Content uploaded by Vitaly Gelman
Author content
All content in this area was uploaded by Vitaly Gelman on Feb 09, 2016
Content may be subject to copyright.
PHOTO BY SIEMENS
BRAKING ENERGY
RECUPERATION
Reversible Thyristor-Controlled Rectifiers
Vitaly Gelman
S
ilicon diode rectifier (SDR)
used in a majority of U.S. trac-
tion systems is a 40-year-old
technology; it provides no
active voltage control and does
not allow for power recuperation
into the ac line. New technolo-
gies such as thyristor-controlled
rectifiers (TCRs) provide active
voltage control, and reversible
TCRs (RTCRs) allow for power
recuperation into the ac line.
The first U.S. traction TCR
was put into revenue service in
Dallas in 1996. The Dallas system
was expanded in 2001, and Phoe-
nix recently started its TCR reve-
nue service as well. In the early
1990s, much attention was drawn
to the TCR, but then, the interest
faded away, returning now when the
energy savings became a priority.
TCR and SDR Basics
SDR’s output characteristics are deter-
mined by rectifier transformer parame-
ters and, to a lesser extent, by system
impedance. Transportation agencies have
accumulated enough experience with SDR so
that they can specify the transformer and, thus,
assure the desired voltage regulation curve, fault
current, etc. The SDR does not require a controller.
The TCR consists of two parts: power circuit and con-
troller with regulator. RTCR is a TCR with an additional power
module to conduct the current in the reverse direction. It is the job
of the regulators to send a firing pulse (also called gating signal) to each
silicon-controlled rectifier (SCR) to achieve a desirable voltage control, resulting
in voltage regulation, current limiting, and proper inverter operation.
Digital Object Identifier 10.1109/MVT.2009.933480
82 ||| 1556-6072/09/$26.00©2009IEEE IEEEVEHICULAR TECHNOLOGY MAGAZINE |SEPTEMBER 2009
Proper design of the regulator requires the use of feed-
back, typically multiloop arrangements (regulation of the
two parameters is needed: voltage and current). More-
over, the operation of the TCR is nonlinear: there is a volt-
age regulation operating region, a voltage-starving region
(the firing angle is minimal, the operation is similar to
SDR), and a current regulation region. Such design neces-
sitates the application of control theory.
It had been field proven in Dallas Area Rapid Transit
(DART) that TCRs with proper regulator work reliably and
without oscillations; they provide voltage/current within
power circuit limitations. While specifying the TCRs, the
transportation agencies need to characterize both static
and dynamic regulator performance along with power
circuit parameters. It is important to request and check
U.S. field references; other locations often have different
requirements and standards.
TCR Voltage Gain over SDR
Electric trains are specified to have 20% overvoltage mar-
gin, i.e., trains with 750-V dc-rated voltage are specified to
operate reliably at 900 V. If the voltage exceeds 900 V, the
train converters shut down. These are the parameters of
the M7 trains used by Long Island Rail Road, Metro North,
and others.
With SDR system, an operating voltage around 700–
750 V is selected. With a ratedvoltage of 700 V and 6% regu-
lation, there is a no-load voltage of 745-V dc (12-pulse recti-
fier will have no-load voltage higher by additional 3%). The
incoming ac-line voltage increase of 10% gives a no-load
voltage of 820-V dc. This leaves just enough margins to
provide the system receptivity for regenerative braking
(the current needs to travel to the accelerating train).
With TCR, a constant voltage region from 0 to 150% and
a load of 825 V are selected; the TCR will adjust the firing
angle to compensate for any ac-voltage variation.
Figure 1 shows the regulation curves for both TCR and
SDR. We can see that TCR provides a 125-V dc voltage gain
at 100% load and a 150-V dc voltage gain at 150% load. If a
minimum voltage at the train of 500-V dc at 100% load as
design criteria is assumed, then for SDR, a voltage margin
to spend on the rail’s (both running and power rails) volt-
age drop is 700500 ¼200-V dc. For the TCR, under the
same condition, the margin of 825500 ¼325-V dc or 60%
gain and even higher if the voltages at 150% load are com-
pared. This voltage gain directly translates into a poten-
tial increase of substation spacing (and lower number of
substations). In the existing SDR system, the load is
increased by 60% by converting to TCR/RTCR without
adding new substations.
Smaller increase in the distances or load (say 30%) can
be selected to optimize other parameters: running rail
voltage, losses in the rails, etc. In a Phoenix project, the
number of substations was reduced 28%, from 18 (SDR) to
14 (TCR), right along with the estimates above.
Some systems (e.g., DART and Phoenix) opted for even
higher voltage of 845–850-V dc. Though this optimizes sub-
station spacing and rail losses, it does not leave enough
room to provide regenerative current rails’ voltage drop
(see Figure 7, train voltage).
Though the TCRs are more expensive than SDRs, they
comprise only a small portion, about 5–10% of the in-
stalled cost of the substation. The reduced number of sub-
stations translates into substantial capital cost savings
(see Table 4).
Train Run Simulation
Simulation Assumption
To simulate energy flow, the following assumptions
are used:
nthe substations are located at 1-mi intervals
nthe passenger stations are located at the substations
nthe distance between the passenger stations is 2 mi
nthe train operates in power mode; the power level P
is determined by the acceleration profile and friction
losses; the power (or acceleration) is set by the train
controller, and the current is determined by the avail-
able voltage I¼P=V, where Iis the current, Pis the
power, and Vis the voltage
nthe train is 10-car M7, and each car is 145,000 lb
nthe acceleration is 2 mi/h/s to 20 mi/h, then inversely
proportional to the speed up to 60 mi/h
THE FIRST U.S. TRACTION TCR
WAS PUT INTO REVENUE SERVICE
IN DALLAS IN 1996.
400
450
500
550
600
650
700
750
800
850
0% 100% 200% 300% 400%
Diode Rect
TCR
F
IGURE
1
TCR and SDR voltage regulation curves.
SEPTEMBER 2009 |IEEE VEHICULAR TECHNOLOGY MAGAZINE ||| 83
nthe deceleration is inversely proportional to speed
above 20 mi/h; at 60 mi/h, it is 0.667 mi/h/s, increases
to 2 mi/h/s at 20 mi/h, and stays 2 mi/h/s between 20
mi/h and zero speed
nthe train accelerates to 60 mi/h, travels at 60 mi/h,
and then decelerates to stop at the second substation
2 mi away
nthe rail impedance is 56 mX/mi (10 mX/1,000 ft)
nboth SDR and TCR are 6 MW units
nthe TCR rectifier voltage is 825-V dc at all loads
nthe SDR has 700 V rated load voltage and 6% regula-
tion (745 V no load voltage, Rrect ¼5:25 mX)
nto account for the losses in the car power train and
the rectifier transformer, 1) the efficiency gtrain of the
car power train is constant 80%, both for acceleration
and braking and 2) the efficiency grect of the rectifier-
transformer unit is 98.5% are assumed. We neglect
the effects of other trains on the energy flow.
Calculations
The losses in the rails are calculated based on the train
position and current.
The friction force (in pounds) is calculated using
Davis formula:
FDAV(v)¼"(0:0025 þ(N1)0:00034)140v2
þv30:03 M
1,000 þ29 343Nþ1:33M
1,000 #,(1)
where Mis the total train mass in kilograms, Nis the num-
ber of cars (10), each car has four axles, and vis the train
speed in miles per hour.
The friction force converted to metric units is
Ffrict(v)¼FDAV v(mi=h) 1:6
3:6
30:454
0:102 ,(2)
where Ffrict(v) is a friction force in Nand vis the speed
in mi/h.
Assume the train to be xmeters from the left substa-
tion (rectifier 1), Lxmeters from the right substation
(rectifier 2), the distance Lbetween the substations is 1 mi
(1,600 m), and the rail’s impedance between the substa-
tions RLL ¼0:0056 X(see [1]). The currents from rectifiers
1 and 2 are I1and I2, respectively.
Figure 2 shows the equivalent circuit of our system.
The rectifiers are presented here as a series connection of
voltage source and equivalent resistance Rrect. For the
SDR, V1¼V2¼745 V and Rrect ¼5:25 mX; for the TCR,
assume V1¼V2¼825-V dc and Rrect ¼0.
Since V1¼V2, the potentials of their top terminals is
the same and can be replaced by a single voltage source
V1; the equivalent circuit is shown in Figure 3.
The equations to calculate the required variables are
as follows:
Ptrain ¼Pmech
gtrain
¼1
gtrain
(Ma þFfrict(v))v,
Ptrain ¼Pmech ¼IVtrain ¼I(VI3Req),
1
Req
¼1
Rrect þx
LRLL
þ1
Rrect þLx
LRLL
,
a¼dv
dt ;v¼dx
dt ,(3)
where Mis the total train mass, vis the train speed, ais
the train acceleration, Lis the distance between substa-
tions, and xis the distance from the train to substation.
To find out the losses in the rails, currents from each
substation I1and I2and, finally, the losses in the left and
RTCR IS A TCR WITH AN ADDITIONAL POWER
MODULE TO CONDUCT THE CURRENT IN THE
REVERSE DIRECTION.
Train
Current
Train Voltage
V1
l
1
l2
lV2
RLL × x/L
Rrect Rrect
Vtrain
Rectifier 1 Rectifier 2
RLL × (L–x )/L
F
IGURE
2
System circuit.
Train
Current
Train Voltage
V
1
l
1
l
2
l
R
LL
× x/L
R
rect
R
rect
V
train
R
LL
× (L–x )/L
F
IGURE
3
Converted equivalent circuit.
84 ||| IEEE VEHICULAR TECHNOLOGY MAGAZINE |SEPTEMBER 2009
right segments of the rail are calculated:
I1Rrect þx
LRLL
¼I2Rrect þLx
LRLL
,
I1þI2¼I,
Loss ¼x
LRLLI2
1þLx
LRLLI2
2:(4)
The last step is to calculate the energies: mechanical
energy of the train, energy spent on friction and drag,
energy lost in the rail’s resistance, and total energy con-
sumed from the ac source.
Emech ¼ZMav ¼Mv2
2,
Efrict ¼ZFfrict(v)v,
Erail ¼ZLoss,
Etot ¼1
grect ZI1(VRrectI1)þI2(VRrect I2), (5)
where Emech is the mechanical energy of the train, Efrict the
energy spent on friction and drag, Erail the energy lost in
the rails and overhead catenary system (OCS) resistance,
and Etot is the total energy consumed from an ac source.
For train braking, the equations are very similar,
except that the current and friction losses have opposite
polarity, and where when multiplied by efficiency we need
to divide, and vice versa, to account for the opposite
energy flow
Ptrain ¼gtrainPmech ¼gtrain (Ma Ffrict(v))v,
Ptrain ¼Pmech ¼IVtrain ¼I(VþIReq),
1
Req
¼1
Rrect þx
LRLL
þ1
Rrect þLx
LRLL
,
a¼dv
dt ;v¼dx
dt ,
Vtrain ¼VþIReq:(6)
An equation for the train voltage Vtrain is added and
needs confirmation that it will not exceed 900 V. The
energy equations are the same, except the one for the total
energy returned:
Etot ¼grect ZI1(VþRrectI1)þI2(VþRrect I2):(7)
Regeneration into the ac line is possible only with
RTCR. The simulation was performed using the Math-
Cad12 program. The speed and position were obtained by
integrating the acceleration. Power, voltage, and current
were obtained from (3) and (4), and for the regeneration,
from (6) and (4). Finally, the integrals (5) and (7) were cal-
culated to get energy balance and savings.
Train Acceleration
Figure 4 shows the results of train acceleration simulation:
traveled distance, speed, acceleration, and train power
during train acceleration from passenger substation to 60
mi/h for both SDR and TCR/RTCR. The acceleration (mi/h/
s) and distance (km) use the right Yscale.
The speed and train power use left scale, and the X
scale is the time in seconds. Assume here that the train
controller sets the acceleration and through it indirectly
sets the speed, position, and train power so that they will
be the same for both SDR and TCR. However, the train cur-
rent, rail losses, and total energy will be different because
TCR has a higher voltage.
Initially, at constant acceleration, the train current and
power increase linearly with time as expected, the force is
constant, and the power is proportional to speed. Above
20 mi/h, the acceleration is inversely proportional to
speed, corresponding to constant power. Actually, power
increases slightly because the friction force increases
with speed.
Figure 5 shows the results of simulation: train current
and rail losses during train acceleration from passenger
substation to 60 mi/h for both SDR and TCR. The current
follows the power, with the TCR current being substan-
tially lower than SDR because of the higher TCR voltage.
Two traces on the bottom show rail losses for TCR and
SDR. Since in both cases the train moves identically,
NET ENERGY IS THE SUM OF ACCELERATION
PLUS CONSTANT SPEED ENERGIES MINUS
RECOVERED ENERGY.
22
20
18
16
14
12
10
8
6
4
2
0
0 5 10 15 20 25 30 35 40 45 50 55 60
60
0
t
Train Power (MW)
Speed (mi/h/10)
Distance (km)
2
1.5
1
0.5
Acceleration (mi/h/s)
F
IGURE
4
Train acceleration simulation with SDR and TCR/RTCR.
SEPTEMBER 2009 |IEEE VEHICULAR TECHNOLOGY MAGAZINE ||| 85
equivalent rail impedance is the same for both cases, and
the rail losses are proportional to the square of currents.
While the train is close to the substation (first 10 s), the
rail losses are very small despite the high current. It hap-
pens because the distance xand rail impedance are small
[see (4)]. Once the train moves further away, the rail
losses rise rapidly; they reach a maximum in the middle
point between the substations and decline to zero once
the train approaches the next substation. A set speed of
60 mi/h at 800 m (half a mile) is reached from the original
substation or at the middle point.
Once the train gets closer to the middle point, the SDR
current rises disproportionably because of the higher volt-
age drop in the rails—higher current is needed to provide
the power. At the middle point, the SDR train voltage is
about 500-V dc, and for the TCR and RTCR, the train volt-
age is about 624 V. The TCR/RTCR provides higher train
voltage leading to lower train current and lower rail losses.
TCR and SDR Energy Balance for
Train Acceleration
Energy balance for acceleration for the same train using
an SDR and a TCR/RTCR is shown in Table 1. All values are
in megajoules (MJ). Mechanical energy, friction losses,
and car power train losses are the same for both cases,
because the train moves identically in both cases. How-
ever, the rail losses are much lower for the TCR because of
the lower current. Since the transformer/rectifier effi-
ciency is assumed to be the same for SDR and TCR, those
losses are lower for the TCR also. Total energy is lower by
about 6% (355 versus 378) for the TCR.
This is similar to reducing losses in a
transmission line by increasing the
voltage and thus lowering the current.
Train Deceleration
Figure 6 shows the results of train
deceleration simulation: traveled dis-
tance, speed, acceleration, and train
T
ABLE
1
Energy balance (MJ) for acceleration to 60 mi/h.
Rectifier
Type
Mechanical
Energy
Friction
Losses
Power
Train
Losses
Rail
Losses
Rectifier/
Transformer
Losses
Total
Energy
SDR 234 16.9 62.7 59.1 5.7 378.4
TCR 35.7 5.3 354.6
22
20
18
16
14
12
10
8
6
4
2
0
0 5 10 15 20 25 30
t
35 40 45 50 55 60
Train Power (MW)
Speed (mi/h/10)
Distance (km)
2
1.5
1
0.5
Accel (mi/h/s)
F
IGURE
6
Train deceleration.
22
20
18
16
14
12
10
8
6
4
2
0
0 5 10 15 20 25
t
30 35 40 45 50 55 60
Current (kA)
Rail Losses (MW)
(Train Voltage – 800 V)/10
F
IGURE
7
Train deceleration with RTCR.
22
20
18
16
14
12
10
8
6
4
2
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Train Current SDR (kA)
Train Current TCR (kA)
Rail Losses SDR (MW)
Rail Losses
TCR (MW)
F
IGURE
5
Train acceleration with SDR and TCR/RTCR.
86 ||| IEEE VEHICULAR TECHNOLOGY MAGAZINE |SEPTEMBER 2009
power during train deceleration from
60 mi/h to a stop at a passenger station.
Assume that the train controller
sets the deceleration and through it
indirectly sets the speed, position,
and train power so that they will be
the same for both SDR and TCR/RTCR.
Figure 7 shows the simulation re-
sults for the train decelerating from
60 mi/h to zero to stop at the passen-
ger station with substation. Since SDR
and TCR cannot absorb energy, there
are only results for RTCR.
The power and current are about
half of the values for acceleration
because of the losses in the cars
power train and also friction losses. At
an efficiency of 0.8, passing the energy
in both directions leaves only 0:82¼
0:64 of the initial energy; the rail losses and friction take
their toll also. So about 50% seems to be right. The train
power increases while its speed goes down because of
lower friction force at lower speed. The current increases
faster than the power because of the additional effect of
rail losses reduction.
RTCR and SDR Energy Balance
for Train Deceleration
Table 2 shows the energy balance for the deceleration.
The train mechanical energy is the same as in Table 1
(it depends only on mass and speed). The energy
recovered into the ac line is the difference between the
mechanical energy and the sum of all losses: friction,
power train losses, rail losses, and rectifier trans-
former losses.
With RTCR, 165 of 234 MJ can be recovered, which is
about 70% of the mechanical energy. With SDR, some of
this energy can be absorbed by nearby trains, and the rest
is wasted as heat. Since, in our simplified analysis, we do
not consider the effects of other trains, we have zero for
recovered energy in Table 3 for SDR.
RTCR and SDR Total Energy Balance
The train travels 800 m (half a mile) during both accelera-
tion and deceleration. Because the distance between the
passenger stations, in this case, is 2 mi, the train needs to
travel 1 mi at 60 mi/h; it will take 60 s. The constant speed
power is 890 kW (friction losses from Davis formula,
power train losses, and rectifier transformer losses), this
gives a constant speed energy of 54.2 MJ.
Table 3 shows energy balance for a train moving
between the two passenger stations 2 mi apart. Net energy
is the sum of acceleration plus constant speed energies
minus recovered energy. RTCR has a little lower accelera-
tion energy because of lower rails’ losses (higher voltage,
lower current), and constant speed energy is the same in
both cases. SDR does not recover any energy. It can be
seen from the table that RTCR offers 243.6 versus 432.6 MJ
or 44% propulsion energy savings.
Figures 8 and 9 show pie charts representing SDR and
TCR/RTCR energy balances, respectively (see Tables 1–3).
Total energy consumption for the same train run is
lower for the TCR compared with that of SDR by about 6%
(408 versus 432 MJ). This is mostly because of the reduced
rail loss.
RTCR recovers mechanical energy of the train. The
small circle on Figure 9 shows the energy balance for recu-
perating mechanical energy into the ac line by RTCR dur-
ing train deceleration (see Table 2).
Capital Cost and Energy Savings with RTCR
To estimate RTCR cost savings, the data from Table 3 are
used. Consider a short line with six SDR substations.
Mech
Energy 54%
Friction 4%
Pwr
Train Loss
14%
Rect/Xfmr
Loss 1%
Rail Loss
14%
Const
Speed
13%
F
IGURE
8
SDR energy balance.
T
ABLE
3
Net energy comparison for SDR and TCR (MJ).
Rectifier
Type
Acceleration
Energy
Constant
Speed
Energy
Recovered
Energy
Net
Energy Savings
RTCR 354.7 54.2 165.3 243.6 44%
SDR 378.4 54.2 0 432.6 0
T
ABLE
2
Energy balance (MJ) for deceleration from 60 mi/h to 0.
Rectifier
Type
Mechanical
Energy
Friction
Losses
Car Power
Train
Losses
Rail
Losses
Rectifier/
transformer
Losses
Total
Recovered
Energy
RTCR 234 16.9 43.4 6.8 2.5 165.3
SDR 234 0 0 0
SEPTEMBER 2009 |IEEE VEHICULAR TECHNOLOGY MAGAZINE ||| 87
Assuming a moderate distance gain of 20%, substitute
them with five TCR or RTCR substations. Assuming at
3-MW power level, the costs of SDR, RTCR with recupera-
tion to the ac line, TCR without recuperation, and the
installed cost of SDR substation are US$150,000,
US$350,000, US$270,000, and US$3 million, respectively.
Further assume that an average load of the SDR substation
at 25% of rated capacity is 0.75 MW for 24 h, and the cost
of energy is US$100/MW h. Energy savings is about 3% for
regular TCR and 30% for RTCR with recuperation; here,
the energy savings was reduced from 6% to 3% for regular
TCR and from 44% to 30% for regeneration RTCR to
account for longer spacing between
the TCR substations. These esti-
mates do not account for energy
absorption by nearby trains.
The results are compiled in
Table 4. The table shows that, as an
alternative to SDR, TCR and RTCR
provide both lower capital cost and
energy savings. The RTCR is a little
higher in capital cost (US$0.4 mil-
lion), but the payback time is just
five months.
RTCR versus SDR savings: initial
(capital) cost is 10% lower with
substantial energy savings.
Figure 10 shows the total cost
and saving over the 30-year period. After 12 years of
running, the savings will exceed the initial cost of
RTCR substations.
Savings for the RTCR Upgrade
Consider an upgrade of the existing SDR system to the
RTCR. For example, the same system with six SDR substa-
tions is used with rated power 3 MW each and consider
the effect of upgrading it to RTCRs. The number of substa-
tions is assumed to be the same. To simplify the analysis,
assume that both rectifier and rectifier transformer needs
to be replaced, and the old units to be discarded with no
resale values. The costs of the
new RTCR, new transformer, and
their installation are US$350,000,
US$160,000, and US$100,000, re-
spectively. Assume that the same
energy cost for the SDR system as
before (US$.942 million/year) and
the savings due to RTCR energy
recuperation into the ac line of 40%
(a higher number than in Table 4 is
used because there is no increase
in space). As before, we do not
account for energy absorption by
nearby trains. Table 5 contains the
cost of the upgrade data.
From Table 5, the payback pe-
riod for the upgrade to the RTCR is
less than three years. The addi-
tional advantage of the upgrade
system throughput and train per-
formance improvements due to
increased dc bus voltage conse-
quently increased the train volt-
age (see the ‘‘TCR Voltage Gain
over SDR’’ section).
In a situation where reduced head-
way or heavier trains necessitate an
T
ABLE
5
Costs of upgrade to the RTCR.
RTCR SDR
Energy savings (US$ million/year) 1.58 0
Equipment and installation cost (US$ million) 3.66 0
Pay-back period (years) 2.3 0
Total savings after ten years (US$ million) 12.1 0
Total savings after 20 years (US$ million) 27.9 0
Total energy cost (US$ million/year) 2.365 3.942
Number of substations 6 6
New rectifier cost (US$) 350,000 0
New transformer cost (US$) 160,000 0
Installation cost, per substation (US$) 100,000 0
T
ABLE
4
Capital cost and energy savings.
Rectifier Type RTCR TCR SDR
Energy savings (US$ million/year) 1.2 0.1 0
Capital savings (US$ million) 2.0 2.4 0
Total savings after six years (US$ million) 9.1 3.1 0
Total savings after 12 years (US$ million) 16.2 3.8 0
Total capital (US$ million) 16.0 15.6 18.0
Total energy (US$ million/year) 2.759 3.824 3.942
Number of substations 5 5 6
Installed substation cost (US$ million) 3.20 3.12 3.00
Rectifier cost (US$) 350,000 270,000 150,000
Friction 4%
Rect/Xfmr
Loss 1%
Rail Loss
9%
Const
Speed
13%
Mech
Energy 57%
Pwr
Train Loss
15%
Friction 4%
Rect/Xfmr
Loss D 1%
Rail
Loss D 2% Pwr
Train Loss
11%
Recovered
ac 40%
F
IGURE
9
TCR/RTCR energy balance.
88 ||| IEEE VEHICULAR TECHNOLOGY MAGAZINE |SEPTEMBER 2009
upgrade of existing SDR system calling for adding addi-
tional SDR substations, an upgrade to the RTCR offers an
attractive alternative:
nno new substations with related real estate cost
ncapital expenditures with payback period of less than
two years
nsubstantial energy savings
nthe performance improvement up to 60% is a free
benefit.
Figure 11 shows the savings of the substation upgrade
from SDR to RTCR.
Conclusions
This article estimates energy savings through braking
energy recuperation and capital cost savings through
increased substation spacing. The TCRs provide advan-
tages over diode rectifiers: better voltage regulation and
fault current limiting translating into some operational
savings (energy savings through increased dc bus voltage,
improved service life) and capital savings (reduced num-
ber of substations).
Assuming just 15% spacing increase in the new substa-
tion installations, capital savings with TCR are more than
10%. The additional savings with RTCR over SDR are
through braking energy recuperation back to the ac line.
Energy savings can be as high as 50% depending on the
train speed profile, train car efficiency, rail resistance, etc.
Upgrading existing SDR substations to RTCR gives
substantial energy savings and has a payback of two to two
and a half years, improving throughput up to 60% without
incurring additional real estate and construction expenses.
Acknowledgments
The author acknowledges Tom Young of Reuel for empha-
sizing the subject of TCR energy recuperation; Bob Puci-
loski and Asha Handa-Pierre from Long Island Rail Road
(LIRR) and Gordon Yu from SYSTRA for supplying applica-
tion data on M7 trains operation; Chuck Ross of PGH Wong
Engineering, John Frederick of Precision Power Systems
and Technology (PPST), Steve Sims of Bay Area Rapid
Transit (BART), and Raymond Stritmatter of Parsons for
supplying equipment and installation data and helpful dis-
cussions on the subject.
Author Information
Vitaly Gelman (vgelman@vgcontrols.com) received
his M.S.E.E. degree in 1976 from Moscow Power Univer-
sity. He is the president and founder of VG Controls,
which has been providing traction products and other
industrial electronic products since 1984 for compa-
nies like Asea Brown Boveri (ABB), BART, LIRR, Powell,
Controlled Power Corporation (CPC), Phelps Dodge,
and others.
Reference
[1] V. Gelman and S. Sagareli, ‘‘Implementation of new technologies in
traction power systems,’’ in Proc. JRC 2004: 2004 ASME/IEEE Joint Rail
Conf., pp. 141–146.
RTCR/SDR Upgrade Saving
Years
Total Cost (million dollars)
50.0
40.0
30.0
20.0
10.0
–10.0
0.0
51015202530
F
IGURE
11
SDR to RTCR substation upgrade savings.
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
0 5 10 15 20 25 30
Years
Total Cost (million dollars)
RTCR Cost
TSR Cost
SDR Cost
RTCR/SDR Saving
F
IGURE
10
Costs of ownership with different rectifiers.
SILICON DIODE RECTIFIER (SDR) USED IN A
MAJORITY OF U.S. TRACTION SYSTEMS IS A
40-YEAR-OLD TECHNOLOGY.
SEPTEMBER 2009 |IEEE VEHICULAR TECHNOLOGY MAGAZINE ||| 89