VIEWS: 5 PAGES: 35 CATEGORY: Energy & Green Technologies POSTED ON: 2/17/2013 Public Domain
BEST PRACTICE MAN UAL TRANSFORMERS 1 CON T E N T S 1.1 B ACK GR OU N D .......................................................................................................................................................................3 1.2 A GU I D E T O T H I S GU I D E .................................................................................................................................................3 2 F U ND AME N T AL S .................................................................................................................................................................4 2.1 P R I N CI P L E OF T R AN S F OR ME R ACT I ON ............................................................................................................................4 2.2 L OS S E S I N T R AN S F OR ME R S ...............................................................................................................................................6 2.2.1 Dielectr ic Los s es .....................................................................................................................................................6 2.2.2 Hys ter is is Los s ............................................................................................................................................................6 2.2.3 Eddy Cur r ent Los s es in T he Cor e.........................................................................................................................7 2.2.4 Res is tive los s es in the windings ...........................................................................................................................8 2.2.5 Eddy Cur r ent Los s es in conductor s : .....................................................................................................................9 2.2.6 Ex tr a Eddy Los s es in S tr uctur al Par ts ................................................................................................................10 3 T R ANS F OR ME R OP E R AT I ON ......................................................................................................................................11 3.1 V AR I AT I ON OF L OS S E S D U R I N G OP E R AT I ON .......................................................................................................11 3.1.1 Var iation of los s es with loading level ......................................................................................................11 3.1.2 Var iation in Cons tant los s es ..........................................................................................................................12 3.1.3 Var iation in Load Los s es ..................................................................................................................................12 3.2 L OS S M I N I MI S AT I ON I N A P P L I CAT I ON & OP E R AT I ON ...................................................................................12 3.2.1 S election of Rating and Number of T r ans for mer s ............................................................................12 3.2.2 Ener gy S aving by Under - utilis ation of tr ans for mer s ......................................................................13 3.2.3 Reduction of los s es due to impr ovement of power factor ...........................................................14 3.2.4 S egr egation of nonlinear loads ....................................................................................................................14 3.3 E F F E CT OF OP E R AT I N G T E MP E R AT U R E .....................................................................................................................14 3.4 A S S E S S I N G T H E E F F E CT S OF H AR MON I CS .............................................................................................................15 3.4.1 U.S . Pr actice – K - Factor .................................................................................................................................15 3.4.2 Eur opean Pr actice- ‘Factor K ’ .......................................................................................................................17 4 R E D U CT I ON OF L OS S E S AT D E S I GN S T AGE ......................................................................................................19 4.1 I N T R OD U CT I ON ................................................................................................................................................................19 4.2 M I N I MI S I N G I R ON L OS S E S .........................................................................................................................................19 4.3 M I N I MI S I N G COP P E R L OS S E S ....................................................................................................................................19 5 E CON OMI C AN AL YS I S ...................................................................................................................................................21 5.1 I N T R OD U CT I ON ................................................................................................................................................................21 5.2 T OT AL OW N E R S H I P COS T OF T R AN S F OR ME R S .....................................................................................................21 5.3 D E CI S I ON S F OR CH AN GE OVE R T O N E W E QU I P ME N T ..........................................................................................22 6 CAS E S T U D I E S ...................................................................................................................................................................23 6.1 I N T R OD U CT I ON ................................................................................................................................................................23 6.2 CAS E S T U D Y 1 .................................................................................................................................................................23 6.2.1 I llus tr ative calculations : ..................................................................................................................................24 6.2.2 Factor for Har monics .........................................................................................................................................24 6.2.3 Per centage of Eddy Los s es in Load Los s es : ........................................................................................25 6.2.4 Full load los s es for Har monic Loading: ...................................................................................................25 6.2.5 Relative economics for low los s tr ans for mer s (All Dr y type) for 1250 kVA and 1600 kVA tr ans for mer s . ...............................................................................................................................................................26 6.2.6 S ummar y: ................................................................................................................................................................27 6.3 CAS E S T U D Y -2 : N ON F E R R OU S ME T AL S E CT OR ..................................................................................................28 6.4 CAS E S T U D Y -3 : P AP E R & P U L P COMP AN Y .........................................................................................................29 6.5 CAS E S T U D Y -4 CH E MI CAL I N D U S T R Y ..............................................................................................................30 6.6 CAS E S T U D Y 5 CAS E OF A L AR GE D AT A H OT E L S T AR T U P ........................................................................30 6.7 S U MMAR Y OF E U R OP E AN CAS E S T U D I E S : .............................................................................................................31 6.8 CAS E S T U D Y : T E A I N D U S T R Y ( I N D I A) ................................................................................................................31 2 1 I N T R OD U CT I ON 1.1 B ackgr ound D is tribution trans formers are very efficient, with los s es of les s than 0.5% in large units . S maller units have efficiencies of 97% or above. I t is es timated that trans former los s es in power dis tribution networks can exceed 3% of the total electrical power generated. I n I ndia, for an annual electricity cons umption of about 500 billion kWh, this would come to around 15 billion kWh. Reducing los s es can incr eas e tr ans for mer efficiency. T her e ar e two components that make up tr ans for mer los s es . T he fir s t is " cor e" los s (als o called no- load los s ), which is the r es ult of the magnetizing and de- magnetizing of the cor e dur ing nor mal oper ation. Cor e los s occur s whenever the tr ans for mer is ener gized; cor e los s does not var y with load. T he s econd component of los s is called coil or load los s , becaus e the efficiency los s es occur in the pr imar y and s econdar y coils of the tr ans for mer . Coil los s is a function of the r es is tance of the winding mater ials and var ies with the load on the tr ans for mer . I n s electing equipments , one often conveniently avoid the concept of life cycle cos ting. But the truth is that even the mos t efficient energy trans fer equipment like a trans former, concept of life cycle cos t is very much relevant. T he total cos t of owning and operating a trans former mus t be evaluated, s ince the unit will be in s ervice for decades . T he only proper method to evaluate alternatives is to reques t the manufacturer or bidder to s upply the load and no-load los s es , in watts . T hen, s imple calculations can reveal anticipated los s es at planned loading levels . Frequently, a s mall increas e in purchas e price will s ecure a unit with lower operating cos ts . T he load pr ofile of electr onic equipment—fr om the computer in the office to the var iable s peed dr ive in the factor y—dr ives both additional los s es and unwanted dis tor tion. S ince tr ans for mer manufactur er s tes t only under ideal (linear ) conditions , a s ubs tantial gap ex is ts between publis hed los s data and actual los s es incur r ed after ins tallation. I n fact, tes t r es ults publis hed in a 1996 I EEE T r ans action paper documented an almos t tr ipling of tr ans for mer los s es when feeding 60kW of computer load r ather than linear load. S lightly differ ent pr actices ar e followed in US A and UK to account for har monics while s electing tr ans for mer s . 1.2 A guide to this guide T his Bes t Practice Manual for Electric T rans formers s ummaris e the approach for energy cons ervation meas ures pertaining to s election, application and operation of electric dis tribution trans formers . T he details of des ign methodology and the varied approaches for materials , cons truction are not in the s cope of this manual. However, s ome theoretical as pects are dis cus s ed where ever deemed fit. Chapter-2 dis cus s es principles of trans former action, des cription of los s es and effect of non linear loads on trans former efficiency. Chapter-3 dis cus s es des ign as pects of trans formers to improve efficiency Chapter-4 dis cus s es los s minimis ation in application and operation Chapter-5 dis cus s es principles of economic evaluation of trans formers Chapter-6 dis cus s es cas e s tudies from I ndian and I nternational s cenario 3 2 F U ND AME N T AL S 2.1 P r inciple of t r ans for mer act ion A current flowing through a coil produces a magnetic field around the coil. T he magnetic field s trength H, required to produce a magnetic field of flux dens ity B, is proportional to the current flowing in the coil. Figure 2.1 s hown below explains the above principle Figure 2.1: Relations hip between current, magnetic field s trength and flux T he above principle is us ed in all trans formers . A trans former is a s tatic piece of apparatus us ed for trans ferring power from one circuit to another at a different voltage, but without change in frequency. I t can rais e or lower the voltage with a corres ponding decreas e or increas e of current. ∆B Vp = −NpA ∆t Magneticfield ∆B Vs = −NsA Ip Is ∆t Vp Vs R Vs Ns Np = Ns Vp Np Fig 2.2: T rans former s chematic When a changing voltage is applied to the primary winding, the back emf generated by the primary is given by Faraday’s law, ∆B EMF = Vp = − N pA ----(1) ∆t 4 A Current in the primary winding produces a magnetic field in the core. T he magnetic field is almos t totally confined in the iron core and couples around through the s econdary coil. T he induced voltage in the s econdary winding is als o given by Faraday’s law ∆B Vs = − NsA -----(2) ∆t T he r ate of change of flux is the s ame as that in pr imar y winding. Dividing equation (2) by (1) gives Vs Ns = Vp Np I n Figur e 2.1, the pr imar y and s econdar y coils ar e s hown on s epar ate legs of the magnetic cir cuit s o that we can eas ily under s tand how the tr ans for mer wor ks . Actually, half of the pr imar y and s econdar y coils ar e wound on each of the two legs , with s ufficient ins ulation between the two coils and the cor e to pr oper ly ins ulate the windings fr om one another and the cor e. A tr ans for mer wound, s uch as in Figur e 2.2, will oper ate at a gr eatly r educed effectivenes s due to the magnetic leakage. Magnetic leakage is the par t of the magnetic flux that pas s es thr ough either one of the coils , but not thr ough both. T he lar ger the dis tance between the pr imar y and s econdar y windings , the longer the magnetic cir cuit and the gr eater the leakage. T he voltage developed by tr ans for mer action is given by E = 4.44 x f x N x B max x Acor e, wher e E = r ated coil voltage (volts ), f = oper ating fr equency (her tz), N = number of tur ns in the winding, B max = max imum flux dens ity in the cor e (tes la), and Acor e, = cr os s - s ectional ar ea of the cor e mater ial in S q. metr es . I n addition to the voltage equation, a power equation ex pr es s ing the volt- amper e r ating in ter ms of the other input par ameter s is als o us ed in tr ans for mer des ign. S pecifically, the for m of the equation is kVA = 444 x f x N x B max x Acor e x J x Acond, wher e, N, B max , Acor e and f ar e as defined above, J is the cur r ent dens ity (A/ s q. mm), and Acond is the coil cr os s - s ectional ar ea (mm 2 ) in the cor e window; of the conducting mater ial for pr imar y winding. J depends upon heat dis s ipation and cooling. S am pl e cal cu l at i on A 50 Hz tr ans for mer with 1000 tur ns on pr imar y and 100 tur ns on s econdar y, max imum flux dens ity of 1.5 T es la and cor e ar ea of 0.01 m 2 . J is taken as 2 Amps ./S q. mm and Acond as 30 mm 2 for this illus tr ation. Voltage developed is given by I n pr imar y winding, E pr imar y = 4.44 x f x Np x B max x Acor e, = 4.44 X 50 X 1000 X 1.5 X 0.01 = 3330 Volts 5 E s econdar y = 4.44 x f x Ns x B max x Acor e, = 4.44 X 50 X 1000 X 1.5 X 0.01 = 333 Volts Volt- amper e capability is given by the following : Power r ating = 4.44 x f x Np x B max x Acor e x J x Acond, X 0.001 K VA. = 4.44 X 50 X 1000 X 1.5 X 0.01 X 2 X 30 X 0.001 = 200 kVA appr ox imately. Actual Rated K VA = Rated Voltage X Rated Cur r ent X 10 - 3 for s ingle phas e tr ans for mer s . Rated K VA = V- 3 X Rated Line Voltage X Rated Line Cur r ent X 10 - 3 for thr ee phas e tr ans for mer s . 2.2 L os s es in T r ans for mer s T he los s es in a tr ans for mer ar e as under . 1. Dielectr ic Los s 2. Hys ter es is Los s es in the Cor e 3. Eddy cur r ent los s es in the Cor e 4. Res is tive Los s es in the winding conductor s 5. I ncr eas ed r es is tive los s es due to Eddy Cur r ent Los s es in conductor s . 6. For oil immer s ed tr ans for mer s , ex tr a eddy cur r ent los s es in the tank s tr uctur e. B as ic des cr iption of the factor s affecting thes e los s es is ex plained below. 2.2.1 Dielect r ic L os s es T his los s occur s due to electr os tatic s tr es s r ever s als in the ins ulation. I t is r oughly pr opor tional to developed high voltage and the type and thicknes s of ins ulation. I t var ies with fr equency. I t is negligibly s mall and is r oughly cons tant. ( Gener ally ignor ed in medium voltage tr ans for mer s while computing efficiency ). 2.2.2 H ys t er is is L os s A sizeable contribution to no-load losses comes from hysteresis losses. Hysteresis losses originate from the molecular magnetic domains in the core laminations, resisting being magnetized and demagnetized by the alternating magnetic field. Each time the magnetis ing for ce pr oduced by the pr imar y of a tr ans for mer changes becaus e of the applied ac voltage, the domains r ealign thems elves in the dir ection of the for ce. T he ener gy to accomplis h this r ealignment of the magnetic domains comes fr om the input power and is not tr ans fer r ed to the s econdar y winding. I t is ther efor e a los s . B ecaus e var ious types of cor e mater ials have differ ent magnetizing abilities , the s election of cor e mater ial is an impor tant factor in r educing cor e los s es . Hys ter es is is a par t of cor e los s . T his depends upon the ar ea of the magnetis ing B - H loop and fr equency. Refer Fig 2.3 for a typical B H Loop. 6 Fig 2.3: B - H Loop: Ener gy input and r etr ieval while incr eas ing and decr eas ing cur r ent. Los s per half cycle equals half of the ar ea of Hys ter es is Loop. T he B - H loop ar ea depends upon the type of cor e mater ial and max imum flux dens ity. I t is thus dependent upon the max imum limits of flux ex cur s ions i.e. B max, the type of mater ial and fr equency. T ypically, this accounts for 50% of the cons tant cor e los s es for CRGO (Cold Rolled Gr ain Or iented)s heet s teel with nor mal des ign pr actice. Hys ter is is Los s es , W h = Kh × f × Bm 1.6 Watts /K g. Wher e K h = T he hys ter is is cons tant f = Fr equency in Her tz Bm = Max imum flux dens ity in T es la 2.2.3 E ddy Cur r ent L os s es in T he Cor e T he alter nating flux induces an EMF in the bulk of the cor e pr opor tional to flux dens ity and fr equency. T he r es ulting cir culating cur r ents depends inver s ely upon the r es is tivity of the mater ial and dir ectly upon the thicknes s of the cor e. T he los s es per unit mas s of cor e mater ial, thus var y with s quar e of the flux dens ity, fr equency and thicknes s of the cor e laminations . B y us ing a laminated cor e, (thin s heets of s ilicon s teel ins tead of a s olid cor e) the path of the eddy cur r ent is br oken up without incr eas ing the r eluctance of the magnetic cir cuit. Refer fig 2.4 below for a compar is on of s olid ir on cor e and a laminated ir on cor e. Fig. 2.4B s hows a s olid cor e, which is s plit up by laminations of thicknes s ‘ d1’ and depth d2 as s hown in C. T his is s hown pictor ially in 2.4 A. 7 Fig 2.4: Cor e lamination to r educe eddy cur r ent los s es Eddy Los s es , W e = Ke × Bm 2 × f 2 × t 2 Watts /K g. Wher e K e = T he eddy cur r ent cons tant f = Fr equency in Her tz. Bm = Max imum flux dens ity in T es la t = T hicknes s of lamination s tr ips . For r educing eddy los s es , higher r es is tivity cor e mater ial and thinner (T ypical thicknes s of laminations is 0.35 mm) lamination of cor e ar e employed. T his los s decr eas es ver y s lightly with incr eas e in temper atur e. T his var iation is ver y s mall and is neglected for all pr actical pur pos es . Eddy los s es contr ibute to about 50% of the cor e los s es . 2.2.4 R es is t ive los s es in t he w indings T hes e r epr es ent the main component of the load dependent or the var iable los s es , des ignated as I 2 R or copper los s es . T hey var y as s quar e of the r .m.s cur r ent in the windings and dir ectly with d.c. r es is tance of winding. T he r es is tance in tur n var ies with the r es is tivity, the conductor dimens ions ; and the temper atur e. ρ ×l R= A Wher e R = Winding r es is tance, Ω ρ = Res is tivity in Ohms - mm 2 /m. l = Length of conductor in metr es A = Ar ea of cr os s s ection of the conductor , mm 2 I n addition, thes e los s es var y with winding temper atur e and thus will var y with the ex tent of loading and method of cooling. T he winding r es is tance at a temper atur e T L is given by the following equation. TL + 235 RL = R 0 × T he cons tant 235 is for Copper . For Aluminium, us e 225 or T 0 + 235 227 for Alloyed Aluminium. Wher e R 0 = Winding r es is tance at temper atur e T 0 , Ω RL = Winding r es is tance at temper atur e, T L, Ω 8 T he r .m.s value of cur r ent will depend upon the load level and als o the har monic dis tor tion of the cur r ent. 2.2.5 E ddy Cur r ent L os s es in conduct or s : Conductors in trans former windings are s ubj ected to alternating leakage fluxes created by winding currents . Leakage flux paths , which pas s through the cros s s ection of the conductor, induce voltages , which vary over the cros s s ection. T hes e varying linkages are due to s elf-linkage as als o due to proximity of adj acent current carrying conductors . T hes e induced voltages , create circulating currents within the conductor caus ing additional los s es . T hes e los s es are varying as the s quare of the frequency. For an is olated conductor in s pace, the varying s elf-linkage over the s ection, leads to clus tering of the current near the conductor periphery. T his is known as S kin Effect. T he s ame effect, with the addition of flux from s urrounding conductors , (Proximity effect) leads to extra los s es in thick conductors for trans former windings . T hes e los s es are termed as Eddy Current Los s es in conductors . T he T es t Certificate mentions the load los s es , which include thes e eddy los s es in conductors at s upply frequency (50 Hertz) as als o the eddy los s es in tank s tructure in general at the s ame frequency in the cas e of oil cooled trans formers . For dry type trans formers , tank los s es are abs ent. T he contribution of eddy los s es including tank los s es , over the bas ic copper los s es for an equivalent D.C. current, can be es timated from the difference in meas ured load los s es and expected copper los s es at the tes t current at the tes t temperature. For normal des igns it ranges from 5% to 15% . Detailed s ubdivis ion is available only from des ign data. I t can be taken as 10% of load los s es in the abs ence of s pecific des ign data. T hes e extra los s es vary with s quare of frequency and s quare of per unit harmonic current. T he eddy los s es in the tank s tructure are equivalent to the dis s ipation in a loaded s econdary with leakage reactance. T he variation is not as the s quare of frequency, and it is cus tomary to take a value of 0.8 for the exponent. T he Eddy los s es in a thick conductor can be reduced by decreas ing the radial thicknes s by s ectionalis ing the conductors ( multi-s tranded) and increas ing the axial dimens ion. T he s ectionalis ed conductor has to be trans pos ed to make it occupy all pos s ible pos itions to equalis e the e.m.fs to the extent pos s ible. A simplified expression for eddy current losses in conductors is given below. w 1 2 N-1 N LC L W/N Fig 2.5 : Sectionalised transformer winding - Schematic 9 T he total radial thicknes s of conductor of W cm is s ubdivided into N parts of W/N thicknes s each. Ke is the ratio of the total los s es including eddy los s , to the los s due to D.C. current. 2 N 1 + (αW / N) × 4 Ke = 9 Where α = (π × 4π × 10 −7 ) × f × Lc where 4 π X 10 –7 is permeability of s pace. ρ ×L Where Lc = Axial length of coil. L = Window Height W = Radial total conductor width in metres W’= Width per s ubdivis ion W/N in centimetres . ρ = Res is tivity, in Ohm-metres Lc For Copper at 60C, α ≈ 100 × . ρ= 2 X 10 –8 Ohm-metres L I f W’ is in cm, W = W’/100 Lc Hence αW/N ≈ w '× , α4 is thus proportional to f2 . L As the number of s ubdivis ions increas e, W’ becomes s maller and Ke comes nearer to 1; but always above 1. For a given geometry, eddy los s es increas e as s quare of frequency. I t is important to trans pos e each layer s o that each layer is connected in s eries with a path in each one of the pos s ible N pos itions before being paralleled. T hus circulating current is forced to flow in a relatively very thin conductor. 2 .2 .6 E xt r a E ddy L os s es in S t r uct ur al P ar t s S ome leakage flux, invariably goes in air paths away from the trans former. S trength of this s tray flux diminis hes and varies invers ely with dis tance. I f it links with any conducting material, it will produce eddy los s es in that material. For oil immers ed trans formers , s ome s tray flux links with s ome parts of the tank and caus es extra eddy current los s es in the s tructure. T hes e los s es are abs ent in dry type trans formers . S imilarly, extra flux due to outgoing L.T . conductors carrying large currents caus e extra eddy current los s es in the s tructural portion s urrounding the leads . 0.8 Both thes e los s es vary with frequency , as s tated earlier. T he above dis cus s ion on trans former los s es is given only to gain familiarity with the fundamental principles . T he mos t important los s es are core los s and copper los s . T he other los s es are des cribed mainly to give a complete picture on los s es . 10 3 T R AN S F OR ME R OP E R AT I ON 3.1 Var iation of losses dur ing oper ation T he los s es vary during the operation of a trans former due to loading, voltage changes , harmonics and operating temperature. 3.1.1 Var iat ion of los s es w it h loading level = Output ×100 % Efficiency Output + Losses = P × kVA rating × p.f . × 1000 × 100 P × kVA rating × p.f . × 1000 + N.L + L.L. × P 2 × T Where, P = Per unit loading N.L. = No load los s es in Watts L.L. = Load los s es in Watts at full load, at 75 C T = T emperature correction factor p.f. = Load power factor T he bas ic D.C. res is tance copper los s es are as s umed to be 90% of the load los s es . Eddy current los s es ( in conductors ) are as s umed to be 10% of the load los s es . Bas ic I 2 R los s es increas e with temperature, while eddy los s es decreas es with increas e in temperature. T hus , 90% of the load los s es vary directly with ris e in temperature and 10% of the load los s es vary invers ely with temperature. Calculations are us ually done for an as s umed temperature ris e, and the ris e in temperature is dependant on the total los s es to be dis s ipated. Operating temperature = Ambient temperature + T emperature ris e T o es timate the variation in res is tance with temperature, which in turn depends on the loading of the trans former, the following relations hip is us ed. RT − op F + Tamb + Trise = RT − ref F + Tref Where F= 234.5 for Copper, = 225 for Aluminium = 227 for alloyed Aluminium R T -op = Res is tance at operating temperature T ref = S tandard reference temperature, 75 C T emperature correction factor, T = Load losses at operating temperatur e Load losses at reference temperatur e RT − op RT − ref = 0.9 × + 0.1 × RT − ref RT − op I f a more realis tic s ubdivis ion of load los s es is known from des ign data, the above expres s ion can be modified accordingly. RT − op 234.5 + 100 I f operating temperature is 100 C, = = 1.0808 RT − ref 234.5 + 75 11 Hence T = 0.9 x 1.0808 + 0.1/1.0808 = 1.06523 3.1.2 Var iat ion in Cons t ant los s es T he ir on los s meas ur ed by no load tes t is cons tant for a given applied voltage. T hes e los s es var y as the s quar e of the voltage. Var iat ion in ir on los s es due t o s ys t em volt age har monics : T he s ys tem input voltage may contain voltage harmonics due to aggregate s ys tem pollution in the grid. T he current harmonics of the local harmonic load adds to this by caus ing additional harmonic voltage drop depending upon magnitude of a particular harmonic and the s ys tem s hort circuit impedance at the point of s upply, and the trans former impedance for that s pecific harmonic frequency. T he combined total harmonics affect the flux waveform and give added iron los s es . T he increas e in cons tant los s is quite s mall, due to this voltage dis tortion. 3.1.3 Var iat ion in L oad L os s es About 90% of the load los s es as meas ur ed by s hor t cir cuit tes t ar e due to I 2 R los s es in the windings . T hey var y with the s quar e of the cur r ent and als o with winding temper atur e. Load Los s es = (Per Unit Loading)2 × Load Losses at Full Load × F + Top F + Tref F = T emperature coefficient = 234.5 for Copper and 227 for Aluminium. T ref = 75 °C us ually, or as pres cribed in the tes t certificate Var iat ion in load los s es due t o load pow er f act or : Any reduction in current for the s ame kW load by improvement in p.f. reduces load los s es . Var iat ion in los s es due t o cur r ent har monics : T he s ys tem current harmonics increas e the r.m.s current and thus increas e the bas ic I 2 R los s es . I n addition, the maj or increas e comes from the variation in eddy current los s es in the windings (Us ually 5 to 10% of the total load los s es ), which vary with the s quare of the frequency. 3.2 L oss Minimisation in Application & Oper ation T rans formers have a long life and do not generally s uffer from technical obs oles cence. T he application details are not clearly known during s election and the load and the type of load als o changes with time. Hence trans former rating is likely to be over-s pecified. However, this is generally not a dis advantage from the view point of energy cons umption. T he us ual bes t efficiency point is near 50% load. 3.2.1 S elect ion of R at ing and Number of T r ans f or mer s I n general, s election of only one trans former of large rating gives maximum efficiency and s impler ins tallation. For large plants with long in plant dis tances , two or more trans formers of equal rating may be s elected. Moreover for critical continuous operation plants , power may be had from two independent feeders at s imilar or different voltage levels . I n all s uch cas es , each trans former may be s ufficient to run the plant. T hus normal operation may be at 50% load. S uch a s ituation can lead to lower than 25% load at times . For non- continuous operation of plants with holidays or s eas onal indus tries , s witching off one trans former to s ave part load los s es is generally cons idered. Planning for growth of loads and addition of non linear loads is becoming increas ingly important. T he factors to be cons idered are: 12 • Expected growth of load over around five to ten years • Margin for minimum 15 to 20% growth • 10 to 15% margin for non-linear loads • Availability of s tandard rating Generally, 30 to 50% exces s capacity, reduces load los s es , but the extra firs t cos t is rarely j us tified by energy s aving alone. On the contrary, a clos e realis tic es timate permits extra firs t cos t on a s maller trans former des igned on the bas is of Leas t T otal Owners hip Cos t ( T OC) bas is . Economic evaluation of trans formers is dis cus s ed in chapter 5. For nonlinear loads , trans formers with minimum eddy los s es in total load los s is preferred. T rans former los s es may be s pecified at a s tandard reference temperature of 75 C. T hey have to be corrected to expected s ite operating temperature. Bas ic I 2 R los s es increas e with temperature, while eddy los s es decreas e with increas e in temperature. For nonlinear loads , the derating factor may be worked out taking a K-factor of 20. Details of K factor evaluation is given in s ection 3.4 of this chapter. T his will need derating of 12% for 10% nonlinear load to about 27% for 40% nonlinear load. T he load factor affects the load los s es materially and an es timate of annual r.m.s . load current value is us eful. T rans formers with relatively low no load los s es ( Amorphous Core T ype) will maintain good efficiency at very low loads and will help in cas es where high growth is expected, but ris k of s low growth is to be minimis ed. 3.2.2 E ner gy S aving by U nder -ut ilis at ion of t r ans f or mer s T able 3.1 s ummaris es the variation in los s es and efficiency for a 100 kVA trans former and als o s hows the difference in los s es by us ing a 1600 kVA trans former for the s ame. T he 1000 kVA trans former has a no load los s of 1700 watts and load los s of 10500 Watts at 100% load. T he corres ponding figures for 1600 kVA trans former are 2600 Watts and 17000 Watts res pectively. Loading is by linear loads . T emperatures as s umed equal. T able 3.1: Comparis on of trans former los s es 1 0 0 0 kVA, 1 6 0 0 kVA. Dif f er ence No load los s es = 1 7 0 0 W No load los s es = in los s es , 2600 W W P er L oad T ot al Out put , E f f iciency, L oad T ot al unit los s es , los s es , kW % los s es , los s es , load W W W W 0.1 105 1805 100 98.23 60 2660 861 0.2 420 2120 200 98.95 265 2865 745 0.3 945 2645 300 99.13 597 3197 552 0.4 1680 3380 400 99.16 1062 3662 282 0.5 2625 4325 500 99.14 1660 4267 -58 0.6 3780 5480 600 99.09 2390 4990 -490 0.7 5145 6845 700 99.03 3258 5853 -992 0.8 6720 8420 800 98.96 4250 6850 -1570 0.9 8505 10205 900 98.88 5379 7979 -2226 1.0 10500 12200 1000 98.78 6640 9240 -2960 T he efficiency of 1000 kVA trans former is maximum at about 40% load.Us ing a 1600 kVA trans former caus es underloading for 1000 kW load. T he las t column s hows the extra power los s due to overs ized trans former. As expected, at light loads , there is extra los s due to 13 dominance of no load los s es . Beyond 50% load, there is s aving which is 2.96 kW at 1000 kW load. T he s aving by us ing a 1600 kVA trans former in place of a 1000 kVA trans former at 1000 kW load for 8760 hours /annum is 25960 kWh/year. @ Rs 5.0 /kWh ,this is worth Rs 1.29 lakhs . T he extra firs t cos t would be around Rs 15.0 lakhs . Hence deliberate overs izing is not economically viable. 3.2.3 R educt ion of los s es due t o impr ovement of pow er f act or T rans former load los s es vary as s quare of current. I ndus trial power factor vary from 0.6 to 0.8. T hus the loads tend to draw 60% to 25% exces s current due to poor power factor. For the s ame kW load, current drawn is proportional to KW/pf. I f p.f. is improved to unity at load end or trans former s econdary, the s aving in load los s es is as under. S aving in load los s es 1 2 2 − 1 pf = (Per unit loading as per kW) X Load los s es at full load X T hus , if p.f is 0.8 and it is improved to unity, the s aving will be 56.25% over exis ting level of load los s es . T his is a relatively s imple opportunity to make the mos t of the exis ting trans former and it s hould not be mis s ed. I t s hould als o be kept in mind that correction of p.f downs tream s aves on cable los s es , which may be almos t twice in value compared to trans former los s es . 3.2.4 S egr egat ion of nonlinear loads I n new ins tallations , non-linear loads s hould be s egregated from linear loads . Apart from eas e of s eparation and monitoring of harmonics , it can be s upplied from a trans former which is s pecially des igned for handling harmonics . T he propagation of harmonics can be controlled much more eas ily and problems can be confined to known network. Perhaps a s maller than us ual trans former will help in coordinating s hort circuit protection for network as well as active devices . T he only dis advantage apart from additional cos t is the increas ed interdependence of s ens itive loads . 3.3 E ffect of oper ating temper atur e T he los s es have to be dis s ipated through the s urface area. When the trans former volume increas es , the ratio of s urface area to volume reduces . T hus , larger trans formers are difficult to cool. Oil cooling us es a liquid ins ulating medium for heat trans fer. I n cold countries the ambient temperature is lower, giving a lower operating temperature. I n tropical countries , ambient temperature is higher giving a higher operating temperature. Oil cooled trans formers operate at lower temperatures compared to dry type trans formers . Every 1C ris e in operating temperature gives about 0.4% ris e in load los s es . A reference temperature of 75 C is s elected for expres s ing the los s es referred to a s tandard temperature. T he operating temperature limit is decided by the type of ins ulation us ed and the difficulties of cooling. T his gives an additional factor for comparing los s es during des ign. Higher temperature permits reduction in material content and firs t cos t. Operating temperature beyond the limits pres cribed for the ins ulation, reduces life expectancy materially. Oil cooled trans formers operate at lower temperatures compared to dry type trans formers . 14 3.4 Assessing the effects of H ar monics Load los s performance of a des ign or an ins talled trans former with known data can be done if the levels of harmonic current are known or es timated. I EC 61378-1 ‘T rans formers for I ndus trial Applications ’ gives a general expres s ion for es timating load los s es for loads with harmonics . T his s tandard is s pecifically meant for trans formers and reactors which are an integral part of converters . I t is not meant for power dis tribution trans formers . T he method is applicable for es timation in power dis tribution trans formers . I t can be us ed for oil cooled trans formers or dry type trans formers . T he alternative approaches for power dis tribution trans formers us ing K-Factor and Factor-K are given later. As per I EC 61378-1 the total load los s es with current harmonics are given as under IL 2 n Ih 2 2 n Ih 2 0.8 PT = PDC1 × + PWE1 × ∑ × h + (PCE1 + PSE1)× ∑ × h I1 1 I1 1 I1 Where PT = T otal load los s es and ‘h’ is the order of the harmonic. 2 IL2 = ∑ In 2 n =1 PDC1 = Bas ic copper los s es for fundamental frequency PWE1 = Winding eddy los s es for fundamental PCE1 = Eddy los s es in s tructural parts due to current leads for fundamental PS E1 = Eddy los s es in s tructural parts for fundamental In = Current for harmonic order n I1 = Fundamental current PCE1 and PS E1 are not applicable to dry type trans formers 3.4.1 U .S . P r act ice – K - F act or T he K-Factor rating as s igned to a trans former and marked on the trans former cas e in accordance with the lis ting of Underwriters Laboratories , is an index of the trans former's ability to s upply harmonic content in its load current while remaining within its operating temperature limits . T he K-Factor is the ratio of eddy current los s es when s upplying non-linear loads as compared to los s es while s upplying linear loads . I n U.S ., dry type of trans formers are us ed in maj ority of applications . 2 k = ∑ In 2 .n2 n =1 I n= Per unit harmonic current , and n = Order of harmonic. For s pecification in general, the U.S . practice is to es timate the K – Factor which gives ready reference ratio K for eddy los s es while s upplying non-linear loads as compared to linear loads . K = 1 for Res is tance heating motor s , dis tr ibution tr ans for mer s etc. 15 K = 4 for welder s I nduction heater s , Fluor es cent lights K = 13 For T elecommunication equipment. K = 20 For main fr ame computer s , var iable s peed dr ives and des ktop computer s . T he eddy los s es in conductor s , ar e as s umed to var y as (I I ) × n n 2 2 wher e I is the total r .m.s . cur r ent and is as s umed to be 100 % i.e. r ated value. I= (I 1 2 + I2 2 + .... + In 2 ) wher e I 1 is taken as 1. Now, s ince I is defined, los s var iation is taken as (I I ) × n n 2 2 including fundamental. K is r atio of Eddy los s es at 100 % cur r ent with har monics and Eddy los s es at 100 % cur r ent with fundamental. ∑( ) × n (I / I ) × 1 n 2 2 2 K = In 2 n =1 I 1 1 ∑( ) ×n n 2 K = In 2 n =1 I T he K - Factor is us ed dir ectly to s pecify tr ans for mer s for a given duty. T he total los s es , if needed can be es timated at any X % loading as under if the contr ibution of eddy los s es in load los s es at fundamental fr equency tes t is known fr om des ign; or as s umed typically as 10 % . Copper los s es ar e then as s umed to be the balance 90 % . T otal load los s es at 100 % load = ( 0.9 + 0.1 x K ) I f K = 11, eddy los s es at 100% load with this har monic patter n ar e 11 times the eddy los s es at fundamental. T otal load los s es at 100% load = 0.9 + 1.1 = 2 2 T otal load los s es at x % load = x x 2. I f total load los s es ar e as s umed to be 100% or 1 for s ame temper atur e r is e, then x 2= 1/K = 1/2. x = 1/K 0.5 or 70.7 % . T hus the tr ans for mer can wor k at 70% of its r ated load cur r ent s pecified for linear loads . A s ample K - factor calculation is given for a given s et of har monic meas ur ements , bas ed on the above r elations hips . 16 T able 3-2: Es timation for K factor Har monic RMS I n/I 1 (I n/I 1 ) 2 (I n/I ) (I n/I ) 2 (I n/I ) 2 x n2 No. Cur r ent 1 1 1 1 0.6761 0.4571 0.4571 3 0.82 0.82 0.6724 0.5544 0.3073 2.7663 5 0.58 0.58 0.3364 0.3921 0.1538 3.8444 7 0.38 0.38 0.1444 0.2569 0.0660 3.2344 9 0.18 0.18 0.0324 0.1217 0.0148 1.2000 11 0.045 0.045 0.0020 0.0304 0.0009 0.1120 T otal r .m.s 1.479 S um 2.1876 11.6138 I r .m.s . = √ 2.1876 = 1.479 = I . K - Factor is given by las t column. K factor = 11.618 A K 13 r ated tr ans for mer is r ecommended for this load. 3.4.2 E ur opean P r act ice- ‘F act or K ’ T he Eur opean pr actice as defined in B S 7821 Par t 4 and HD 538.3.S 1 defines a der ating factor for a given tr ans for mer by a ‘Factor- K ’. 0.5 e I1 2 N In 2 1 + ×∑n × q K = I 1 + e I n+2 1 e = Eddy cur r ent los s at fundamental fr equency divided by los s due to a D.C. cur r ent equal to the r .m.s . value of the s inus oidal cur r ent. In = magnitude of nth har monic cur r ent. q = Ex ponential cons tant dependent on type of winding and fr equency = 1.7 for r ound / r ectangular s ection = 1.5 for foil type low voltage winding. I = R.M.S . value of the cur r ent including all har monics 0. 5 n =N 2 = ∑ In n =1 T he obj ective is to es timate the total load los s es at 100% current, when that current contains harmonics . T he bas e current is thus I the r.m.s . current which is 100% . T his is equal to the rated current at which the load los s es are meas ured at fundamental frequency. T he bas ic copper los s es vary as the s quare of the r.m.s . current and hence are equal to the meas ured los s es at fundamental frequency. T otal load los s es at fundamental are taken as unity i.e. 1. 2 2 1= I R + Eddy Los s es , Eddy Los s es = e x I R los s es . 2 1= I R ( 1 + e) 2 2 Eddy Los s es as a fraction of total load los s es = e x I R / I R( 1 + e ) = e / 1 + e ( Eddy Los s es at I ( 100% ) = e / 1 + e x n= 1 ) Σ In / I ( ) 2 x nq S ince harmonics are expres s ed as fractions of fundamental, e I1 × ×∑ n 2 ( I12 × 1q + I3 2 × 3q + ... + In 2 × nq ) Eddy Los s es = 1 + e I n =1 I12 17 e I1 n × × 1 + ∑ 2 ( I1 2 × 1q + I3 2 × 3 q + ... + In 2 × n q ) = 1 + e I n =n + 2 I1 2 2 2 2 I = I1 + IH where IH equals the s um of s quares for harmonics , but excluding fundamental. T otal los s es = 2 e I12 + IH 2 I R+ × ( ) e IH e I1 2 − 2 × 2 + n × × ∑ ×n In q 2 2 I 1+ e I 1+ e 1+ e I n = n + 2 I1 2 I f the term for I H is neglected, there is an error on s afe s ide with a total deviation of only 2% to 4% depending upon I H, s ince e/ 1+ e its elf is about 9% to 10% of total los s es at fundamental. T he addition to eddy los s es may be 10 to 15 times due to harmonics . T he firs t two terms equal the total los s es at fundamental and thus equals 1.T he Factor K is taken as the s quare root of total los s es . T he expres s ion thus s implifies to the form s tated earlier. T he s ummation term is for n > 1 and thus covers harmonics only. 2 2 At X % load, Load Los s es = X K and s ince new load los s es s hould be equal to 1, X = 1/K . T ypical calculation (taking q as 1.7 and as s uming that eddy cur r ent los s at fundamental as 10% of r es is tive los s i.e. e= 0.1) is given below. T able 3-1: es timation of Factor K Har monic No. RMS Cur r ent I n/I 1 (I n/I 1 ) 2 nq nq (I n/I ) 2 1 1 1 1 1 1 3 0.82 0.82 0.6724 6.473 4.3525 5 0.58 0.58 0.3364 15.426 5.1893 7 0.38 0.38 0.1444 27.332 3.9467 9 0.18 0.18 0.0324 41.900 1.3576 11 0.045 0.045 0.0020 58.934 0.1193 S um 2.1876 = Ó 15.9653 I r .m.s . = √ 2.1876 = 1.457. 2 2 K = 1 + (0.1/ 1.1) x ( 1/1.457) x ( 15.9653 –1 ) = 1.641 K = 1.28 T r ans for mer der ating factor = 1 /K = 1/1.28 x 100 = 78.06% 18 4 R E D U CT I O N OF L OS S E S AT D E S I GN S T AGE 4.1 I ntr oduction T he des ign approaches for reduction of los s es are well known and proven. T hey cons is ts of 1. Us ing more material 2. Better material 3. New Material 4. I mproved dis tribution of materials 5. I mprovement in cooling medium and methods Each des ign tries to achieve des ired s pecifications with minimum cos t of materials or minimum weight or volume or minimum overall cos t of owners hip. Worldwide, more and more cons umers are now purchas ing trans formers bas ed on the total owners hip cos ts , than j us t the firs t cos t. 4.2 Minimising I r on L osses T he evolution of materials us ed in trans former core is s ummaris ed below. YEAR CORE MAT ERI AL T HI CK NES S Los s (appr ox .) (mm) (W/kg at 50Hz) 1910 War m r olled FeS i 0.35 2 (1.5T ) 1950 Cold r olled CRGO 0.35 1 (1.5T ) 1960 Cold r olled CRGO 0.3 0.9 (1.5T ) 1965 Cold r olled CRGO 0.27 0.84 (1.5T ) 1975 Amor phous metal 0.03 0.2 (1.3T ) 1980 Cold r olled CRGO 0.23 0.75 (1.5T ) 1985 Cold r olled CRGO 0.18 0.67 (1.5T ) T here are two important core materials are us ed in trans former manufacturing. Amorphous metal and CRGO. I t can be s een that los s es in amorphous metal core is les s than 25% of that in CRGO. T his material gives high permeability and is available in very thin formations ( like ribbons ) res ulting in much les s core los s es than CRGO. T he trade off between the both types is interes ting. T he us e of higher flux dens ities in CRGO (upto 1.5 T ) res ults in higher core los s es ; however, les s amount of copper winding is required, as the volume of core is les s . T his reduces the copper los s es . I n amorphous core, the flux dens ity is les s and thinner laminations als o helps in reducing core los s es . However, there is relatively a larger volume to be dealt with, res ulting in longer turns of winding, i.e higher res is tance res ulting in more copper los s es . T hus iron los s es depend upon the material and flux dens ity s elected, but affect als o the copper los s es . I t becomes clear that a figure for total los s es can be compared while evaluating operating cos t of the trans formers . T he total operating cos t due to los s es and total inves tment cos t forms the bas is of T otal Owners hip Cos t of a trans former. 4.3 Minimising Copper losses 2 T he maj or portion of copper los s es are I R los s es . Us ing a thicker s ection of the conductor i.e. s electing a lower current dens ity can reduce the bas ic I 2R los s es . However, an arbitrary increas e in thicknes s can increas e eddy current los s es . I n general, decreas ing radial thicknes s by s ectionalis ation leads to reduction in eddy current los s es . A properly configured foil 19 winding is us eful in this context. T he des igner has to take care of the proper buildup of turns with trans pos ition and als o take care of the mechanical s trength to s us tain s hort circuit in addition to needed ins ulation and s urge voltage dis tribution. All the s ame, des igners can always try to get minimum bas ic I 2 R and minimum eddy current los s es for a given des ign and s pecified harmonic loading. 20 5 E CO N O M I C A N A L Y S I S 5.1 I ntr oduction For any inves tment decis ion, the cos t of capital has to be weighed agains t the cos t/benefits accrued. Benefits may be in cas h or kind, tangible or intangible and immediate or deferred. T he benefits will have to be converted into their equivalent money value and deferred benefits have to be converted into their pres ent worth in money value for a proper evaluation. S imilarly , future expens es have to be accounted for. T he cos t of capital is reckoned as the rate of interes t, where as the purchas ing power of the currency meas ured agains t commodities determines the relative value of money in a given economic domain. T hus interes t rates increas es value of capital where as inflation degrades the value of capital. T he deferred monetary gains /expens es are expres s ed in terms of their pres ent worth( PW). I f Rs 90.91 is inves ted at an annual interes t of 10% , it will yield 90.91x (1+ 10/100) = Rs 100/- at the end of one year. Hence the pres ent worth of Rs 100 after one year is Rs 90.91/- , if the annual rate of interes t is 10% . n 1 + a 1− 1+ i PW = where PW is pres ent worth. i−a a = per unit inflation index, annual i = per unit interes t rate n = number of years Purchas e of a trans former involves firs t cos t and s ubs equent payment of energy charges during a given period. T he effective firs t cos t or the total owners hip cos t can be had by adding the pres ent worth of future energy charges . T he T OC E F C i.e. . T otal Owners hip Cos t: - Effective Firs t Cos t adds an appropriate amount to account for energy expens es and s hows a better meas ure of comparing an equipment with higher firs t cos t, but having a higher efficiency and thus lower running charges . T he concept of evaluation can be applied to trans formers with the as s umptions that the annual los s es and the load level remain s teady at an equivalent annual value, the tariff is cons tant and the rates of inflation and interes t are cons tant. T hes e as s umptions have obvious limitations , but the T OC E F C concept is widely us ed method for evaluation. T he period of ‘n’ years may be 10 to 15 years . T he longer the period, greater the uncertainty. Generally, ‘n’ will be roughly equal to the economic life of the equipment governed by the technical obs oles cence, phys ical life and perceptions of return of capital of the agency making the inves tment decis ion. 5.2 T otal Ow ner ship cost of tr ansfor mer s T OC E F C = P r ice + Cos t of Cor e los s + Cos t of L oad los s Cos t of cor e los s EFC = A X Core los s in Watts Cos t of L oad los s EFC = B X Load los s in Watts Where A = Equivalent firs t cos t of No load los s es , Rs /Watt 21 PW × EL × HPY = 1000 PW = Pres ent worth, explained in previous s ection 5.1 EL = Cos t of electricity, Rs / Kwh, to the owner of the trans former HPY = Hours of operation per year B = Equivalent firs t cos t of load los s es = A × p2 × T P = Per Unit load on trans former T = T emperature correction factor, details of calculation given in s ection 3.1.1. 5.3 Decisions for changeover to new equipment I n this cas e there is an added cos t of the exis ting working equipment. T he value left in a working equipment can be evaluated either by its technical worth, taking its left over life into cons ideration or by the economic evaluation by its depreciated value as per convenience. For trans formers , the prediction of life is very difficult due to varying operating parameters . Moreover, for any equipment, there is a s alvage value, which can be taken as equivalent immediate returns . T hus T OCE F C = ( Pres ent depreciated effective cos t of old equipment – S alvage value ) + A X Core los s + B X Load los s 22 6 CA S E S T U D I E S 6.1 I ntr oduction Five cas e s tudies are pres ented from European data as pres ented in the publication ‘ Energy S aving in I ndus trial Dis tribution T rans formers ’ From KEMA, Netherlands . One cas e s t udy from I ndian indus try is given. T he cas e s tudies fr om K EMA, as s ume full details of No Load Los s and Load Los s as well as por tion of Eddy Los s es in Load Los s as being available fr om tr ans for mer manufactur er or fr om r elevant s tandar d. No tes ts ar e conducted at s ite. T he har monic content of the load is given for each typical application. T he applicability of Low Los s des igns in each r ating is analys ed and payback per iod is found out. T he cas e s tudies als o give the ener gy s aving gains in ter ms of r eduction in car bon diox ide (Co2) emis s ion. T he likely penalty/gain per T on of Co2 in monetar y ter ms ar e taken as 0.3 kg/kWh to 0.6 kg/kWh with a cos t r anging fr om Eur o 10 to Eur o 33/ ton. T his gives a monetar y factor of 0.003 Eur o/ kWh to 0.02 Eur o/kWh. T he ener gy pr ice is taken as 0.04 Eur o/kWh. T hus Co2 cos t can be 15 % to 50 % of Ener gy cos t. T his factor however is not applicable for payback and it is thus not cons ider ed for payback in the tables pr es ented. T he payback is cons ider ed for ex tr a pr ice to be paid for the low los s tr ans for mer and it is ar ound 2 to 3 year s . T he Load Los s figur es given in the tables give the Load Los s es cons ider ing the har monics in the load. I n the fir s t cas e s tudy, the factor for enhanced eddy los s es in the fir s t load los s is s hown for illus tr ation only for illus tr ating r ough or der of values . All s tudies ar e pr es ented in the year 2002. 6.2 Case S tudy 1 T he cas e s tudy cons ider s a lar ge company in the I r on and S teel s ector . T he aver age loading is 400 MW out of which about 60 MW is thr ough H.T . utilization by H.T . Motor s . T he r emaining 340 MW is thr ough dis tr ibution tr ans for mer s . Load is cons tant dur ing 24 hour s a day, 7 days a week. T r ans for mer r atings var y fr om 800 kVA to 4800 kVA. T her e ar e about 400 T r ans for mer s . About 200 Nos . ar e of 1250 kVA, and about 100 Nos . of 1600 kVA while the r emaining 100 Number s ar e of differ ent r atings . Mos t of the tr ans for mer s ar e r eplaced between 1982 to 1990. Almos t all the tr ans for mer s ar e of Dr y T ype due to pr oblems faced in the ear lier oil cooled tr ans for mer s . T he company follows the total owners hip cos t (T OC) concept and has us ed A and B figures of EUR 2.27/W for no load los s es and EUR 1.63/W for load los s es . T he comparative figures are given for 1250 kVA trans formers . T able 5.1 input data 1250 kV tr ans for mer T r ans for mer load 65% (cons tant load, 24/24h) with 6 puls e har monics Economic lifetime 10 year s I nter es t r ate 7% Ener gy pr ice EUR 40/MWh Har monic s pectr um 1 3 5 7 9 11 13 15 17 19 21 23 25 % 100 0 29 11 0 6 5 0 3 3 0 2 2 A (no- load los s evaluation) EUR 2,46 /W B (load los s evaluation) EUR 1,04 /W 23 6.2.1 I llus t r at ive calculat ions : I nflation is not cons ider ed and hence the pr es ent wor th ex pr es s ion is s implified us ing a = zer o. n 1 + a 1− 1− 1 Pr es ent wor th = 1+ i = (1 + i )n = (1 + i )n − 1 i−a i (1 + i ) n i I nter es t Rate 7 % i.e. 0.07 per unit. Per iod is 10 year s Pw = (1 + 0.07 )10 − 1 = 7.0236 0.07(1.07 ) 10 EL = 0.04 EUR/kWh Pw × EL × 8760 A= = EUR2.46 / Watt 1000 B = A x P2 x T , P = 65 % , i.e. 0.65, T = 1 B = 2.46 x 0.65 x 0.65 = 1.039 = EUR 1.04/watt 6.2.2 F act or f or H ar monics ( I ) ×h h =n 2 Factor for eddy los s es = ∑ Ih 2 1 h =1 I f har monics ar e abs ent, this factor is one, T he tes ted load los s es have eddy los s es at fundamental. I f data fr om des ign is available for per centage of eddy los s at fundamental, it s hould be us ed in the calculation. I n the abs ence of s pecific data, 2 copper los s es due to I R can be taken as 90 % and 10% of the s pecified Load Los s es can be attr ibuted to eddy los s es at fundamental fr equency. T hus Load Los s es at fundamental fr equency = Load Los s es x [ p.u. loading] 2 x [ 0.9 + (0.1) x 1] T he Ex tr a addition is over and above eddy los s es due to fundamental fr equency and hence ex tr a har monic factor ( I1) × h ( I1) × h h=n 2 h=n 2 K h = ∑ Ih 2 −1 Or K h = ∑ Ih 2 h =1 h =3 For the given s ix puls e har monics , the fifth has 29% value of the fundamental. K 5 = (0.29 ) × 5 × 5 2 Hence = 2.1025 24 K h = (0 .29 ) × 25 + (0 .11 ) × 49 + (0 .06 ) × 121 + (0 .05 ) × 169 + (0 . 03 ) × 287 2 2 2 2 2 + (0 . 03 ) × 361 + (0 .02 ) × 529 + (0 .02 ) × 625 2 2 2 = 2.1025 + 0.5929 + 0.4356 + 0.4225 + 0.2601 + 0.3249 + 0.2116 + 0.25 = 4.6001 T otal eddy los s factor = 4.6001 + 1 = 5.6 6.2.3 P er cent age of E ddy L os s es in L oad L os s es : T he nex t s tep is to evaluate full load los s es with har monic loading for the given tr ans for mer and als o for the r elatively low los s tr ans for mer of s imilar r ating being cons ider ed for r eplacement. T his r equir es data on per centage of Eddy Los s es in conductor s in the total Load Los s es for the ex is ting tr ans for mer and the near es t low los s s ubs titute. For 1250 kVA r ating, the ex is ting and new low los s des ign have following data for the s ubdivis ion of eddy los s es , the figur es ar e infer r ed fr om the final load los s figur es given in the K EMA publication. Ex is ting 1250 kVA Low Los s 1250 kVA No Load 2400 W 2200 W Rated Load Los s 9500 W 8200 W As s umed % Copper Los s es 90.69% 90.69% As s umed % Eddy Los s es 9.31 % 9.31% 6.2.4 F ull load los s es f or H ar monic L oading: Ex is ting T r ans for mer : Full load load los s es on Har monic Load = Rated load load los s es on linear loads x [ p.u. Copper + K (p.u. Eddy los s )] = 9500 × (0.9069 + 5.6 × 0.093) = 9500 × 1.42826 = 13568.47 Or 13568 Watts For S ugges ted Low Los s es tr ans for mer Full load load los s es = 8200 x 1.42826 = 11711.73 or 11712 watts . I t can be noted that infer r ed dis tr ibution is ver y clos e to as s umed dis tr ibution of 90% and 10% . T his is not always tr ue as can be s een fr om tables given in the annex ur e. For 1600 kVA tr ans for mer , the dis tr ibution wor ks out to 88.68% copper los s es and 11.32% for eddy los s es . For s imilar har monic load factor of 5.6 the multiplier comes to 1.5207. T hus r ated full load los s (Linear ) of 10000 w yields a figur e of 10000 x 1.5207 = 15207 w. T he low los s s ubs titute has full load los s (linear ) = 9500 x 1.5207 = 14447 w T he actual figur e s tated is 14218 w. T hus a s lightly differ ent dis tr ibution is cons ider ed for the low los s s ubs titute. T he method thus illus tr ates the s teps to calculate full load los s (har monic 25 loads ) if the dis tr ibution is known. I f des ign data is not available, 90% and 10% s ubdivis ion can give a r eas onable value. I ncidentally it s hows that due to har monic loads the full load los s es have gone up by 42% in 1250 kVA, and 52% in 1600 kVA tr ans for mer . T he needed der ating would be (11.42) = 0.839 and (11.52) = 0.811 For 1250 kVA and 1600 kVA r es pectively for har monic loading. T he actual loading is only 65% and hence all alter natives cons ider ed ar e s afe fr om the view point of temper atur e r is e. 6.2.5 R elat ive economics f or low los s t r ans f or mer s ( All Dr y t ype) f or 1 2 5 0 kVA and 1 6 0 0 kVA t r ans f or mer s . T he data wor ked out for 1250 kVA and 1600 kVA ar e given in T able 5.2and T able 5.3. T able 5.2 1250 kVA tr ans for mer U nit Dr y Dr y D i f f er en ce t r an s f or m er t r an s f or m er , l ow l os s es T r ans for mer r ating kVA 1250 1250 Rated no- load los s W 2400 2200 - 200 Rated load los s W 13568 11712 - 1856 T ot al an n u al l os s es kW h / a 71241 -8 6 2 3 62618 CO2 em i s s i on @ 0 ,4 2 8 ,5 2 5 ,0 -3 ,5 kg/ kW h t on / a Pur chas e pr ice EUR 12250 13000 750 Pr es ent value no- load los s EUR 5907 5414 - 493 Pr es ent value load los s EUR 14108 12178 - 1930 Capi t al i s ed cos t s EUR 32265 30592 -1 6 7 3 P ay back ( year s ) 2 ,2 I n t er n al r at e of r et u r n 45% T able 5.4 1600 kVA tr ans for mer U nit Dr y D r y t r an s f or m er , D i f f er en ce t r an s f or m er l ow l os s es T r ans for mer r ating kVA 1600 1600 Rated no- load los s W 2800 2670 - 130 Rated load los s W 15207 14218 - 989 T ot al an n u al l os s es kW h / 80809 -4 7 9 7 a 76012 CO2 em i s s i on @ 0 ,4 32,3 30,4 -1 ,9 kg/ kW h t on / a Pur chas e pr ice EUR 14451 14990 539 Pr es ent value no- load EUR 6891 6571 - 320 los s Pr es ent value load los s EUR 15812 14784 - 1028 Capi t al i s ed cos t s EUR 37154 36345 -8 0 9 P ay B ack ( year s ) 2 ,8 I n t er n al r at e of r et u r n 34% 26 Com m en t s : T he figur es for 1250 kVA, ex is ting tr ans for mer ar e illus tr ated fir s t. Rated No Load Los s = 2400 w = 2.4 kW Rated load los s = 13568 W = 13.568 kW (full load) Annual los s es for 65% loading for 8760 hour s = 2.4 x 8760 + 13.568 x 0.65 x 0.65 x 8760 kWh = 21024 + 50216.5 = 71240.5 = 71241 kWh/annum. Car bon Diox ide emis s ion at 0.4 kg/kWh = 71241 x 0.4 = 28496 kg = 28.5 T ons /annum Pur chas e Pr ice is given as EUR 12250 (About Rs .673750) Pr es ent value of noload los s es 2.46 x 2400 = 5904 T aken as EUR 5907 Pr es ent value of Load Los s = 13568 x 1.04 = EUR 14110 T aken as EUR 14108 T otal Capitalis ed Cos t = EUR 32265 A s imilar figur e for low los s tr ans for mer is EUR 30592 T his figur e favour s the low los s type with an initial pur chas e pr ice of EUR 13000 which is EUR 750 of added inves tment. Payback for ex tr a inves tment of EUR 750: T he low los s tr ans for mer cons umes 62618 kWh/annim, s aving ther eby 8623 kWh/annum. T hus the annual s aving = EUR 0.04 x 8623 = EUR 345 750 S imple payback = = 2.17 or 2.2 year s . (For about 6.12 % Ex tr a I nves tment) 345 100 I nter nal Rate of Retur n = = 45 % about. 2.2 A similar calculation for 1600 kVA shows a saving of 4797 kWh and a payback of 2.8 years for an added investment of EUR 539 (about 3.73 % extra cost). IRR 34 %. 6.2.6 S ummar y: 1. Due to s omewhat higher load los s figur es us ed for T OC dur ing initial pur chas e, higher inves tments have been pr efer r ed. Hence it is not ver y attr active to r eplace ex is ting tr ans for mer s by s cr apping. 2. I f a tr ans for mer is to be r eplaced for any r eas on, the low los s s ubs titutes s how an attr active payback of 2.2 to 2.8 year s . T he total s aving potential for r eplacing ALL 400 tr ans for mer s is given below in T able _ _ _ _ _ _ _ . T able 5.5 (Page34) T he total s aving potential of 2939 Mwh/year is equivalent to EUR 117564/year and is 0.084% of the total cons umption of 3.5 x 10 6 Mwh/year . 27 6.3 Cas e S t u dy-2 : Non fer r ous metal sector I n a large company in the non ferrous metal s ector, the total loading is about 190 MW. But almos t 180 MW are cons umed through dedicated high voltage trans formers for electrolys is . T he s cope for dis tribution trans formers is limited is only to 10 MW. Out of it, the load variation is about 45% during 10 hours , 35% during 14 hours . T otal number of trans formers is 25, wherein a good number is at 1000 kVA. Excepting 3 new dry type ins talled in 1999, mos t of the trans formers are old(1965 to 1970). T he los s pattern is No load = 1900 Watts Load los s = 10250 Watts Calculations for1000 kVA old trans former with the loading pattern and 5 years of life wit 7% interes t rate gives the A factor = EUR 1.44/watt And B factor = EUR 0.24/Watt. Harmonics are not cons idered. S ince the loading is low, giving a very low B factor, direct replacement is not economically viable. T able 5.5 s ummaris es the data for dry trans formers and oil cooled trans formers for future replacement. T able 5.5 outcome 1000 kVA tr ans for mer U nit Dr y HD 538 Oi l C-C’ D i f f er en ce t r an s f or m er r an s f or m er T r ans for mer r ating kVA 1000 1000 Rated no- load los s W 2000 1100 - 900 Rated load los s W 8600 9500 900 T ot al an n u al l os s es kW h / a 30336 -6 5 4 3 23793 CO2 em i s s i on @ 0 ,4 12,1 9,5 -2 ,6 kg/ kW h t on / a Pur chas e pr ice EUR 10074 8007 - 2067 Pr es ent value no- load EUR 2873 1580 - 1293 los s Pr es ent value load los s EUR 2102 2322 220 Capi t al i s ed cos t s EUR 15049 11909 -3 1 4 0 P ay back ( year s ) N/ A I n t er n al r at e of r et u r n N/ A I n this cas e, the oil trans former has a lower firs t cos t and als o lower los s es . Hence it is the mos t favoured choice and the rate of return is not applicable; s ince the low los s trans former als o happens to have a lower firs t cos t. T able 5.6 s ummaris e the overall potential for the s aving. T his is equal to EUR 6560 and 0.0099% of the total electricity charges becaus e only a s mall fraction of the total load is qualifying for calculation of s avings . T able 5.6 Annual s avings potential T r an s f or m er T ot al E n er gy s avi n g CO2 em i s s i on s avi n g s iz e n u m ber [ MW h ] [ t on n es ] 1000 kVA 12 78,5 31,2 Other 13 85,1 33,8 T ot al 25 164 65 28 6.4 Case S tudy-3: P aper & P ulp Company A paper mill s tarted fuctioning s ince 1978 and was expanded in 1986, the peak electrical loading is about 110 MW, out of which 72 MW are us ed at high voltage for HT motors . T he remaining is dis tributed with 52 trans formers with ratings of 1000 kVA and 3150 kVA. T he dominant number( 28) are 3150 kVA trans formers with LV of 690 Volts . Average loading is 65% . T he highlight of the cas e s tudy is that in 1986, the company took s pecial care to s elect trans formers with low los s es for long term gains . T hes e trans formers are better compared to the low los s trans formers available today. T he cas e is pres ented for 3150 kVA trans former for which the input data is given in table 5.7. T able 5.7 : I nput data of 3150 kVA tr ans for mer Transformer size 3150 kVA oil-type T r an s f or m er l oad 65% dur ing 24/24 hour s with 6 puls e har monics E con om i c l i f et i m e 20 year s I n t er es t r at e 7% E n er gy pr i ce EUR 40/MWh 6 puls e accor ding to I EC 146- 1- 1 H ar m on i c s pect r u m A ( n o-l oad l os s EUR 3,71 /W eval u at i on ) B ( l oad l os s eval u at i on ) EUR 1,57 /W T he comparis on of the 1986 low los s trans former is made with the original s upply of 1978 bas ed on the likely prices as prevalent in 2002. T he res ults are s hown in table 5.8. I t is s een that even though 1986 trans former is about 30% more expens ive, it s till gives large s avings with an internal rate of return of 33 % and a payback period for extra inves tment of 3 years . T able 5.8: Outcome of 3150 kVA tr ans for mer U nit Oi l 1 9 7 8 Oi l 1 9 8 6 D i f f er en ce t r an s f or m er T r an s f or m er T r ans for mer r ating kVA 3150 3150 Rated no- load los s W 2870 3150 - 280 Rated load los s W 24500 16800 - 7700 T ot al an n u al l os s es kW h / 181908 -4 6 8 1 6 a 135092 CO2 em i s s i on @ 0 ,4 72,8 54,0 -1 8 ,8 kg/ kW h t on / a Pur chas e pr ice EUR 19329 24987 5658 Pr es ent value no- load los s EUR 10654 11693 1039 Pr es ent value load los s EUR 66432 45553 - 20879 Capi t al i s ed cos t s EUR 96415 82233 -1 4 1 8 2 P ay back ( year s ) 3 ,0 I n t er n al r at e of r et u r n 33% I t is es timated that the company is already s aving 46816 kWh/year due to thes e trans formers . 29 6.5 Cas e S t u dy-4 Ch em i cal I n du s t r y I n the KEMA s tudies , it is obs erved that; des pite variations in the proces s es , common trends are obs erved regarding electrical ins tallations . High reliability requirements lead to redundancy in trans former ins tallations and a low average loading of about 40% . Bas ed on the general obs ervations , a fictitious but repres entative cas e s tudy is prepared. Average loading is 110 MW, out of which 40 MW are for electrolys is or H.V. motors and thus out of the purview. Loading is continuous round the clock and loads are non-linear. A typical rating is 1250 KVA ( 60 out of 71 trans formers ). T he remaining trans formers are 630, 1000 and 1600 KVA. T he s tudy compares 1250 KVA HD538 trans former and 1250 KVA low los s trans former. Life time is cons idered 5 years and harmonics are not cons idered. I nteres t rate is taken as 7% . Energy price EUR 50/MWH. A= EUR 1.8/W and B= EUR 0.29/W (40% loading ). Highlights : For the chos en parameters , the differences are marginal. T he extra cos t of Low Los s type is EUR 750 over EUR 12250, and payback is 4.2 years with a rate of return of 6% . T his is a cas e where the chos en parameters of lifetime, harmonics etc. can s ignificantly affect the decis ion. I f the Low Los s type is chos en, the potential s avings can be 214.4 MWH/yr. Which can als o mean s avings in CO 2 emis s ion of 85.8 T on. 6.6 Case S tudy 5 Case of A L ar ge Data H otel S tar t U p T his is a high growth rate bus ines s with computers as a maj or load. T he s tartup load connection is typically 100 MW in the growth expectation of 200% to 300% ris e per year for a few years . T he economic life time is cons idered as only one year and interes t 7% . Figures as s umed are 25% initial 24 Hrs . loading which reaches 70% at the end of one year. Energy at EUR 60/MWH, high harmonic loading, A= EUR 0.52/W initial and als o the s ame value for no load los s es . B= EUR 0.03/W initial and 0.24/W at the end of the year. Highlights : T he s tudy s hows that due to s election of one year as economic life, the preference is clearly in favour of lowes t firs t cos t. I t is revealed that compared to 1600 KVA Dry type normal and 1600 KVA low los s Dry type, the cheapes t would be an oilcooled CC’ type trans former. T he capitalis ed cos ts with harmonics are EUR 16714, and 17132 (low los s ) res pectively initially. At the end of one year the figures are EUR 22311 and 22366. T hus the low los s trans former is s till not attractive. T here is a net s aving of 8222 KWH/year after one year which equals about EUR 411. T he extra price of EUR 539 can not be recovered in the economic life pres cribed. T he oilcooled trans former is a winner in the s hort run, with a capitalis ed cos t for initial period as EUR 12951 including harmonics . Even for this trans former, higher operating temperature due to harmonics s ugges ts a dras tic decreas e in operating life from 30 years to 6 years . Even then the s elected s hort economic life s pan makes this choice viable, provided the hot s pot temperature is acceptable. By the s ame cons ideration a s maller rating 1000 KVA trans former gives a capital s aving of 25% even though it has an energy penalty. I t is important to note that the payback period is not affected by the choice of economic life s pan, but the relatively longer payback los es its s ignificance due to s hort time inves tment perception. I n s uch a cas e, enforcing minimum los s norms only can help. Alternatively the inves tment in the trans former can be made by the utility with a long term perception to make energy s aving pos s ible. T he utility can s hift the trans former later to a s uitable load as needed. 30 6.7 S ummar y of E ur opean Case S tudies: T here is an interes ting s ummary of the s ens itivity of the payback period to input parameters . T able 5.9 gives a s ummary of effect of Low, Medium and High values of parameters on the payback period. Loading and electricity price are two mos t important factors . Loading s hould be carefully evaluated for a proper choice. T able 5.9 Par ameter s ens itivity on the payback per iod P ar am et er P ar am et er var i at i on P ayback t i m e ( year s ) U nit L M H L M H Har monic s pectr um None 12 puls e 6 puls e 3,3 3,1 2,7 Electr icity pr ice EUR/MWh 40 60 80 4,5 3,1 2,4 CO2 emis s ions kg/kWh 0,3 0,4 0,6 3,2 3,1 3,0 CO2 cos ts EUR/tonne 0 10 33 3,3 3,1 2,7 Loading pr ofile % 20 40 60 5,2 3,1 1,9 Economic lifetime year s 1 5 10 3,1 3,1 3,1 I nter es t % 5 7 9 3,1 3,1 3,1 Pur chas e pr ice % 80 100 120 2,5 3,1 3,7 6.8 Case S tudy: T ea I ndustr y ( I ndia) Ener gy Audit for T ea Factor ies making C.T .C. T ea, managed by H/S C.W.S (I ndia) Ltd., Dis tr ict Coimbator e. Audit was conducted in may 1990 for Mayur a and Par lai T ea factor ies . Power is r eceived at 22 kV and 11 kV by s epar ate lines . T his is s tepped down by two 500 kVA T r ans for mer o 22 kV/433 V an 11 kV/433 V which fee s egr egated loads . T he typical los s figur es for 500 kVA tr ans for mer s ar e 1660 W for no load and 6900 W as load los s es for 100% load. R ecom m en dat i on : Par allel both tr ans for mer s for a total 500 kVA load on s econdar y s ide and in lean s eas on and holidays when the load is 5% to below 25% , cut off one tr ans for mer on H.V. and H.V. s ides . B r i ef An al ys i s : For total load of 500 kVA, T her e ar e thr ee options . a) Only one tr ans for mer takes full 500 K VA LOAD. Los s es = 1.66 (No Load) + (500/500) 2 x 6.9 kW (load los s es ) b) One tr ans for mer takes s egr egated 300 kVA while s econd akes 200 kVA s egr egated load. Los s = 1.66 + (300/500) 2 x 6.9 + 1.66 + (200/500) 2 x 6.9 kW c) B oth ar e par alleled to take 250 kVA each. Los s = 2 (1.66 + (250/500) 2 x 6.9)kW = 6.77 kW. T hus on maj or load, the los s es ar e minimum by par alleling both tr ans for mer s . 31 Oper ation at higher loads dur ing leave s eas on : a) T wo par alleled tr ans for mer s Los s es = 2 [ 1. 66 + (0.25/2) 2 x 6.9} kW = 3.54 kW at 25% load Los s es = 2 { (1.66) + (0.05/2) 2 x 6.9} kW = 3.33 kW at 5% load b) Only one tr ans for mer is ener gized Los s es = 1.66 x (0.25) 2 x 6.9 = 2.09 kW at 25% load Los s es = 1.66 x (0.05) 2 x 6.9 = 1.68 kw at 5% load T hus los s es ar e minimum at low loads us ing only one tr ans for mer . T he tar iff was kVA of M.D. x Rs . 60 + Rs . 0.89 x kWh + Rs . 150 meter r ent. T he total annual cons umption for the factor y was 1856479 kWh per year and the electr icity bill was Rs . 2038694 giving Rs . 1.0094/kWh as aver age cos t. T he s aving by par alleling and s witching off one tr ans for mer wer e cons er vatively es timated at a minimum of 10000 kWh/year with no inves tment giving a little over Rs . 10000/year as a s aving. Power factor impr ovement was alr eady made but s ome s cope for fur ther impr ovement was s ugges ted. T his would r educe M.D. and s ave on M.D. char ges and als o give s avings on tr ans for mer and cable los s es . 32 APPENDIX-I: DATA REGARDING AVAILABLE DESIGNS A.1 Data S our ce : T his data is bas ed on the data given in the r efer ence viz ‘Ener gy S aving in I ndus tr ial Dis tr ibution T r ans for mer s ’ May 2002 by W.T .J. Huls thor s t and J.F. Gr oeman of K EMA – Nether lands . T he data is intended give a bas ic feel about the los s levels and dis tr ibution for dis tr ibution tr ans for mer s and their r elative cos ts /pr ices ; as per cur r ent Eur opean Pr actice. T he pr ices ar e for compar is ion only but a gener al conver s ion factor of 1 Eur o = Rs . 55 is cons ider ed whenever applicable. T he ener gy pr ice is s tated to be r oughly in the r ange of 40 Eur o/Mwh i.e. 0.04 Eur o/kWh. T able A1. Data For T r ans for mer s Us ed I n T he Utilities Rating KVA 100 400 1600 HV KV 20 10 20 LV V 400 400 690 Los s - HD428 A- A’ C- C’ A C A- A’ A- A’ C- C’ C- C’ A C A- A’ A- A’ C- C’ C- C’ A C Level AMDT AMDT AMDT AMDT AMDT AMDT No- Load W 320 210 60 80 930 930 810 810 150 180 2.600 2.600 1.700 1.700 380 420 Los s es Load W 1.75 1.47 1.750 1.475 4.600 4.600 3.85 3.85 4600 3.850 14.00 14.00 17.00 1700 17.00 14.00 Los s es 0 5 0 0 0 0 0 0 0 0 T otal Kg 520 650 740 770 1.190 1.200 1300 1400 1590 1750 3.300 3.240 3.370 9680 4.310 4550 Mas s Cor e Kg 150 220 220 225 435 440 450 540 570 600 1.100 1.210 1.200 1.460 1400 1.550 Mas s Flux T 183 1.45 1.35 1.35 1.83 1.84 1.85 1.6 1.35 1.35 1.84 1.84 1.7 1.6 1.35 1.35 Dens ity Conduct Cu/Al Cu Cu Cu Cu Cu Al Cu Al Cu Cu Cu Al Cu Al Cu Cu or Mater ial Winding Kg 85 115 130 155 203 145 350 220 360 450 505 295 725 465 1.120 1.225 Mas s Cur r ent A/mm 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 2.1 Dens ity 2 Height Mm 1300 1300 1300 1300 1330 1.420 1350 1550 1400 1400 1.890 1.820 1860 2000 1870 1900 Length Mm 890 830 1050 1100 1320 1.100 1010 1130 1340 1240 1.820 2000 1710 1850 17770 1770 Width Mm 600 560 620 620 800 840 800 780 770 800 1.180 1280 1100 1020 1320 1200 E fficienc % 97.9 98.3 98.19 98.46 98.62 98.82 98.8 98.8 98.81 99.00 98.78 98.78 99.02 99.02 98.91 99.10 y (* ) 4 2 9 9 S ound Db (A) 57 36 59 59 61 68 56 58 68 68 68 72 63 63 76 76 Power Unit E ur o 2539 2800 3458 3667 4385 4236 4881 4705 6373 6797 9692 9261 10307 1011 15050 15531 Cos t 9 Unit % 90.7 100 121.6 127.5 93.2 91.1 103. 100 135.5 144.5 95.8 91.4 101.9 100 143.7 153.5 Cos t 7 Data for Oil immer s ed tr ans for mer s , 100 K VA to 1600 K VA us ed in the utilities . AMDT r efer s to Amor phous Cor e Dr y T ype. 33 T able A- 2: Data for calculated los s es for I ndus tr y T r ans for mer s of 1000 to 4000 K VA. Typical Industry Transformer Parameters rating kVA 1000 1600 2500 4000 HV kV 10 10 10 10 LV V 420 420 420 420 Uk % 6 6 8 8 LOSS-LEVEL Oil CC' Oil DD' Dry base Dry Low Oil CC' Oil DD' Dry base Dry Low Oil CC' Oil DD' Dry base 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 % 100 129 126 139 100 118 133 138 100 131 131 140 100 118 102 110 (*) at full load and cos phi = 1 T able A- 3: Data for calculated es timation fr om des ign data for per centage of ex tr a eddy los s es in windings and s tr uctur al par ts . For Oil immes s ed and Dr y T ype T r ans for mer s (K EMA T able 4.6 Page 29- 30). 1 0 0 0 kVA 1 6 0 0 kVA 2 5 0 0 kVA 4 0 0 0 kVA In Other In Other In Other In Other winding PS E + winding PS E + winding PS E + winding PS E + PWE PCE PWE PCE PWE PCE PWE PCE Oi l CC’ H D 6% 5% 9% 13 % 11% 14 % 13% 28 % 428 Oi l D D ’ H D 6% 5% 9% 13 % 11% 14 % 13% 28 % 428 D r y t ype 6% – 9% – 11% – 13% – HD538 D r y t ype 6% – 9% – 11% – 13% – l ow l os s es T able A- 4: Dis tr ibution T r ans for mer s Los s S tandar ds Load Losses No-Load Losses Rated Oil-Filled (HD428) Dry Type Oil-Filled (HD428) Dry Type Power Up to 24 kV2 (HD538) Up to 24 kV2 (HD538) List A List B List C 12 kV Primary List A’ List B’ List C’ 12 kV Primary KVA W W W W W W W W 50 1100 1350 875 N/A 190 145 125 N/A 100 1750 2150 1475 2000 320 260 210 440 160 2350 3100 2000 2700 460 375 300 610 250 3250 4200 2750 3500 650 530 425 820 400 4600 6000 3850 4900 930 750 610 1150 630 /4% 6500 8400 5400 7300 1300 1030 860 1500 630 /5% 6750 8700 5600 7600 1200 940 800 1370 1000 10500 13000 9500 10000 1700 1400 1100 2000 1600 17000 20000 14000 14000 2600 2200 1700 2800 2500 26500 32000 22000 21000 3800 3200 2500 4300 4000 N/A N/A N/A N/A N/A N/A N/A N/A 1 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. 34 REFERENCES 1. Energy Saving in Industrial Distribution transformers- W.T.J. Hulsthorst & J.F. Groeman, KEMA, Netherlands 2. Transformers- BHEL, Tata Mc GrawHill (I) Ltd 3. Harmonics related documents from Underwriters Laboratory, USA 4. The Scope of Energy Saving in European Union Through Use of Energy Efficient Distribution Transformers – European Copper Institute, Belgium 35