying quite energy-efficient transformers in 1986 as described in case 5.3. The used transformers are cost-effective and even more energy efficient than the losses of the DD’ transformer as given in this report! Transformer standards apparently are not very demanding regarding energy efficiency. Almost every interviewed company has placed dry type transformers, or is considering to place dry type transformers instead of oil immersed transformers. Comparing the losses of a dry type transformer with an oil-immersed transformer and the purchase price, this is remarkable. The losses of a dry type transformer are very high compared with the oil-immersed transformers. In other words, the efforts to lower the losses of an oil transformer in the last decades are wept out by using dry type transformers with higher losses. 6.1.3 Loss evaluation, including the effects of harmonic pollution Special attention has been paid to the effect of harmonic pollution. Harmonics create extra losses and therefore a temperature rise, which has a negative influence of the lifetime of transformers. The influence of harmonics on the transformer is for the dry transformer less than the oil-immersed transformer. In the case studies, the increase of energy losses due to harmonics was significant (over 30%), while the lifetime reduction was rather limited. Transformer loss evaluation with A and B factors (according to HD 428 and 538) is applied in some cases, but this method neglects the influence of higher harmonics: differences up to 30% may occur. In fact, the loss evaluation method with the A and B factors is not well applicable if significant harmonic distortion is present. One of the complications is that in the presence of harmonics, A and B, as presently defined, become dependent on the transformer construction, so can not be used anymore to compare different transformers in a simple way. 6.1.4 Energy saving potential The case studies showed that, although there are some older (oil) transformers used in the industry, it is normally not cost effective to replace these transformers with new ones with lower losses, an exception being the situation where very old (<1970), inefficient transformers are applied. For new transformers, energy-efficient designs are often attractive, provided a “reasonable” payback time (5-7 years) for the increased investment is allowed. – Page 54 of 57 – 40130017-TDC 02-24226 The energy savings potential by application of energy-efficient transformers differs significantly from company to company. A rough estimate of the European potential is 5,5 TWh/a for Europe, equivalent to CO2 emission savings of 2,2 million tonnes/year, or 0,65% of the 340 Mton target for Europe in the year 2012. Because of the long life span of distribution transformers, market penetration of energyefficient transformers in the private sector will only be achieved gradually. 6.2 Recommendations 6.2.1 Purchasing policies and promotion One of the main problems to improve the use of energy efficient transformers is the ignorance of electricity consumers for buying and using energy efficient transformers. Since a lot of transformers are bought as part of turnkey projects, there is a little (or sometimes no) attention for the losses of transformers. Low attention for the evaluation costs will result into an inefficient transformer during the complete lifetime. Therefore the purchaser should be aware of the loss evaluation formula at the moment of ordering the transformer. A problem here is the estimation of these evaluation factors by the purchaser. A good estimation of these factors can save the environment, electricity and costs over a long period of time. The awareness of the purchaser for the no load and load loss evaluation factors should be the way to promote energy efficient transformers. 6.2.2 Standardisation: HD428 and 538 The transformers as given in the HD 428 and HD 538 are based on distribution transformers in the utilities. The industry uses similar transformers, but there is a difference in rating of the used transformers for utilities and industries, so it is proposed to extend and refine the range to make it suitable for the industry. The extension should be up to 4000 kVA rated power, and the rating should become more fine as the load can often be better predicted, which allows a better match between transformer and load. Table 6.1 gives the proposed (extended) range of preferred transformer ratings. Table 6.1 proposed (extended) range of preferred transformer ratings Present range in Proposed range Proposed secondHD428/538 HD 428/538 ary voltages 50 50 420 V 100 100 420 V 160 160 420 V 250 250 420 V 400 400 420 V (500) 420 V 630 630 420 V 800 420 V 1000 1000 420 V 1250 420 V 1600 1600 420 V 2000 420 V – Page 55 of 57 – 2500 2500 3150 4000 40130017-TDC 02-24226 420 V, 690 V 420 V, 690 V 420 V, 690 V Attention should be paid to the influence of harmonics in the HD 428 or HD 538. The use of the economic evaluation factors A and B needs adaptation in order to correctly represent the influence of harmonics. It is proposed to develop a simple system for this. HD 428 contains three lists for no-load and load loss levels. Following modifications are proposed (and similar modifications for HD 538): List C-C’ should be the preferred series. The list with higher no-load loss levels (A’ and B’) should be removed. It is proposed to add a list D’ for reduced no-load losses (15% less no-load losses than list C’) and E’ for strongly reduced no-load losses (70% less no-load losses than list C’, i.e. amorphous cores). The list for the highest load loss levels (B) should be removed, too. The load loss level A (15-20% higher than level C) will be justified in some cases of very light loading. It is recommended to add a list D for reduced load losses (15% less load losses than list C) and a list E for further reduced load losses (30% less load losses than list C) 6.2.3 Savings potential of special transformers and high-voltage motors Although the using of energy saving transformers can be cost effective and saves energy, a lot of electricity consumed in the industries is at a higher voltage rating than 420 or 690 V. A lot of energy is consumed by HV motors or equipment for electrolysis. The most widespread high-voltage load in industrial networks is the high-voltage motor, with rated voltages up to several kV and powers in the megawatt range. It might be worthwhile to investigate the scope for energy savings by application of energy efficient high-voltage motors. Electrolysis is often used in chemical and non-ferrous metal industry, which in Europe are responsible for about 27% of the total electricity consumption. For electrolysis a converter transformer in combination with a rectifier is used to produce a low DC voltage. The rectifier produces high levels of harmonic pollution in the load current. Filtering of these harmonics is very limited due to the high cost of filtering in such high-power installations. Therefore the harmonic loading and the extra losses in the converter transformers may be expected to be very high. It is recommended to assess the potential for energy savings by application of energy efficient converter transformers. 6.2.4 Optimised system voltage in industrial networks Some industrial companies use 3 and 6 kV networks and installations, in order to feed large motors. Networks of these voltages have following drawbacks: components for these voltages are ‘specials’ (so expensive) safety regulations are very stringent for 3 and 6 kV networks, which are formally considered ‘high voltage’. 690 V installations are still ‘low voltage’, requiring less qualified personnel. – Page 56 of 57 – 40130017-TDC 02-24226 Since a lot of electrical equipment is standardised for 420, 690 V or 10 kV, it is proposed to investigate if it would be cost effective to replace existing 3 and 6 kV installations with a 690 V or 10 kV installation [14]. – Page 57 of 57 – 533565406 2001-10-23 REFERENCES [1] The scope for energy saving in the EU through the use of energy-efficient electricity distribution transformers, European Copper Institute, December 1999. [2] Quick scan: Power demand of ICT companies, KEMA TDP-01-16731, January 2001 on behalf of NWN (now NUON), (in Dutch, confidential). [3] Harmonics, Transformers and K-Factors, Copper development Association, CDA publication 144, September 2000. [4] Loading guide for dry-type power transformers, IEC 60905. [5] Loading guide for oil immersed transformers, IEC 60354. [6] Energy in the Netherlands, EnergieNed 2000. [7] Dry-type power transformers, IEC 60726. [8] Power transformers, IEC 60076. [9] Transformers for industrial applications, IEC 61378-1. [10] Semiconductor convertors, general requirements and line commutated convertors part 1-2: Aplication guide, IEC 146-1-2. [11] J & P Transformer book, twelfth edition 1998, Martin J. Heathcote. [12] Availability of the power supply system in the Netherlands – Report for systems from 0,4 kV to 150 kV in 2000; KEMA report (G.A. Bloemhof, W.T.J. Hulshorst), reference 40110014-TDP 01-21151A d.d. Aug 23, 2001. [13] Boete van 50 Euro voor ton teveel kooldioxide; http:www.energiemanagement.net (in Dutch). [14] Use of 690 V for LV industrial distribution network to save capital cost and improve network efficiency, P.F.Lionetto, R.Brambilla, P.Vezzani and E.Picatoste, CIRED 2001