You are on page 1of 28

Electrical Systems: Power Factor Correction

Resistive devices, like electric resistance heaters and incandescent lights transform all the power supplied to the device into heat or useful energy. Inductive devices, like motors, use some of the power supplied to the device to energize the inductive windings and create a magnetic field. This power, called reactive power, is alternately stored and given up by the windings, but is not used to do actual work. When this happens, the line supplying power to the device now carries the actual power used by the device and the reactive power created by the device. Actual power used by the device is measured in kW, reactive power created by induction devices is measured in k Ar, and the apparent power in the supply lines is measured in k A. The mathematical relationships between these types of power are described by the !power triangle" shown below. #or e\$ample, %s & '%a( ) %r(*+,(

%s & supplied power 'k A* %r & reactive power 'k Ar* %a & actual power 'kW*

The ratio of the actual power consumed by e-uipment '%a* to the power supplied to e-uipment '%s* is called the power factor. %# & %a , %s & kW , k A & cos .evices which generate,re-uire large amounts of reactive power in relation to actual power consumed have low power factors. /uch devices include0 1otors 2I. and fluorescent lights with low %# ballasts .evices which convert A3 power to .3 power such as0 .3 drives Welding machines #.s Induction furnaces

#ully loaded motors generally have a power factor of about 456. 2owever, if the motor is under loaded, the fraction of reactive power 'for the coil* to actual power 'for mechanical work* increases and the power factor decreases. Two potential problems are associated with low power factor. #irst many utilities have e\$plicit or implicit charges for low power factor. /econd, low power factor increases the current, and hence losses, in transformers and the electrical distribution system. These losses cost money and generate e\$cess heat in the electrical distribution system, which may shorten e-uipment lifetime or cause production shut downs. These potential problems are discussed in the sections that follow. %ower #actor 3harges 1any utilities charge for low power. To measure power factor, the most common type of utility meter measures the total k Ar7hours and k A7hours over the billing period and calculates the average power factor as0 %# & 3os 8 Arc/in 'k Arh , k Ah* 9 The most common methods of charging for low power factor are0 +. Adding a demand penalty when the power factor dips below a set amount 'usually :56* #or e\$ample0 the 3inergy ./ rate specifies a demand penalty of0
5.: 7 %# kWactual when %# ; 5.: %#

If actual power was +55 kW, the power factor was 456, and the avoided cost of demand were <+=.>? ,kW, the monthly power factor charge would be0
5.: 7 5.4 +55 kW =+(.= kW 5.4

+(.= kW \$ <+=.>? ,kW & <+:> (. @asing the demand charge on the supplied power %s 'k A*, rather than the actual power used %a 'kW*.

#or e\$ample, assume billing demand is based on k A rather than kW and the demand charge is <+=.55 ,k A7month. If actual demand was +55 kW, the power factor was 456, the implicit monthly power factor charge would be0 k A%# & 456 & kW , %# & +55 kW , 5.4 kW,k A & +(= k A k A%# & +556 & kW , %# & +55 kW , +.5 kW,k A & +55 k A %enalty & <+= ,k A7month \$ 'A A%# & 456 7 A A%# & +556* %enalty & <+= ,k A7month \$ '+(= k A B +55 k a* & <C?= ,month C. @asing part of the overall charge on the reactive power k Ar, which increases as power factor decreases. #or e\$ample, the .ayton %ower and Dight Eeneral /ervice %rimary Rate specifies a charge of <5.C5 per k Ar. The relationship between reactive and actual power is0 %r 'k Ar* & %a 'kW* \$ Tan83os7+'%#*9 If the actual power was +55 kW and the power factor was 456, then the power factor charge would be0 %r 'k Ar* & +55 'kW* \$ Tan83os7+'5.4*9 & ?= k Ar ?= k Ar \$ <5.C5 ,k Ar & <(C %ower Dosses and F\$cess 2eat Eeneration In addition to possible power factor charges, low power factor also results in e\$cess current in the electrical distribution system upstream from the device. The e\$cess line current results in increased resistive losses, and hence heat gain, in the wiring and electrical distribution e-uipment. The -uantity of line losses associated with low power factor correction can be calculated as follows0 DD+ & Dine loss before power factor correction DD( & Dine loss after power factor correction 6 Dine Doss /avings & 'DD+ B DD(* , DD+ DD+ & I(+R+ & 'k A+, Thus0 DD( & 8kW( R ,
( +

*( R+ &8'kW+,%#+* ,

9( R+ & 8kW( R ,

9+ , %#+(

9( , %#((

Assuming everything remains constant e\$cept for the power factor0 8kW( R ,
(

9+ & 8kW( R ,

9( & 8kW( R ,

And, 6 Dine Doss /avings & 'DD+ B DD(* , DD+ 6 Dine Doss /avings & 8'kW( R , (* , %#+( B 'kW( R , 6 Dine Doss /avings & 8+ , %#+( B + , %#((9 , + , %#+( 6 Dine Doss /avings & + B '%#+, %#(*(
(

* , %#((9 , 'kW( R ,

*+ , %#+(

#or e\$ample, if the power factor were improved from 456 to :56, the percent line loss savings would be0 6 Dine Doss /avings & + B '%#+, %#(*( & + B '456, :56*( & (+6 In addition, the heat generation in upstream electrical distribution e-uipment would be reduced by (+6. This may or may not be significant. If the electrical circuits are fully loaded and tripping due to e\$cess current, then power factor correction could mitigate this problem. Although percent line loss savings are relatively high, total energy savings are typically small since line losses are small. #or e\$ample, if line losses are (6 of the total power draw, the total power savings from correcting the power factor would be0 (6 \$ (+6 & 5.G(6 /ome manufactures of power factor correction e-uipment claim that actual losses are much greater than those calculated here, but there is little documented evidence of this in the open literature. @ecause of these effects, it is generally in the clientHs interest to maintain a relatively high power factor. To maintain a high power factor0 purchase e-uipment with high power factor ratings, such as high power factor lighting ballasts avoid or replace dramatically oversized motors, since under7loaded motors have low power factors.

If power factor is still a problem, consider adding electrical capacitors. /izing 3apacitors and Fstimating /avings 3apacitors work by canceling reactive power and current on the primary or upstream side of the capacitor. #or e\$ample, if a motor operates at ?56 power factor, installing a capacitor in the power supply line to the motor would reduce reactive power and line current on the primary side of the capacitor, but would not change the line current on the secondary 'motor* side of the capacitor. Thus, installing capacitors directly upstream from low7power7factor loads reduces line current throughout the plantIs electrical distribution systemJ whereas installing capacitors directly down stream of the utility

meter at the electrical service entrance to the plant, results in power factor correction for utility billing purposes, but will not reduce line losses and overheating throughout the plant.

k Aprimary ; k Asecondary %#primary K %#secondary

The two primary types of capacitors are oil7filled and gas7filled. Lil7filled capacitors typically last about >5,555 hours and may introduce voltage transients to the downstream e-uipment. In addition, the oil is potentially flammable and may contain to\$ins re-uiring special disposal. Mitrogen and helium gas7filled capacitors last about +(5,555 hours, filter voltage transients, are non7flammable and non7to\$ic. 2owever, the substantially lower cost of oil7filled capacitors make them much more popular than gas7filled capacitors. 3apacitors are sized by the amount of reactive power 'k Ar* they can cancel. /imple capacitors are sized to compensate for a fi\$ed amount of power. !/tepped" capacitors have internal controls that adNust the amount of reactive power compensation. Adding too much capacitance can push the system from !lagging" to !leading"J for e\$ample, adding too much capacitance may change the power factor from :=6 to +5=6. Although leading power factor does not harm e-uipment, purchasing e\$cess capacitors is e\$pensive and serves no useful purpose. In addition some meters may read a leading +5=6 power factor as :=6. If so, you would not get credit for the power factor correction from :=6 to +556. 'Lther utility meters would read a power factor of +5=6 as +556*. @ecause of these reasons, we recommend a conservative approach to power factor correction in which we never overcorrect the power factor past +556. A simple method to size the amount of capacitor k Ar re-uired is described in the steps that follow0 +. #ind k Ar for each month0 %r 'k Ar* & %a 'kW* \$ Tan83os7+'%#*9 (. To increase %# as close to +.5 as possible, recommend additional capacitance e-ual to minimum monthly k Ar during the past +( months. This approach minimizes the possibility of adding too much capacitance.

C. /ubtract the recommended capacitance 'k Ar* from recorded 'k Ar* for each month. This difference represents the reactive power 'k Ar* if the recommended capacitance were added. G. Recalculate %#, k A or k Ar and electricity costs for each month, using the reactive power calculated in the previous step. These costs represent the costs if the recommended capacitance had been added. =. 3alculate savings as the difference between the actual costs and the costs calculated in the previous step. >. To estimate the implentation cost, we note that the installed cost of capacitors is about about <(5 ,k Ar 7 <=5 ,k Ar, depending on control comple\$ity and size. F\$amples follow0

AR x: Consider Installing Capacitors to Improve Power Factor

Annual /avings %roNect 3ost 3L( Resource .ollars 3apital Lther Total 'tonne* Flectrical #ees Mone <+(,C5= <G>,555 <G>,555 <:(,555 AR30 (.C(+(.C /imple %ayback :5 months

Analysis Flectrical e-uipment that generates inductive loads, such as motors, creates current that is not in phase with the current supplied by the electric utility. Inductive loads cause the current waveform to lag behind the voltage waveform. This causes a portion of the energy to return to the source, hence leaving less usable power for the e-uipment. The power associated with this unusable current is called reactive power 'k Ar*. @ecause some energy is returned to the source, a higher supply power 'k A* must be generated by the utility in order to meet e-uipment needs. The ratio of power consumed by e-uipment 'kW* to total power in the electrical lines 'k A* is called the power factor. 1ost utilities charge for low power factor. Dow power factor can be corrected by adding capacitors, which are rated in terms of k Ar.

%ower #actor of +. All %ower is usable. %ower factor of 5.?+. Onusable power is created. '/ource0 www.wikipedia.com*

The diagram below shows the relationship between the various types of power0 supplied power 'k A*, reactive power 'k Ar*, and actual used power 'kW*. The -uantity of each type of power can be calculated using trigonometric relations defined by the power triangle.
k A 'power supplied by utility* kW %# & kVA

kW 'actual power used by machines*

%ower factor is ratio of the actual power, kW, and supplied power, k A. The power factor can be calculated using the following relationships0 %# & kW , k A & cos 'tan7+ 'k Ar,kW** '+*

Reactive power is a measure of the unused power in the lines. The reactive power can be calculated using the following relationship0 Reactive %ower 'k Ar* & kW \$ tan 'cos7+ '%#** '(*

The supplied power 'k A* can be calculated from the reactive power 'k Ar* and power factor from the following relationship0 k A & k Ar,sin 'cos7+'%#** 'C*

According to your rate structure !Darge Eeneral /ervice Rate", the power factor charge is about <5.G4 per k Ar per month. Pour utility company charges for all k Ar in e\$cess of +56 of the billed demand. According to the utility bills, the current power factor averages about 5.4=. /avings would be achieved if the power factor were increased from 5.4= to +.5. Recommendation We recommend considering adding an additional (,C55 k Ar of capacitance to improve the average power factor. Estimated Savings The data listed below is e\$tracted from the utility bills for the years (55> and (55?. The utility bills summary shows actual demand 'kW* and your current lagging reactive demand 'k Ar*. According to your rate structure, the power factor charge is about <5.G4 ,k Ar in e\$cess of +56 of billed demand. As indicated in the table below, installing (,C55 k Ar of capacitance would increase your average annual power factor from 5.4= 'kW,k A* to +.5 'kW,k A*. /avings are calculated as the product of power factor charge '<,k A* and the current billed reactive demand 'k A*.
Meter Reading Date 4,(?,5? :,(>,5? +5,(>,5? ++,(?,5? +(,(?,5? +,(=,54 (,(>,54 C,(?,54 G,(=,54 =,(?,54 >,(=,54 ?,(=,54 Tot,Avg &illed )nit Power Metered Actual Actual (agging Power Proposed Reactive Factor Consumption Demand Reactive Factor Capacitance Demand C#arge !"#\$period% !"% Demand !'Ar% !"\$!'A% !'Ar% !'Ar% *\$!'A% (,?45,4(= (,>(C,=CG (,>>:,?4: (,:=:,?(4 (,>+>,=G= (,>CG,:C+ (,:C>,>>( (,>=C,?C= (,=>>,:5+ (,=4G,(+? (,C++,+C= (,+G=,(:5 C+,G4C,(:( G,(+: G,(C4 G,+4= G,(G5 G,(C= G,(C= G,GGC G,(C= G,(C= G,54+ G,(=+ G,(+? G,(CG (,5>= (,5=? (,(G? (,++G (,+C4 (,+C4 (,+C4 (,+C4 (,+C4 (,54+ (,+C> (,(>( (,+C4 (,=C4 (,=C( (,?+: (,=:5 (,>+( (,>+( (,>(G (,>+( (,>+( (,=G5 (,>+G (,?C: (,>+( 5.4> 5.4> 5.4G 5.4= 5.4= 5.4= 5.4> 5.4= 5.4= 5.4= 5.4= 5.4G 5.4= <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 <5.G4 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 (,C55 +ew Reactive Demand !'Ar% (C4 (C( G+: (:5 C+( C+( C(G C+( C+( (G5 C+G GC: C+( +ew &illed +ew Power Reactive Factor Demand !"\$!'A% !'Ar% 5 5 + 5 5 5 5 5 5 5 5 +? ( +.55 +.55 +.55 +.55 +.55 +.55 +.55 +.55 +.55 +.55 +.55 5.:: +.55 Savings *\$Period% <::+ <:4? <+,5?4 <+,5+= <+,5(> <+,5(> <+,5(> <+,5(> <+,5(> <::: <+,5(= <+,5?4 <+(,C5=

The table shows that +,C55 k Ar of additional capacitance would be sufficient to avoid almost all power factor penalties, and the savings from installing additional capacitance would be about <+(,C5= per year. 2ence, we propose adding an additional (,C55 k Ar of capacitance. Estimated Implementation Cost @ased on previous e\$perience, simple capacitors can be installed for about <G5 per k Ar. Therefore, the estimated implementation cost would be about0 '(,C55 k Ar \$ <G5 ,k Ar* & <:(,555 We estimate that about =56 of this is for materials and =56 for labor. Estimated Simple Pay,ac! '<:(,555 , <+(,C5= ,year* \$ +( months & :5 months

AR x: Install Capacitors to Improve Power Factor

Annual /avings %roNect 3ost 3L( Resource .ollars 3apital Lther Total 'tonne* Flectrical #ees Mone <(>,=C? <+5,>(= <+5,>(= <(+,(=5 AR30 (.C(+(.C /imple %ayback +5 months

Analysis Flectrical e-uipment that generates inductive loads, such as motors, generates current that is not in phase with the current supplied by the electric utility. The power associated with this unusable current is called reactive power 'k Ar*. The total power in the line 'k A* includes both the useable 'kW* and the reactive power 'k Ar*. The ratio of power consumed by e-uipment 'kW* to total power in the electrical lines 'k A* is called the power factor. 1ost utilities charge for low power factor. Dow power factor can be corrected by adding capacitors, which are rated in terms of k Ar. The diagram below shows the relationship between the various types of power0 total power 'k A*, reactive power 'k Ar*, and actual used power 'kW*. The -uantity of each type of power can be calculated using trigonometric relations defined by the power triangle.
k A 'total power*

kW %# & kVA

kW 'actual power used by machines*

%ower factor is ratio of the actual power, kW, and total power, k A. The power factor can be calculated using the following relationships0 %# & kW , k A & cos 'tan7+ 'k Ar,kW** The reactive power can be calculated using the following relationship0 Reactive %ower 'k Ar* & kW \$ tan 'cos7+ '%#** '(* '+*

The total power 'k A* can be calculated from the reactive power 'k Ar* and power factor from the following relationship0 k A & k Ar,sin 'cos7+'%#** 'C*

In the !Otility Analysis" section of this report, both on7peak and off7peak power factors less than 5.: increase the billed demand. The cost for power factor below 5.: during on7 peak is0

& 'measured on7peak kW* , 'on7peak %#* \$ 5.: \$ 'on7peak demand cost* The cost for power factor below 5.: during off7peak is0 & Q8'ma\$ kW* , off7peak %# \$ 5.:9 7 8'measured on7peak kW* , 'on7peak %#* \$ 5.:9R \$ 'off7peak demand cost* Recommendation We recommend considering adding an additional G(= k Ar to improve the average power factor. Estimated Savings The table below shows on7peak and off7peak actual demand 'kW*, power factor 'kW,k A*, supplied power 'k A*, reactive power 'k Ar*, and billing demand 'kW*. The on7peak and off7peak power factor remains lower than 5.: for the entire year. Osing the rate structure form the !Otility Analysis" section of this report, the cost for this low power factor is about <(>,=C? per year.
Meter Reading Date :,++,5> +5,+5,5> ++,4,5> +(,++,5> +,+(,5? (,+(,5? C,+(,5? G,+(,5? =,++,5? >,+(,5? ?,+(,5? 4,+5,5? 0ot\$Avg Days C( C+ (: CC C( C+ (4 C+ (: C( C5 (: 123 -n.pea! Demand !"% 4>G 4>( 4G( 4C4 4+5 4G( 45G 4+G 4(C 4(? 45> 4G= 456 -//.Pea! Demand !"% 4>5 4>G 4G5 4C4 4+( 4(? 45> 4(? 4+? 4(? ?:C 4G= 417 -n.pea! Power Factor !"\$!'A% 5.?C 5.?C 5.?C 5.?C 5.?G 5.?> 5.?G 5.?> 5.?G 5.?= 5.?C 5.?= 7839 -//.Pea! Power Factor !"\$!'A% 5.?G 5.?C 5.?C 5.?G 5.?= 5.?G 5.?= 5.?= 5.?C 5.?G 5.?C 5.?G 7839 -n.pea! Supplied Power !'A% +,+4? +,+?G +,+=? +,+G= +,5:C +,++G +,5:C +,5?> +,+5> +,+5? +,5:: +,+(C :;:51 -//.pea! Supplied Power !'A% +,+=? +,+:( +,+=+ +,+C5 +,5?? +,++= +,5?> +,+5C +,++? +,++( +,54G +,+G> :;:55 -n.Pea! Reactive Power !'Ar% 4+G ?:? ?:C ?45 ?CG ?(: ?G+ ?5C ?C: ?C> ?G4 ?G5 366 -//.Pea! Reactive Power !'Ar% ??G 4(+ ?4? ?=? ?54 ?G? ?+C ?C5 ?>( ?GC ?G5 ??= 366 -n.Pea! &illing Demand !"% +,5>4 +,5=? +,5G+ +,5C5 :4G +,55C :4G :>4 ::> ::? :4: +,5++ :;7:: -//.Pea! &illing Demand !"% 5 +> 5 5 5 + 5 (= +5 G 5 (+ 2 Calculated Power Factor Cost *\$period% <(,>5C <(,C?+ <(,G5( <(,C(( <(,5:4 <+,:C? <(,+?? <+,44= <(,5:G <(,+>= <(,CG5 <(,+GC *52;613

The on7peak and off7peak reactive power 'k Ar* measured at the meter will be reduced by G(= k Ar per month by installing G(= k Ar of capacitance. This is depicted in the table below. G(= k Ar of capacitance will raise the on7peak and off7peak power factor above 5.:. Therefore, the annual power factor cost would be about <5 per year.
Meter Reading Date :,++,5> +5,+5,5> ++,4,5> +(,++,5> +,+(,5? (,+(,5? C,+(,5? G,+(,5? =,++,5? >,+(,5? ?,+(,5? 4,+5,5? 0ot\$Avg Days C( C+ (: CC C( C+ (4 C+ (: C( C5 (: 123 -n.pea! Demand !"% 4>G 4>( 4G( 4C4 4+5 4G( 45G 4+G 4(C 4(? 45> 4G= 456 -//.Pea! Demand !"% 4>5 4>G 4G5 4C4 4+( 4(? 45> 4(? 4+? 4(? ?:C 4G= 417 -n.pea! Power Factor !"\$!'A% 5.:+ 5.:( 5.:( 5.:( 5.:C 5.:G 5.:C 5.:= 5.:C 5.:G 5.:C 5.:G 78<1 -//.Pea! Power Factor !"\$!'A% 5.:C 5.:+ 5.:( 5.:C 5.:G 5.:C 5.:G 5.:G 5.:( 5.:C 5.:C 5.:( 78<1 -n.pea! Supplied Power !'A% :G? :C: :+: :+5 4>? 4:> 4>G 4>5 44+ 44G 4>4 :5( 4<6 -//.pea! Supplied Power !'A% :(4 :=5 :+= :5( 4>5 444 4=> 44( 44C 44> 4=C :+G 4<1 -n.Pea! Reactive Power !'Ar% C4: C?( C>4 C== C5: C5G C+> (?4 C+G C++ C(C C+= 117 -//.Pea! Reactive Power !'Ar% CG: C:> C>( CC( (4C C(( (44 C5= CC? C+4 C+= C=5 117 -n.Pea! &illing Demand !"% 4>G 4>( 4G( 4C4 4+5 4G( 45G 4+G 4(C 4(? 45> 4G= 41: -//.Pea! &illing Demand !"% 5 5 5 5 5 5 5 5 5 5 5 5 7 Calculated Power Factor Cost *\$period% <5 <5 <5 <5 <5 <5 <5 <5 <5 <5 <5 <5 *7

The annual savings for installing G(= k Ar of capacitance would be about0 <(>,=C? ,yr 7 <5 ,yr & <(>,=C? ,yr Estimated Implementation Cost @ased on previous e\$perience, simple capacitors can be installed for about <=5 per k Ar. Therefore, the estimated implementation cost for installing G(= k Ar of capacitance would be about0 G(= k Ar \$ <=5 ,k Ar & <(+,(=5 We estimate that about =56 of this is for materials and =56 for labor. Estimated Simple Pay,ac! '<(+,(=5 , <(>,=C? ,yr* \$ +( months & +5 months

AR x: Install Capacitors to Improve Power Factor

Annual /avings %roNect 3ost 3L( Resource .ollars 3apital Lther Total 'tonne* Flectrical #ees Mone <GG,C>( <>,?=5 <>,?=5 <+C,=55 AR30 (.C(+(.C /imple %ayback G months

Analysis Inductive loads, such as motors, generate a reactive current that is not in phase with the current supplied by the utility, hence, it is unusable. The power associated with this reactive current is called reactive power, and is measured in k Ar. The useful power drawn by machines and supplied by the utility is measured in kW. The total power in the lines is the sum of the useful and reactive power and is measured in k A. The ratio of power consumed by e-uipment 'kW* to power in the lines 'k A* is called the power factor. 1ost utilities charge for low power factor. Dow power factor can be corrected by adding capacitors, which are rated in terms of k Ar. The diagram below shows the relationship between the various types of power. The -uantity of each type of power can be calculated using trigonometric relations defined by the power triangle.
k A 'line power*

kW %# & kVA

k Ar 'reactive power*

kW 'actual power used by machines*

The power factor can be calculated using the following relationships0 %# & kW , k A & cos 'tan7+ 'k Ar,kW** '+*

Reactive power is a measure of the unused power in the lines. The reactive power can be calculated using the following relationship0 Reactive %ower 'k Ar* & kW \$ tan 'cos7+ '%#** '(*

The line power 'k A* can be calculated from the reactive power 'k Ar* and power factor from the following relationship0 k A & k Ar,sin 'cos7+'%#** 'C*

In the utility analysis section of this report, the power factor charge is k A \$ 5.:. The utility rate for meter + and C, !Eeneral Rate ./5+ /ervice at /econdary .istribution

oltage=; the monthly power factor charge for each meter with a power factor less than 5.: and a billed demand less than +,555 kW is0 @illed kW,mo \$ C55 kWh,kW \$ '<5.5CC5?> 7 <5.5(55:G* kWh ) @illed kW \$ <+=.+G=G ,kW The monthly power factor charge for each meter with a power factor less than 5.: and a billed demand greater than +,555 kW is0 @illed kW,mo \$ C55 kWh,kW \$ '<5.5CC5?> 7 <5.5(55:G* kWh ) @illed kW \$ <+=.+G=G ,kW According to the utility bills, the current power factors for meter + S C are less than 5.: for every month. /avings would be achieved if the power factor were increased to 5.:5. Recommendation We recommend considering adding an additional (=5 k Ar to 1eter + and an additional (55 k Ar of capacitance to 1eter C to improve the average power factor for each meter. Estimated Savings The table below shows meter + actual demand 'kW* current reactive demand 'k Ar*, current power factor and the total calculated electrical costs based upon the current rate structure for the previous seven billing periods. The calculated electrical cost for meter + is about <(:5,54: for the last seven billing periods.
1eter + Meter Reading Date +5,=,(55? ++,=,(55? +(,>,(55? +,:,(554 (,?,(554 C,?,(554 G,4,(554 Days (: C+ C+ CG (: (: C( Metered Consumption !"#\$period% (G5,5>G =54,(C> =:>,5:> >?4,>55 ?(G,:GG G+4,?(= G4(,4(C Avg8 Daily Consumption !"#\$day% 4,(?4 +>,C:= +:,((: +:,:=: (G,::4 +G,GC: +=,544 Actual Demand !"% 45C :?> +,5>( +,(=> +,C+? 4=C :?: Supplied Power !'A% +,55= +,+:( +,CC> +,=G: +,=4> +,+55 +,(GC Current Reactive Demand !'Ar% >5G >4= 4+5 :5? 44C >:G ?>= Power Factor !"\$>'A% 5.45 5.4( 5.45 5.4+ 5.4C 5.?4 5.?: Power Factor Penality !"% +5( :? +G5 +C4 ++5 +C? +C: &illed Demand !"% :5= +,5?C +,(5( +,C:G +,G(? ::5 +,++4 Calculated 0otal Cost *\$period% <(>,G5( <C:,>:G <G=,+(C <=+,+>> <=C,C>: <CG,4G5 <C:,G:=

0ot\$Avg

5:6

1;29<;944

:2;<39

:;716

:;543

329

7847

421874

:;:64

*5<7;74<

The table below shows meter + actual demand 'kW* new reactive demand 'k Ar*, new power factor and the total calculated electrical costs based upon the current rate structure for the previous seven billing periods. The calculated electrical cost for meter + by adding (=5 k Ar of reactive power would be about <(?>,5>C for the last seven billing periods.

1eter + Meter Reading Date +5,=,(55? ++,=,(55? +(,>,(55? +,:,(554 (,?,(554 C,?,(554 G,4,(554 Days (: C+ C+ CG (: (: C( Metered Consumption !"#\$period% (G5,5>G =54,(C> =:>,5:> >?4,>55 ?(G,:GG G+4,?(= G4(,4(C Avg8 Daily Consumption !"#\$day% 4,(?4 +>,C:= +:,((: +:,:=: (G,::4 +G,GC: +=,544 Actual Demand !"% 45C :?> +5>( +(=> +C+? 4=C :?: +ew Supplied Power !'A% 4?4 +54G ++45 +C:= +G>C :G? +54? +ew Reactive Demand !'Ar% C=G GC= =>5 >=? >CC GGG =+= Power Factor !"\$>'A% 5.:+ 5.:5 5.:5 5.:5 5.:5 5.:5 5.:5 Power Factor Penality !"% 5.55 5.55 5.55 5.55 5.55 5.55 5.55 &illed Demand !"% 45C :?> +,5>( +,(=> +,C+? 4=C :?: Calculated 0otal Cost *\$period% <(G,G>: <C?,::5 <G(,4>C <G:,C+( <=(,++> <C(,(C+ <C?,54(

0ot\$Avg

5:6

1;29<;944

:2;<39

:;716

:;:94

6:9

78<7

7877

:;716

*532;721

The table below shows meter C actual demand 'kW* current reactive demand 'k Ar*, current power factor and the total calculated electrical costs based upon the current rate structure for the previous seven billing periods. The calculated electrical cost for meter C is about <:G,5?4 for the last seven billing periods.
1eter C Meter Reading Date +5,=,(55? ++,=,(55? +(,>,(55? +,:,(554 (,?,(554 C,?,(554 G,4,(554 Days (: C+ C+ CG (: (: C( Metered Consumption !"#\$period% ?C,4?+ +(4,G:( +=:,(5= +4C,5:? +??,(>= +(G,:(? +G?,C(> Avg8 Daily Consumption !"#\$day% (,=G? G,+G= =,+C> =,C4= >,++C G,C54 G,>5G Actual Demand !"% (C( (4( C>( C>4 C4? C5: C>4 Supplied Power !'A% (:>.C G+?.4 G:G.? =G? =G4 GC? =(? Current Reactive Demand !'Ar% +4G C5: CC4 G5= C4? C5: C?4 Power Factor !"\$>'A% 5.?4 5.>? 5.?C 5.>? 5.?+ 5.?+ 5.?5 Power Factor Penality !"% CG.>? :G.G( 4C.>5 +(G.C: +5=.>G 4G.(C +5>.=? &illed Demand !"% (>? C?> GG= G:( G:C C:C G?= Calculated 0otal Cost *\$period% ?,:+: +(,54G +G,=?C +>,C4( +>,+>: +(,(?( +G,>?:

0ot\$Avg

5:6

<<9;:41

9;259

117

923

117

783:

211865

957

*<9;734

The table below shows meter C actual demand 'kW* new reactive demand 'k Ar*, new power factor and the total calculated electrical costs based upon the current rate structure for the previous seven billing periods. The calculated electrical cost for meter C by adding (55 k Ar of reactive power would be about <4(,((> for the last seven billing periods.
Meter Reading Date +5,=,(55? ++,=,(55? +(,>,(55? +,:,(554 (,?,(554 C,?,(554 G,4,(554 Days (: C+ C+ CG (: (: C( Metered Consumption !"#\$period% ?C,4?+ +(4,G:( +=:,(5= +4C,5:? +??,(>= +(G,:(? +G?,C(> Avg8 Daily Consumption !"#\$day% (,=G? G,+G= =,+C> =,C4= >,++C G,C54 G,>5G Actual Demand !"% (C( (4( C>( C>4 C4? C5: C>4 Supplied Power !'A% (C(.5 C5+.4 C4>.: G(+.( GC5.+ C(?.G G54.> +ew Reactive Demand !'Ar% 5 +5: +C4 (5= +4? +5: +?4 Power Factor !"\$>'A% +.55 5.:C 5.:C 5.4? 5.:5 5.:G 5.:5 Power Factor Penality !"% 5.55 5.55 5.55 ++.5= 5.55 5.55 5.55 &illed Demand !"% (C( (4( C>( C?: C4? C5: C>4 Calculated 0otal Cost *\$period% <?,(=: <+5,(4> <+(,:4+ <+G,((G <+G,+=? <+5,>>4 <+(,>=5

0ot\$Avg

5:6

<<9;:41

9;259

117

164

:15

78<1

::876

11:

*45;552

The annual savings for installing (=5 k Ar of capacitance to meter + and (55 k Ar of capacitance to meter C would be about0 '<(:5,54: 7 <(?>,5>C* , ? mo ) '<:G,5?4 7 <4(,((>* , ? mo & <(=,4?4 , ? mo <(=,4?4 , ? mo \$ +( mo,yr & <GG,C>( ,yr Estimated Implementation Cost @ased on previous e\$perience, simple capacitors can be installed for about <C5 per k Ar. Therefore, the estimated implementation cost for installing G55 k Ar of capacitance would be about0 'G=5 k Ar \$ <C5 ,k Ar* & <+C,=55 We estimate that about =56 of this is for materials and =56 for labor. Estimated Simple Pay,ac! '<+C,=55, <G+,=:=,year* \$ +( months & G months

AR I>C(0 TAAF 3LRRF3TI F A3TILM/ TL I1%RL F %LWFR #A3TLR Annual /avings %roNect /imple 3ost %ayback Resource 3L( 'lb* .ollars Flectrical .emand GC? k A <:G,4+( <4,555 + month Analysis and Recommendation 1otors and other inductance devices re-uire more power in the lines than they actually consume. %ower factor is ratio of power used 'kW* to the power that must be supplied in the lines 'k A*. The relationship between actual, supplied and reactive power is shown in the figure below. k A 'power supplied by utility* %# & kW kVA

k Ar 'measure of unused power*

kW 'power used by machines* According to the your electricity bills, the average plant power factor is about >46. @ased on your electric rate structure, you can lower your demand charges by improving power factor. 2owever, you do not want to over correct the power factor to more than +556. In 1arch (555, power factor peaked at ?G6. Thus, we recommend adding only enough capacitance to raise your power factor to :?6 for that month. According to the analysis below, this would be about G55 k Ar of capacitance. Fstimated /avings .etermining 2ow 1uch 3apacitance to Add The highest power factor registered during the period for which we were supplied with electricity data was for 1arch (555, when the plant power factor was ?G6 and the peak power was =:G kW. If the power factor during this month were corrected to :?6, the power factor during the rest of the year would never e\$ceed :?6. Osing standard trigonometric relations, the angle theta in the power triangle shown above was about0 3os & kW , k A & 3os7+ '%#* & 3os7+ '?G6* & G(.(? o The reactive power was about0 Tan & k Ar , kW k Ar & kW \$ tan

k Ar & =:G kW \$ tan 'G(.(?* & =G5 k Ar If the power factor were increased to :?6, the reactive power would be about0 3os & kW , k A & 3os7+ '%#* & 3os7+ ':?6* & +G.5? o k Ar & kW \$ tan k Ar & =:G kW \$ tan '+G.5?* & +G: k Ar Thus, the amount of capacitance re-uired to boost your power factor from ?G6 to :?6 during 1arch (555 would be about0 =G5 k Ar B +G: k Ar & C:+ k Ar We recommend adding about G55 k Ar. 3alculating /avings According to your electricity bills, the annual average power factor is >46 and the annual average demand is =?: kW. Osing standard trigonometric relations, the angle theta in the power triangle shown above is about0 3os & kW , k A & 3os7+ '%#* & 3os7+ '>46* & G?.+=> o The reactive power is about0 Tan & k Ar , kW k Ar & kW \$ tan k Ar & =?: kW \$ tan 'G?.+=>* & >(G k Ar If G55 k Ar of capacitance were added, the average k Ar would be about0 >(G k Ar B G55 k Ar & ((G k Ar The angle theta would be about0 & Tan7+ 'k Ar,kW* & Tan7+ '((G,=?:* & (+.+=5 o The annual average power factor would be about0 %# & 3os & 3os '(+.+=5* & :C6

The annual average k A would be about0 k A & kW , 3os & =?: , 3os'(+.+=5 o* & >(+ k A The billing k A would be about0 >(+ k A \$ 8+ ) '5.: 7 :C6*9 & >5( k A According to your electricity bills, the annual average power factor is >46 and the annual average k A is 4=( k A. Thus, without power factor correction the annual average billing k A would be about0 Actual k A \$ 8+ ) '5.: B %#*9 & 4=( kW \$ 8+ ) '5.: B >46*9 & +,5C: k A Thus, the savings from correcting the average power factor to :C6 would be about0 +,5C: k A B >5( k A & GC? k A GC? k A \$ <+4.54 ,k A7month \$ +( months,yr & <:G,4+( ,yr Fstimated Implementation 3ost According to -uotes from other utilities, the total installed cost for simple capacitors is no more than <(5 per k Ar. If so, the total implementation cost for installing G55 k Ar of capacitors would be about0 G55 k Ar \$ <(5 ,k Ar & <4,555 Fstimated /imple %ayback <4,555 , <:G,4+( ,yr \$ +( months,yr & + month

AR F5>+=5+0 TAAF 3LRRF3TI F A3TILM/ TL I1%RL F %LWFR #A3TLR Annual /avings %roNect /imple 3ost %ayback Resource 3L( 'lb* .ollars Flectric .emand G( k A <C,>>: <(,455 : months Analysis and Recommendation 1otors and other inductance devices re-uire more power in the lines than they actually consume. %ower factor is ratio of power used 'kW* to the power that must be supplied in the lines 'k A*. The relationship between actual, supplied and reactive power is shown in the figure below.

k Ar 'measure of unused power*

kW 'power used by machines* According to the your electricity bills, the average plant power factor is about :(.G6. @ased on your electric rate structure, you can lower your demand charges by improving power factor. 2owever, you do not want to over correct the power factor to more than +556. In #ebruary (555, power factor peaked at :=6. Thus, we recommend adding only enough capacitance to raise your power factor to ::6 for that month. According to the analysis below, this would be about +G5 k Ar of capacitance. Fstimated /avings .uring #ebruary (555, the plant power factor was :=6 and the peak power was ?>= kW. Osing standard trigonometric relations, the angle theta in the power triangle shown above was about0 3os & kW , k A & 3os7+ '%#* & 3os7+ ':=6* & +4.+:= o The reactive power was about0 Tan & k Ar , kW k Ar & kW \$ tan k Ar & ?>= kW \$ tan '+4.+:=* & (=+ k Ar If the power factor were increased to ::6, the reactive power would be about0 3os & kW , k A & 3os7+ '%#* & 3os7+ '::6* & 4.++ o k Ar & kW \$ tan k Ar & ?>= \$ tan '4.++* & +5: k Ar Thus, the amount of capacitance re-uired to boost your power factor from :=6 to ::6 during #ebruary (555 would be about0 (=+ k Ar B +5: k Ar & +G( k Ar

We recommend adding about +G5 k Ar. According to the your electricity bills, the average plant power factor is about :(.G6. The average power supplied is about 4++ k A and the average power consumed is about ?=5 kW. Osing these numbers, the average reactive power during the year is about0 3os & kW , k A & 3os7+ '%#* & 3os7+ ':(.G6* & ((.G4( o Tan & A Ar , kW k Ar & kW \$ tan k Ar & ?=5 \$ tan '((.G4(* & C+5 k Ar If +G5 k Ar of capacitance were added, the average reactive power would be about0 k Ar( & C+5 k Ar B +G5 k Ar & +?5 k Ar The average power supplied would be about0 ( & Tan7+'k Ar , kW* & Tan7+'+?5 , ?=5* & +(.??+ o k A & kW , 3os( & ?=5 kW , 3os'+(.??+* & ?>: k A The average reduction in k A would be about0 4++ k A B ?>: k A & G( k A Thus, the power factor savings would be about0 G( k A \$ <?.(4 ,k ABmo \$ +( mo,yr & <C,>>: ,yr Fstimated Implementation 3ost According to -uotes from other utilities, the total installed cost for capacitors about <(5 ,k Ar. If so, the total implementation cost for installing +G5 k Ar of capacitors would be about0 +G5 k Ar \$ <(5 ,k Ar & <(,455 Fstimated /imple %ayback <(,455 , <C,>>: ,yr \$ +( months,yr & : months AR F5>(:5+0 IM/TADD =55 A AR L# 3A%A3ITAM3F TL I1%RL F %LWFR #A3TLR Annual /avings %roNect /imple

%ower factor charge

Resource =55 k Ar

3L( 'lb*

.ollars <>,555

3ost <?,=55

%ayback += months

Analysis %ower factor is the ratio of actual power consumed 'kW* to the line power supplied 'k A*. The utility recorded an annual average power factor of 4C6 for the electricity service to the plant. 1ost plants strive to achieve a power factor of :=6 or more to reduce demand costs, reduce line losses and allow electrical e-uipment to run cooler.

k Ar 'measure of unused power*

Recommendation We recommend installing at least =55 k Ar of capacitance to raise the plant power factor and save on electricity charges. Fstimated /avings 1onthly demand, reactive power and power factor from the last year of billing data is shown in the table below.

AW +,C(( +,(=4 +,(GC +,+44 +,(:4 +,+4+ +,54G +,+?= +,(C+ +,+=G +,5>G +,+>C Tot,Avg,1a\$

A Ar =(( >>( :C( :4G +,5=4 44> ?GC ??G ?>C 45= ?5+ ?>? :,=:4

%# :C6 4:6 456 ??6 ?46 456 4C6 4G6 4=6 4(6 4G6 4G6 4(6

The monthly average reactive power varies from a low of =(( k Ar to a high of +,5=4 k Ar. Thus, it would be highly unlikely that you would ever over7correct your power factor by installing =55 k Ar of capacitance. If you did, the annual savings would be about0

=55 k Ar \$ <+ ,k AR7mo \$ +( mo,yr & <>,555 Fstimated Implementation 3ost The contract price for installation of simple capacitors varies from <+5 to <(5 per k Ar. Assuming <+= ,k Ar, the implementation cost would be about0 =55 k Ar \$ <+= ,k Ar & <?,=55 Fstimated /imple %ayback <?,=55 , <>,555 ,yr \$ +( months,yr & += months

AR T0 IM/TADD 3A%A3ITLR/ TL I1%RL F %LWFR #A3TLR %resent (?,5G4 k Ar,yr Recommended C,5G4 k Ar,yr Annual /avings (G,555 k ArJ <?,(55 <>5,555 4.C years

%# charge Implementation 3ost /imple %ayback

Analysis 1any types of electrical e-uipment re-uire that more power be supplied to the e-uipment than is actually consumed by the e-uipment. The ratio of power consumed by e-uipment to power supplied to e-uipment is called the power factor. Pour utility charges you for both how much power your plant actually uses 'measured in kW* and for the e\$cess power that the utility was re-uired to supply 'a measure of which is k Ar*. Thus, low power factor results in higher utility costs. In addition, low power results in overheating in wires, circuit boards and motors. This overheating may cause circuit breakers to shut off the load and interrupt production, and always reduces the lifetime and increases maintenance costs for electrical e-uipment. It is therefore in your interest to maintain a power factor as close to +556 as is economically feasible. k A 'power supplied by utility* kW %# & kVA kW 'power used by machines* The utility recorded an annual average power factor of ?:6 for the electricity service to the main plant. 1ost industrial facilities have a power factor of :56 or above. A ?:6 power factor corresponds to an average of (,(=G k Ar per month, for which the utility charges C5 cents per k Ar. Thus, your average power factor charge is about <>?> per month or about <4,++G per year. Installing (,555 k Ar of capacitors would raise the average monthly power factor to about ::6 without ever switching from lagging to leading. The annual utility savings would be about <?,(55. Additional savings from longer e-uipment lifetimes, lower maintenance costs, and avoided production stoppages are difficult to -uantify, but are potentially even greater than the utility savings. We think that this is especially critical in your plant where the general difficulty of obtaining spare parts can put machines out of production for e\$tended periods. In addition, your past problem with circuit overheating was worsened by your plantIs low power factor.

k Ar 'measure of unused power*

Recommendation We recommend that you contact your utility about a power -uality survey to identify the circuits with the lowest power factors. Pou could then install corrective capacitance on those circuits or on the main utility feed. In this way, you could raise the power factor into the acceptable range of :56 or greater. This will increase the lifetime and efficiency of your e-uipment and help minimize production shutdowns. It will also provide modest savings on your utility bill. Lur recommendation of (,555 k Ar represents the ma\$imum capacitance you currently need. Desser amounts of capacitance would reduce your utility savings accordingly. We estimate the cost of (,555 k Ar to be appro\$imately <>5,555 with simple controls. It could be as high as <+55,555 with more complicated controls and features. @ased on utility savings alone, the simple payback would be 4.C yearsJ however, the payback would be shorter if production and e-uipment maintenance savings were included. Fstimated /avings (,555 k Ar,month \$ +( months,year \$ <5.C5 ,A AR & <?,(55 Fstimated Implementation 3ost (,555 A Ar \$ <C5 ,A Ar & <>5,555 Fstimated /imple %ayback /imple %ayback & <?,(55 , yr & 4.C years AR T0 IM/TADD TWL +557A AR 3A%A3ITLR @AMA/ %resent Flectric .emand 'kW* Fstimated Implementation 3ost0 /imple %ayback0 Recommended Annual /avings G?> k AJ <C,=?5 <+,555 for installation G months
<>5,555

Analysis Pour electric demand charge is assessed in terms of k A, not in kW. This means that the electrical demand charge directly depends on the overall power factor in your plant. A high power factor results in several benefits0 lower demand costs, lower voltage drop in the circuit, lower overheating in wiring, and less maintenance. The power factor in the last +( months varied between 44.46 and :?.:6. /ome capacitor banks were already installed on bus bars high above the floor. Pou also have five +557k AR capacitor banks in stock. Recommendations We tried to strike a balance between raising power factor and not to cross to !lead" power factor. Lur analysis indicated that you could install two +557k AR capacitor banks to your circuits, preferrably at locations where some low power factor e-uipment items such

as welders, rectifiers, and 2I. lamps with low power factor ballasts, are located. In addition, we have learned from the e\$perience in other plants that an audio failure warning system allows for instant detection of capacitor malfunction and thus helps avoid e\$cessive cost penalties if a capacitor bank fails. Without audio failure systems, it could be many months before you discover a capacitor bank failure. Fstimated 3ost /avings %resent The billing demand varied between 4(: k A and +,C4( k A, totalling +C,>=G k A for the year 'August :G to Uuly :=*. After adding two +557k AR capacitor banks The annual total billing demand will be lowered from +C,>=G k A to +C,544 k A for a reduction of G?> k A /avings G?> k A \$ <?.=5,k A & <C,=?5,year in annual demand cost savings Fstimated Implementation 3ost <+,555 for installing two +557k AR capacitor banks Fstimated /imple %ayback G months AR T0 IM/TADD 3A%A3ITLR/ TL RAI/F %LWFR #A3TLR Flectric .emand 'kW* Fstimated Implementation 3ost0 /imple %ayback0 %resent (,55C kW After AR 5 kW Annual /avings (,55C kWJ <(=,4+: <C5,555 +G months

Analysis and Recommendation Pour monthly power factor during the past +( months was consistently below the :56 level re-uired by your utility company. As a result, a total of (,55C kW in accumulated power factor related demand penalties were assessed resulting in <(=,4+: in cost penalties. The power factor varied between 4C.C6 and 4=.?6. This low power factor problem could be corrected by adding capacitor banks at the source of the problem, the electric main, or both. When large electric motors are operated below =56 of their rated capacity, their power factor decreases signficantly. The thermo7mechanical refiner was e-uipped with a huge ?55 hp motor. Lther large motors include a =55 hp motor at the hydropulper, G55 hp and a +=5 hp motors at wood refiner, and a +55 hp motor at primary screen. These are good candidates to check for under loading. If so, capacitors should be installed at the motor.

=55 k Ars of capacitors will be needed to raise the power factor to :56 level. Pou should consult local heavy e-uipment vendors to check out motor loading problems and to determine how to position the capacitors between the motors and the electric main. /ome capacitors have already been installed, but it is suspected that they were either knocked out by lightning or power surges. If e\$isting capacitors and fuses can be repaired and put back into operation, the net need of new capacitors will be reduced accordingly. Fstimated /avings @ased on last yearHs utility bills, we estimate that adding =55 k Ar would save you about <(=,4+: per year in utility penalties. It would also reduce heat generation and voltage drop, and allow you to decrease the size of wiring. The improvement of power factor from 4C6 to :+6 would result in an 4.46 reduction in electric current, an 4.46 reduction in voltage drop across the circuit, and an +46 reduction in e\$cess heat generation. Fstimated Implementation 3ost <C5,555 for =55 7 >55 k AR capacitors and controls 3apacitors installed at the electric main might re-uire a sophisticated controls costing an additional <+5,555. /imple %ayback +G months AR T0 IM/TADD AO.I@DF 3A%A3ITLR #AIDORF ADAR1 Flectric .emand Implementation 3ost0 /imple %ayback0 %resent +(C kW Recommended 5 <+=5 Annual /avings +(C kWJ <+?C+ <+=5 + month

Analysis Recently (55 k Ar of capacitance failed. The visual alarm which should signal when the capacitor bank is not operational also failed. This dual failure caused the power factor of the plant to drop below :56 and cost the company <+,?C+ in demand penalties. Recommendation Although a properly functioning visual alarm may alleviate such problems in the future, we recommend installing an audible alarm. This type of alarm is almost impossible to ignore and assures that you will be alerted to a capacitor bank failure. Fstimated /avings and Implementation 3ost Dooking backward, we can say that a malfunctioning alarm caused the company to be charged for +(C kW in demand penalties costing <+,?C+. A functioning audible alarm

would have avoided this cost. Audible alarms can be purchased and installed for about <+=5. /imple %ayback /% & <+=5 initial cost , <+,?C+ ,yr savings \$ +( months,yr & + month