dc islands in ac smart grids.pdf
TRANSCRIPT
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Abstractmdash The advantages arising from DC distribution
networks is related to the possibilities of achieving higher quality supply and easier reconfigurability of the system This paper presents the concept of introducing some DC islands interconnected with the AC distribution network This will make it easier to connect storage systems electrical drives power converters and renewable sources (ie photovoltaic panels) In order to realize the aforementioned goals a methodology for designing the control strategy of different power converters connected to a DC bus without a centralized system management control unit is proposed In this way a plug and play functionality for connecting new power converter interfaced elements in the DC network able to stabilize the voltage under different working conditions is realized In order to prove the advantages obtainable with such a network a 100 kW DC test facility has been realized at the laboratories of RSE Different working conditions were tested by means both of numerical and experimental results proving the effectiveness of the proposed strategy for voltage regulation continuity of service and the smart use of storage devices
Index Termsmdash DC distribution networks Smart grid Storage Systems
I INTRODUCTION
owadays distribution networks are changing substantially and rapidly Environmental constraints with
particular attention to the reduction of CO2 are leading the governments of major countries to use more and more renewable energy particularly from wind and solar sources [1]-[4] These sources are characterized by i) distribution in an area ii) low power density for land occupation and iii) discontinuity in power availability [5] Moreover power plants based on renewable energy are usually much smaller than traditional electrical generation plants and the use of storage systems is advisable to benefit from the generation during low power demand periods and reduce alternative power flows along transmission lines with consequent advantages in terms of system efficiency [5]-[6]
This scenario is causing a change in distribution networks
Manuscript received October 19 2012 revised January 15 2013 accepted
February 24 2013 S Grillo L Piegari and E Tironi are with the DEIB (Sipartimento di
Elettronica Informazione e Bioingegneria) of Politecnico di Milano I-20131 Milan Italy (email samuelegrillopolimiit luigipiegaripolimiit enricotironipolimiit)
V Musolino is with DIMAC RED Biassono MB Italy (e-mail vmusolinodimacredit)
C Tornelli is with the TampD technologies Department of RSE Via Rubattino 54 I-20134 Milan Italy (email carlotornellirse-webit)
from passive to active grids where the power flows fault currents and temporal power profile are quite different from the traditional ones [7] Power injection on low voltage (LV) and medium voltage (MV) grids makes voltage regulation a difficult task For this reason different distribution system operators (DSOs) require mutual cooperation with the demand-side control because it is difficult to maintain the power quality using only supply-side control [8]-[10]
At the same time the great evolution in informatics and telecommunications is leading the transformation of the traditional grids into smart grids where the continuous dialogue between all of the devices connected to the network implies improvements in the overall efficiency and quality of the system Moreover it is also important to underline the wide diffusion of power electronic converters capable of decoupling the parameters (ie voltage and frequency) of the generators from the parameters of the network
In this scenario it could be convenient in some cases to substitute a DC grid for a traditional AC distribution network [11]-[13] Indeed a DC network besides its well-known advantages with respect to AC systems [14] offers a simplification in relation to the power electronic converters for all of the devices operating on DC ie photovoltaic (PV) systems and many storage systems Moreover it could eliminate all of the rectifiers present as a first stage in electrical drives and suppliers For this reason it is reasonable to imagine future smart grids realized using AC backbones and some DC leaves in the zones characterized by many small generation units storage facilities and users that are particularly susceptible to voltage disturbances In such a network it is possible to separate the backbones from the leaf regulation to ensure
high power quality of the DC network management of power fluxes exchanged between the AC
and DC sides with integrated ancillary services such as reactive power supply and harmonic compensation
no increase in the fault currents even in the presence of new generation units
the inhibition of disturbance propagation from the AC to DC side and vice-versa
Starting from the aforementioned considerations a DC test-facility has been realized with the aim of verifying the potentiality of a DC island interconnected with an AC distribution network The introduction of different power converters especially those interfaced with different storage and renewable resources requires a suitably designed management system in order to avoid uncontrolled power flows between the different elements In order to realize a
DC Islands in AC Smart Grids
S Grillo Member IEEE V Musolino L Piegari Member IEEE E Tironi C Tornelli
N
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
fucodiprwoanDCin go
coislthsyincostuthcowhfothan
ItawiPo
Fig
ully plug and pontrol unit a mfferent conve
resented in neorks without
nd the insertionC network do order to guar
oals obtained t Designing
storage desystem dyn
Designing converter i
Designing account th
Guarantee charge of t
In this way thonnected with lands to improe system the
ystem solutifrastructure
omponents is rudies are desce DC island i
omponents ushile Section V
or each convere results of n
nd discussed
The configuralian Researchith the Depolitecnico di M
g 1 Test facility
play solution wmethodology ferters based ext Sections exchanging i
n or disconneces not requirerantee the northrough this m a DC netwo
evices are connnamic respons disjoint contin order to avo the power co
he behavior of an embeddethe different sthe paper analythe introduct
ove the powermethodology
ion withoutThe descriptireported in Secribed in orderintroduction Ised for numeV proposes thrter of the DCnumerical and
II TES
ration of the th center on Enpartment of Milano is show
configuration
without a centrfor designing on traditionaThis means t
information wction of a sing
e a control sysrmal system o
methodology liork where diffnected in ordese performanctrol loop for oid control oveonverter contrthe controlled
ed self-regulattorage devicesyzes besides tion in a smarr quality and ey to design a t centralizedion of the tection II In Ser to demonstraIn Section IV
erical simulatihe methodologC network Find experimenta
ST FACILITY
test facility renergetic System
Electrical Ewn in fig 1
ralized managthe control lo
al PI controllthat each con
with the othergle element frostem reconfiguoperation Theie in ferent typologer to fulfill difces
each single erlapping
rol loops takind device tion of the sts the relevant art grid of somenergy efficienfully plug and communiest facility aection III somate the advant
V the models ions are disc
gy to set the cnally in Sectioal tests are re
ealized at RSms) in collabo
Engineering o
gement oops of lers is nverter r units om the uration e main
gies of fferent
power
ng into
tate of
aspects me DC ncy of d play ication
and its me case tage of of the
cussed control on VI
eported
E (the oration of the
Ttechthe opedevequ
Inof th
Tcomelem[15]
TkWleveSpacaseflowconvoltconwithcom32 kalonmodEacgridcharDCDC connTheflucpropgenpowbothPV genactihouare a to
The test facilihnical and eco
potentialitiesrated with suices connecteipped with dif
ensure the sAC grid is nDC networkstabilize theoccurs betwrealize a peor intermittethe AC gridimplement aregenerativerealize the that result fr
n the followinhe test facility
The LVDC mimposed of a fments a photo]
The front end which is a sel to connect tain Germanye of insufficiew in the case overter participtage and alsodition by comh their admiss
mposed of two kW capacity ng with two dules (30 kW ch Zebra batted through a 35rging and diDC converterbus will be i
nected to thee supercapacctuations in pagation fromeration chang
wer from the Ph space constrfield-emulato
erated powervity is not con
urs Finally twinstalled in th
otal load of 60
ity was realizonomic constrs of LVDC uitable controed to the netfferent storageecurity for denot available ak is insufficiene DC voltage
ween the powerak shaving acent loads thus an energy rece loads bridging and
rom the presenng a brief descy is given crogrid has a front end ACovoltaic field e
converter (Fignificant leveto an LV AC and so on) [1
ent or no genof a generationpates in the o detects the
mparing the ACsible ranges TZEBRA batteof 64 Ah andsupercapacitofor 4 s) with
ery and capac-kW DCDC bscharging pror interface betindicated by Ze supercapacitcitor banks
the DC vom the AC ges in the DCPV fields avaraints and meteor (PVC) has r can be connstrained by w
wo programmahe DC micro-g0 kW One ba
zed by takingraints with th
active distriol strategies twork The ne units in orderfined intervalsand the power
nt to cover the e level if a sr generated anction in the prs leveling the
overy action
energy manance of renewabcription of the
voltage level CDC converteemulator and
EC) has a rael compared wgrid in severa6] This allow
neration and n surplus Theregulation of e occurrence C grid voltageThe energy steries each witd open circuit or banks eachh a maximum itor bank is cbi-directional ocesses In thtween the ZEBZC while the tors will be in
compensate oltage causedgrid or sud
C grid Becauailable at the eorological cobeen chosen
ntrolled and weather condiable purely regrid these twoank of 30 kW
g into accounte aim of studibution netwofor the diffe
network has r to s of time whenr generated onpower deman
sudden imbalnd consumedresence of pulpower draine
in the presenc
agement functble sources main compon
of 400 V ander energy stocontrollable l
ated power of with the maximal countries (I
ws power flowalso the opp
e ACDC frontf the DC netw
of an islanes and frequentorage systemsth a rated powvoltage of 27
h made of novoltage of 38
coupled to theconverter to ahe followingBRA batteriesDCDC conv
ndicated by Sfor very
d by faults dden load anuse of the limsite as a resu
onditions a 35 In this waythe experim
itions and dayesistive load-bo units can pro
W can be regul
t the dying orks
ferent been
n the n the nd lance
lsing ed by
ce of
tions
nents
d it is orage loads
f 100 mum Italy
w in a osite t end work nding ncies s are
wer of 79 V o 24 84 V e DC allow the s and verter SCC
fast dip
ndor mited ult of 5-kW y the ental
ytime banks ovide lated
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
wicoth
acse
A
tostepe
fapofalothpo
Fig
B
opto cothhat=steloex
C
anlocodr13
ith step changonnected to the fast regulati
In order to chieving the abet and analyzed
i Pulsii Faul
is coiii Reg
Pulsing load
A pulsing mography (CAep load varyineriod of 25 s a
The power rcility Howev
ower of the cilities Thusad has been prat of a real CA
ower absorbed
g 2 Pulsing load
Fault in AC
In the secondpening of the A the AC distr
onverter detecte converter
appens at tim=1635 s Meaeps simulatingads on the D
xperimental tes
Regenerativ
Usually elecnd braking opads For this r
onducted by aprive for a woo9 s with an a5 kW) consta
ges of 1 kWhe DC bus by on of the pow
III CA
test the capabove discussedd sing load lt of the AC nonnected to theenerative load
d
load is reprAT) scanner wng between aband a duty cycrequested by tver it can be s
different dev the DCDC crogrammed toAT although
d by the load d
d simulating powe
C network
d test an AC fAC breaker coribution netwots the AC-sidefrom the AC
me t=47 s wanwhile the lg the connectiDC bus In fst is reported
ve load
ctrical drives perations are reason the tespplying the reodworking plaacceleration pant speed pha
while the otmeans of a D
wer absorbed
ASE STUDY
abilities of thd goals three
network while e DC bus
d
resented by whose typicalbout 60 kW a
cle of 70 this load is toscaled by taki
vices connecteconverter conno absorb a pow
opportunely sduring the tests
er profile of CAT
fault is simulaonnecting the ork The contre lack of volta
C network Twhile the AC load demand ion and disconfig 3 the po
alternating bethe most di
t to simulate real power demant [17] The phase (maximuase (power ab
ther 30 kW bDCDC conver
he DC netwokinds of case
a step variabl
a computer l power diagraand 180 kW w
oo high for thing into accoued to the DCnected 30 kW
wer profile simscaled In fig s is reported
ated with the mfront-end con
rol algorithm age and disco
The simulatedreconnection
changes withnnection of difower profile
etween acceleiffused regeneregenerative lo
mand of an eleworking cyclum power abssorbed 700 W
bank is rter for
ork in s were
le load
axial am is a with a
he test unt the C test
W rated milar to
2 the
manual nverter of the nnects
d fault n is at h some fferent of the
eration erative oads is ectrical e lasts sorbed
W) and
brakmeaThecan insedrivconnexp
Fig
Fig
Indiffelecof toveopeapp
Insign
Agavdev
Tdynpartelec
king phase (man power absoerefore it is cl
lead to a sertion and discve has been nected at timeerimental pow
3 Load power du
4 Power diagram
IV M
n order to coferent convertctrical storage the model to rr the full dynrative workinlications are
an energy futhe average a peak shavabsorbdeliva certain ref
n general tnificantly loweAmong the vae attention toice behavior a
A ZEB
The ZEBRA banamic modelsticular in [18]ctrochemical i
maximum poworbed during olear how the ssignificant enconnection of simulated In
e t=32s and dwer diagram is
uring AC network
m of simulated reg
MODELING OF E
orrectly set uers of the DCdevices is re
represent the namic range ofng conditions
function in whpower requestving function
ver the requestference averagthe first funer charge and arious models
lumped modas seen from th
BRA battery m
attery model cs available in] and [19] leaimpedance spe
wer regeneratone cycle is lostorage action nergy recoveranother load in particular disconnected as reported in fi
k fault test
generative load
ELECTRICAL ST
up the controC system thequired In parbehavior of thf utilization iss of a storag
hich the deviceted by the app
n in which thted peak pow
ge power nction is chadischarge dyn available in
dels capable ohe terminals
model
comes from thn the literatuad acid batteryectroscopy (E
ted 17 kW) ower than 200of the DC sy
ry Moreoverin parallel witha 9 kW loa
at time t=47sig 4
TORAGE
l strategy fore modeling ofrticular the abhe electric stos needed Thege device in
e has to guaraplication he storage haer above or b
aracterized bnamic
the literaturef interpolating
he lead acid baure [18]-[21]y models base
EIS) are descri
The 0 W
ystem the h the ad is The
r the f the bility orage e two
grid
antee
as to elow
by a
e we g the
attery ] In ed on ibed
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
whco
Fig
(SthR1
thac
coeq
E
So
chnocu
A
dymin[2id
Fig
hile in [20] aompromise bet
The model us
g 5 ZEBRA batt
The ZEBRA SoC) represene state of char1 which takese device and
ccount the dynIn addition
onsidered to bquations descri
ESoCE )( 0
C
oCn
11
where
E0 is the opharged SoC isominal batteryurrent supplied
Supercapac
A supercapaynamics of th
mHz up to 10troduced in [2
24] where adentification w
g 6 Supercapacit
In this model Ri represen
available f C(u)=C0
RR
and [21] thirdtween complesed for our pur
tery model
battery is modnting the openrge (SoC) in ss into accountd the parallelnamic behavior
for our purpbe constant duibing the batte
SoCKe 1
dtib
pen circuit vos the state of
y capacity Ke d by the batter
citor model
acitor modele device in th
00 Hz is repr23] was intega simplified
was presented
tor model
l nts the high from the manu+KVyuml u wh
C(u)Ri
1
222
u
2
C
C(u)Ri
1
222
u
2
C
d-order modelxity and precirpose is repres
deled as an elen circuit voltaseries with thet the high freq branchR2-C2
r of the batterypose the batturing the workery device are
oltage when tf charge of this a voltage co
ry
l capable ofhe typical freqresented in f
grated in a moprocedure f
frequency resufacturerrsquos dathere C0 and
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
ls that show aision are descrsented in fig 5
ectromotive foage as a functe internal resisquency resista2 which takey tery temperatk cycle so th
the battery ishe battery Cn
onstant and ib
f representinquency range fig 6 This more general mofor the para
sistance whictasheet
KV are con
)(
)
uC vsc
isc
)(
)
uC vsc
isc
a good ribed 5[22]
orce E tion of stance
ance of es into
ture is hat the
(1)
(2)
s fully is the
b is the
ng the of 10
model odel in ameter
ch was
nstants
TreprFor lowsign
Adefiachithe 1
2
34
Itconconand
Tthe Howprefcapaotheconbe aand worquathessuitmodstorautovaluhas at thvoltto cooconconothevolt
Tassibetwwouano
evaluated byin [24] τ(u)=3yuml(Rdc-Rresistance re
The supercaparesent the redour purposes
w frequencies nificant error
A control strained In ordeieving all ofcontrol strateg
The stabilizaat a steady sThe automaone or moreThe self-recThe optimal
t is evident ttrol has to bverters that ca
d PVC The possibility
use of only owever in ordeferable to havable of workier units The trol laws for tadapted to the
d have to be rking conditiolity on the DCse laws can btably tuned difications conrage system onomously itue The tuningto be coordina
he same timetage Thus theoptimally exprdinated withverters On thtrols of the foer because thtage and curre
The duty of imigned only toween the voluld cause a cother even in
y a constant c
Ri)yumlC(u) wheported in the acitor and battistribution and
s these phenoalmost close t
V CONTRO
ategy for the er to make the aims discgy should ensuation of the D
state for differeatic configurabe devices are uharge of the stl utilization of that to achiebe implemenan regulate th
y of splitting one controller
er to realize a mve separate coing independeproposed con
the four convedifferent devintegrated in
ons for the difC bus As willbe obtained uand modifie
nsists in the shas to im
ts own state g of the dynamated to make e to guarantee e dynamic respploit the conh the dynamhe contrary frour converters hey are basedent at their conmposing a stao one converltages measurontinuous pown the absenc
current charge
here Rdc repmanufacturertery models ud self-discharmena which to DC can be
OL STRATEGY
network conthe DC net
cussed in the ure
DC voltage durent kind of loability of the
unavailable torage system
f all of the devve the first g
nted Howevehe DC voltage
the voltage rr or of coordmore robust coontrollers for tently from thntrol strategy erters The coices they conn
order to ensfferent systeml be shown in using traditioed One ofself-recharge f
mplement in of charge ar
mic responseseach device wthe stabilizati
ponse of each ntrolled devic
mic performanrom an operati
are independd only on thennection nodeationary DC brter otherwisred at the diwer flow frome of loads
e test as descr
presents the rsquos datasheet used are unabrge of the devtake place at neglected wit
nverters has twork capableprevious sect
ring transientsads control schem
ms vices goal DC volr there are e FEC ZC S
regulation requinated controlontrol it woulthe different u
he presence ofis to design
ontrol laws havnect to the DCsure both opt
ms and high pothe next Sect
onal control lf the innovafunction that
order to kround the des of the conve
work at its best ion of the DCconverter is tu
ce but it is nces of the oion viewpoint
dent one from e measuremen
bus voltage case the differfferent conve
m one convertThe stiff vol
ribed
DC
le to vices
very thout
been e of tions
s and
me if
ltage four
SCC
uires llers ld be units f the four
ve to C bus timal ower ions
loops ative each
keep sired
erters and
C bus uned also
other t the each
nt of
an be rence erters er to ltage
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
reofenit imalgthreabnefureimchthSCsydethdufacoth
this suvoth
adpogrduwompebaofth
febavafinun
thAfoncoaranexan
gulation has bf energy in thnergy stored in
is not depmplements a gorithm in orde PV field It ached Indeed
bsorb the eneretwork is not ully charged)gulation but
mplemented Inharging actione state of charCC are depenystems respecefined as a fune optimal stat
uring the nightilure of the
overed by the e loads duringIn order to aey are close t
tuned to thupercapacitorsoltage) In thiey are far fromIn order to ac
dvisable that thossible feedinrid Moreoveruring transientork at time p
minutes The serformances oandwidths of tf different bane different devAs shown beledback chain
andwidth for aalues for the rend the closednits
A Fr
The FEC is ae FEC a phafter locking thn the rotating omponents of re implementend the switchixternal one hand the direct an
been committehe AC networkn the storage spendent on w
maximum pder to draw thshuts down if
d this happensrgy produced available and
The ZC ain order to enn particular i
n the droop isrge Thus the ndent on the ctively An opnction of the gte of charge fttime and loweAC network
PV field whig the nighttimeamplify the acto the optimalhe supercaparsquo energy (pros way their e
m the optimal chieve the opthe power suppng only the mr the supercts of a few seperiods ranginsplit of the pof each devicthe convertersndwidths impvices low each conand PI regula
a power unit egulator const
d loop transfe
ront end conve
a traditional acase lock loop he grid phasereference conthe current [2
ed The internaing componen
as the DC voltnd quadrature
ed to the FEC k is consistensystems and uweather condpower point e maximum pf a maximum s only if the sby the PV fi
d all of the sand SCC connsure stabilityin order to alss realized by voltage refereSoC values fptimal state ogoal of the netfor the batterer during the d
k during the le the PVC cae ction of the sul SoC the SCacitorsrsquo voltaoportional to energy contribstate of chargtimal use of thplied by the FE
mean power recapacitors shoeconds while ng from a fewpower accordie can be obtasrsquo controllers
plies different
nverter is contrator In order tit is necessarytants It is theer functions f
erter
ctive front end(PLL) algorith the control o
ntrolling the di25] Two nestal one uses thnt control signtage and react currents as ou
because the antly greater thunlike the PVditions The
tracking (Mpossible energy
voltage threshsystem is not aield (ie if thstorage systemntribute to vy a droop conso perform thtaking into ac
ences for the Zfor the two sof charge has twork For exaies could be daytime becau
daytime couannot give ene
upercapacitorsCC voltage ref
ge instead othe square o
bution is low e he devices it EC be the smoequest from thould supply the batteries sw seconds toing to the dyained based o Indeed the response tim
rolled by meanto obtain the dy to choose suerefore advisafor the three
d In order to chm is implemof the FEC opirect and quadted feedback
he currents as nals as outputive power as utputs In relat
amount han the V field
PVC MPPT) y from hold is able to he AC ms are voltage ntrol is e self-ccount
ZC and storage
to be ample higher
use the uld be ergy to
s when ference of the of the when
is also oothest he AC power should some
ynamic on the choice
mes for
ns of a desired uitable able to power
control mented perates drature chains inputs ts the inputs tion to
the If thcapathe
Fig
T
R2
wki arcaparepr
Asquareguof ththe
2
I
V
d
DC
Fglobresp
A
Tbidisidewhimeanestin tcomgensche
Fig
Tvoltof dcharslowcan tran
DC voltage che internal chaable of settingFEC can be dr
7 Control schem
The closed loop
dp
pd
VksC
skV
3
32
where Vd is there the constanacitance of thresenting the p
As shown in Fare of the DC ulator is easierhe DC voltagegrid current
2
3
RCs
RVd
d
C
From (3) it bally as a lowponse of the FE
Zebra conver
The ZC is reirectional stepe is connectedile the low voans of an induted feedback the inductance
mponents whierating the reeme of the ext
8 Control schem
The self-rechatage referencedroop action enrge around thw if compared
therefore bensfer function
control only tain is much fag the desired rawn as report
me of FEC
p transfer func
id
i
VksV
Rks
32
e direct compots of the PI reghe DC bus apower absorbeFig 7 the feevoltage In thr In fact in the directly depe
1C
R
is clear that w-pass filter wEC
rter
ealized with p-downstep-ud to the DC boltage side is
uctor In additichains The ine acting on thile the externaeference currternal loop is r
me of ZC
arge function e in function onsures that thehe desired vald to the dyname neglected in
the external chaster than the ecurrents the ted in fig 7
ction is
dV
onent of the ggulator C is thand R is a reed by the DC bedback chain is way indeed
he frequency dends on the di
the control whose bandwid
one inverter up converter bus with an os connected toion the ZC is nternal one cohe duty cycleal one regulaterent as an oureported in fig
is obtained of the state ofe battery tries lue In any ca
mic action of thn formulating
hain is interesexternal one ancontrol schem
rid voltage kp
he total equivaesistive parambus is realized on
d the tuning odomain the sqirect compone
system operdth limits the
leg acting The high volutput capacitao the ZEBRAoperated with
ontrols the cue of the switces the DC vol
utput The cog 8
changing the f charge Thisto keep its staase this actio
he control loopg the closed
sting nd is
me of
(3)
p and alent
meter
n the of the quare ent of
(4)
rates time
as a ltage ance A by h two urrent ching ltage ntrol
DC s sort ate of on is p and loop
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
fucodiprwoanDCin go
coislthsyincostuthcowhfothan
ItawiPo
Fig
ully plug and pontrol unit a mfferent conve
resented in neorks without
nd the insertionC network do order to guar
oals obtained t Designing
storage desystem dyn
Designing converter i
Designing account th
Guarantee charge of t
In this way thonnected with lands to improe system the
ystem solutifrastructure
omponents is rudies are desce DC island i
omponents ushile Section V
or each convere results of n
nd discussed
The configuralian Researchith the Depolitecnico di M
g 1 Test facility
play solution wmethodology ferters based ext Sections exchanging i
n or disconneces not requirerantee the northrough this m a DC netwo
evices are connnamic respons disjoint contin order to avo the power co
he behavior of an embeddethe different sthe paper analythe introduct
ove the powermethodology
ion withoutThe descriptireported in Secribed in orderintroduction Ised for numeV proposes thrter of the DCnumerical and
II TES
ration of the th center on Enpartment of Milano is show
configuration
without a centrfor designing on traditionaThis means t
information wction of a sing
e a control sysrmal system o
methodology liork where diffnected in ordese performanctrol loop for oid control oveonverter contrthe controlled
ed self-regulattorage devicesyzes besides tion in a smarr quality and ey to design a t centralizedion of the tection II In Ser to demonstraIn Section IV
erical simulatihe methodologC network Find experimenta
ST FACILITY
test facility renergetic System
Electrical Ewn in fig 1
ralized managthe control lo
al PI controllthat each con
with the othergle element frostem reconfiguoperation Theie in ferent typologer to fulfill difces
each single erlapping
rol loops takind device tion of the sts the relevant art grid of somenergy efficienfully plug and communiest facility aection III somate the advant
V the models ions are disc
gy to set the cnally in Sectioal tests are re
ealized at RSms) in collabo
Engineering o
gement oops of lers is nverter r units om the uration e main
gies of fferent
power
ng into
tate of
aspects me DC ncy of d play ication
and its me case tage of of the
cussed control on VI
eported
E (the oration of the
Ttechthe opedevequ
Inof th
Tcomelem[15]
TkWleveSpacaseflowconvoltconwithcom32 kalonmodEacgridcharDCDC connTheflucpropgenpowbothPV genactihouare a to
The test facilihnical and eco
potentialitiesrated with suices connecteipped with dif
ensure the sAC grid is nDC networkstabilize theoccurs betwrealize a peor intermittethe AC gridimplement aregenerativerealize the that result fr
n the followinhe test facility
The LVDC mimposed of a fments a photo]
The front end which is a sel to connect tain Germanye of insufficiew in the case overter participtage and alsodition by comh their admiss
mposed of two kW capacity ng with two dules (30 kW ch Zebra batted through a 35rging and diDC converterbus will be i
nected to thee supercapacctuations in pagation fromeration chang
wer from the Ph space constrfield-emulato
erated powervity is not con
urs Finally twinstalled in th
otal load of 60
ity was realizonomic constrs of LVDC uitable controed to the netfferent storageecurity for denot available ak is insufficiene DC voltage
ween the powerak shaving acent loads thus an energy rece loads bridging and
rom the presenng a brief descy is given crogrid has a front end ACovoltaic field e
converter (Fignificant leveto an LV AC and so on) [1
ent or no genof a generationpates in the o detects the
mparing the ACsible ranges TZEBRA batteof 64 Ah andsupercapacitofor 4 s) with
ery and capac-kW DCDC bscharging pror interface betindicated by Ze supercapacitcitor banks
the DC vom the AC ges in the DCPV fields avaraints and meteor (PVC) has r can be connstrained by w
wo programmahe DC micro-g0 kW One ba
zed by takingraints with th
active distriol strategies twork The ne units in orderfined intervalsand the power
nt to cover the e level if a sr generated anction in the prs leveling the
overy action
energy manance of renewabcription of the
voltage level CDC converteemulator and
EC) has a rael compared wgrid in severa6] This allow
neration and n surplus Theregulation of e occurrence C grid voltageThe energy steries each witd open circuit or banks eachh a maximum itor bank is cbi-directional ocesses In thtween the ZEBZC while the tors will be in
compensate oltage causedgrid or sud
C grid Becauailable at the eorological cobeen chosen
ntrolled and weather condiable purely regrid these twoank of 30 kW
g into accounte aim of studibution netwofor the diffe
network has r to s of time whenr generated onpower deman
sudden imbalnd consumedresence of pulpower draine
in the presenc
agement functble sources main compon
of 400 V ander energy stocontrollable l
ated power of with the maximal countries (I
ws power flowalso the opp
e ACDC frontf the DC netw
of an islanes and frequentorage systemsth a rated powvoltage of 27
h made of novoltage of 38
coupled to theconverter to ahe followingBRA batteriesDCDC conv
ndicated by Sfor very
d by faults dden load anuse of the limsite as a resu
onditions a 35 In this waythe experim
itions and dayesistive load-bo units can pro
W can be regul
t the dying orks
ferent been
n the n the nd lance
lsing ed by
ce of
tions
nents
d it is orage loads
f 100 mum Italy
w in a osite t end work nding ncies s are
wer of 79 V o 24 84 V e DC allow the s and verter SCC
fast dip
ndor mited ult of 5-kW y the ental
ytime banks ovide lated
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
wicoth
acse
A
tostepe
fapofalothpo
Fig
B
opto cothhat=steloex
C
anlocodr13
ith step changonnected to the fast regulati
In order to chieving the abet and analyzed
i Pulsii Faul
is coiii Reg
Pulsing load
A pulsing mography (CAep load varyineriod of 25 s a
The power rcility Howev
ower of the cilities Thusad has been prat of a real CA
ower absorbed
g 2 Pulsing load
Fault in AC
In the secondpening of the A the AC distr
onverter detecte converter
appens at tim=1635 s Meaeps simulatingads on the D
xperimental tes
Regenerativ
Usually elecnd braking opads For this r
onducted by aprive for a woo9 s with an a5 kW) consta
ges of 1 kWhe DC bus by on of the pow
III CA
test the capabove discussedd sing load lt of the AC nonnected to theenerative load
d
load is reprAT) scanner wng between aband a duty cycrequested by tver it can be s
different dev the DCDC crogrammed toAT although
d by the load d
d simulating powe
C network
d test an AC fAC breaker coribution netwots the AC-sidefrom the AC
me t=47 s wanwhile the lg the connectiDC bus In fst is reported
ve load
ctrical drives perations are reason the tespplying the reodworking plaacceleration pant speed pha
while the otmeans of a D
wer absorbed
ASE STUDY
abilities of thd goals three
network while e DC bus
d
resented by whose typicalbout 60 kW a
cle of 70 this load is toscaled by taki
vices connecteconverter conno absorb a pow
opportunely sduring the tests
er profile of CAT
fault is simulaonnecting the ork The contre lack of volta
C network Twhile the AC load demand ion and disconfig 3 the po
alternating bethe most di
t to simulate real power demant [17] The phase (maximuase (power ab
ther 30 kW bDCDC conver
he DC netwokinds of case
a step variabl
a computer l power diagraand 180 kW w
oo high for thing into accoued to the DCnected 30 kW
wer profile simscaled In fig s is reported
ated with the mfront-end con
rol algorithm age and disco
The simulatedreconnection
changes withnnection of difower profile
etween acceleiffused regeneregenerative lo
mand of an eleworking cyclum power abssorbed 700 W
bank is rter for
ork in s were
le load
axial am is a with a
he test unt the C test
W rated milar to
2 the
manual nverter of the nnects
d fault n is at h some fferent of the
eration erative oads is ectrical e lasts sorbed
W) and
brakmeaThecan insedrivconnexp
Fig
Fig
Indiffelecof toveopeapp
Insign
Agavdev
Tdynpartelec
king phase (man power absoerefore it is cl
lead to a sertion and discve has been nected at timeerimental pow
3 Load power du
4 Power diagram
IV M
n order to coferent convertctrical storage the model to rr the full dynrative workinlications are
an energy futhe average a peak shavabsorbdeliva certain ref
n general tnificantly loweAmong the vae attention toice behavior a
A ZEB
The ZEBRA banamic modelsticular in [18]ctrochemical i
maximum poworbed during olear how the ssignificant enconnection of simulated In
e t=32s and dwer diagram is
uring AC network
m of simulated reg
MODELING OF E
orrectly set uers of the DCdevices is re
represent the namic range ofng conditions
function in whpower requestving function
ver the requestference averagthe first funer charge and arious models
lumped modas seen from th
BRA battery m
attery model cs available in] and [19] leaimpedance spe
wer regeneratone cycle is lostorage action nergy recoveranother load in particular disconnected as reported in fi
k fault test
generative load
ELECTRICAL ST
up the controC system thequired In parbehavior of thf utilization iss of a storag
hich the deviceted by the app
n in which thted peak pow
ge power nction is chadischarge dyn available in
dels capable ohe terminals
model
comes from thn the literatuad acid batteryectroscopy (E
ted 17 kW) ower than 200of the DC sy
ry Moreoverin parallel witha 9 kW loa
at time t=47sig 4
TORAGE
l strategy fore modeling ofrticular the abhe electric stos needed Thege device in
e has to guaraplication he storage haer above or b
aracterized bnamic
the literaturef interpolating
he lead acid baure [18]-[21]y models base
EIS) are descri
The 0 W
ystem the h the ad is The
r the f the bility orage e two
grid
antee
as to elow
by a
e we g the
attery ] In ed on ibed
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
whco
Fig
(SthR1
thac
coeq
E
So
chnocu
A
dymin[2id
Fig
hile in [20] aompromise bet
The model us
g 5 ZEBRA batt
The ZEBRA SoC) represene state of char1 which takese device and
ccount the dynIn addition
onsidered to bquations descri
ESoCE )( 0
C
oCn
11
where
E0 is the opharged SoC isominal batteryurrent supplied
Supercapac
A supercapaynamics of th
mHz up to 10troduced in [2
24] where adentification w
g 6 Supercapacit
In this model Ri represen
available f C(u)=C0
RR
and [21] thirdtween complesed for our pur
tery model
battery is modnting the openrge (SoC) in ss into accountd the parallelnamic behavior
for our purpbe constant duibing the batte
SoCKe 1
dtib
pen circuit vos the state of
y capacity Ke d by the batter
citor model
acitor modele device in th
00 Hz is repr23] was intega simplified
was presented
tor model
l nts the high from the manu+KVyuml u wh
C(u)Ri
1
222
u
2
C
C(u)Ri
1
222
u
2
C
d-order modelxity and precirpose is repres
deled as an elen circuit voltaseries with thet the high freq branchR2-C2
r of the batterypose the batturing the workery device are
oltage when tf charge of this a voltage co
ry
l capable ofhe typical freqresented in f
grated in a moprocedure f
frequency resufacturerrsquos dathere C0 and
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
ls that show aision are descrsented in fig 5
ectromotive foage as a functe internal resisquency resista2 which takey tery temperatk cycle so th
the battery ishe battery Cn
onstant and ib
f representinquency range fig 6 This more general mofor the para
sistance whictasheet
KV are con
)(
)
uC vsc
isc
)(
)
uC vsc
isc
a good ribed 5[22]
orce E tion of stance
ance of es into
ture is hat the
(1)
(2)
s fully is the
b is the
ng the of 10
model odel in ameter
ch was
nstants
TreprFor lowsign
Adefiachithe 1
2
34
Itconconand
Tthe Howprefcapaotheconbe aand worquathessuitmodstorautovaluhas at thvoltto cooconconothevolt
Tassibetwwouano
evaluated byin [24] τ(u)=3yuml(Rdc-Rresistance re
The supercaparesent the redour purposes
w frequencies nificant error
A control strained In ordeieving all ofcontrol strateg
The stabilizaat a steady sThe automaone or moreThe self-recThe optimal
t is evident ttrol has to bverters that ca
d PVC The possibility
use of only owever in ordeferable to havable of workier units The trol laws for tadapted to the
d have to be rking conditiolity on the DCse laws can btably tuned difications conrage system onomously itue The tuningto be coordina
he same timetage Thus theoptimally exprdinated withverters On thtrols of the foer because thtage and curre
The duty of imigned only toween the voluld cause a cother even in
y a constant c
Ri)yumlC(u) wheported in the acitor and battistribution and
s these phenoalmost close t
V CONTRO
ategy for the er to make the aims discgy should ensuation of the D
state for differeatic configurabe devices are uharge of the stl utilization of that to achiebe implemenan regulate th
y of splitting one controller
er to realize a mve separate coing independeproposed con
the four convedifferent devintegrated in
ons for the difC bus As willbe obtained uand modifie
nsists in the shas to im
ts own state g of the dynamated to make e to guarantee e dynamic respploit the conh the dynamhe contrary frour converters hey are basedent at their conmposing a stao one converltages measurontinuous pown the absenc
current charge
here Rdc repmanufacturertery models ud self-discharmena which to DC can be
OL STRATEGY
network conthe DC net
cussed in the ure
DC voltage durent kind of loability of the
unavailable torage system
f all of the devve the first g
nted Howevehe DC voltage
the voltage rr or of coordmore robust coontrollers for tently from thntrol strategy erters The coices they conn
order to ensfferent systeml be shown in using traditioed One ofself-recharge f
mplement in of charge ar
mic responseseach device wthe stabilizati
ponse of each ntrolled devic
mic performanrom an operati
are independd only on thennection nodeationary DC brter otherwisred at the diwer flow frome of loads
e test as descr
presents the rsquos datasheet used are unabrge of the devtake place at neglected wit
nverters has twork capableprevious sect
ring transientsads control schem
ms vices goal DC volr there are e FEC ZC S
regulation requinated controlontrol it woulthe different u
he presence ofis to design
ontrol laws havnect to the DCsure both opt
ms and high pothe next Sect
onal control lf the innovafunction that
order to kround the des of the conve
work at its best ion of the DCconverter is tu
ce but it is nces of the oion viewpoint
dent one from e measuremen
bus voltage case the differfferent conve
m one convertThe stiff vol
ribed
DC
le to vices
very thout
been e of tions
s and
me if
ltage four
SCC
uires llers ld be units f the four
ve to C bus timal ower ions
loops ative each
keep sired
erters and
C bus uned also
other t the each
nt of
an be rence erters er to ltage
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
reofenit imalgthreabnefureimchthSCsydethdufacoth
this suvoth
adpogrduwompebaofth
febavafinun
thAfoncoaranexan
gulation has bf energy in thnergy stored in
is not depmplements a gorithm in orde PV field It ached Indeed
bsorb the eneretwork is not ully charged)gulation but
mplemented Inharging actione state of charCC are depenystems respecefined as a fune optimal stat
uring the nightilure of the
overed by the e loads duringIn order to aey are close t
tuned to thupercapacitorsoltage) In thiey are far fromIn order to ac
dvisable that thossible feedinrid Moreoveruring transientork at time p
minutes The serformances oandwidths of tf different bane different devAs shown beledback chain
andwidth for aalues for the rend the closednits
A Fr
The FEC is ae FEC a phafter locking thn the rotating omponents of re implementend the switchixternal one hand the direct an
been committehe AC networkn the storage spendent on w
maximum pder to draw thshuts down if
d this happensrgy produced available and
The ZC ain order to enn particular i
n the droop isrge Thus the ndent on the ctively An opnction of the gte of charge fttime and loweAC network
PV field whig the nighttimeamplify the acto the optimalhe supercaparsquo energy (pros way their e
m the optimal chieve the opthe power suppng only the mr the supercts of a few seperiods ranginsplit of the pof each devicthe convertersndwidths impvices low each conand PI regula
a power unit egulator const
d loop transfe
ront end conve
a traditional acase lock loop he grid phasereference conthe current [2
ed The internaing componen
as the DC voltnd quadrature
ed to the FEC k is consistensystems and uweather condpower point e maximum pf a maximum s only if the sby the PV fi
d all of the sand SCC connsure stabilityin order to alss realized by voltage refereSoC values fptimal state ogoal of the netfor the batterer during the d
k during the le the PVC cae ction of the sul SoC the SCacitorsrsquo voltaoportional to energy contribstate of chargtimal use of thplied by the FE
mean power recapacitors shoeconds while ng from a fewpower accordie can be obtasrsquo controllers
plies different
nverter is contrator In order tit is necessarytants It is theer functions f
erter
ctive front end(PLL) algorith the control o
ntrolling the di25] Two nestal one uses thnt control signtage and react currents as ou
because the antly greater thunlike the PVditions The
tracking (Mpossible energy
voltage threshsystem is not aield (ie if thstorage systemntribute to vy a droop conso perform thtaking into ac
ences for the Zfor the two sof charge has twork For exaies could be daytime becau
daytime couannot give ene
upercapacitorsCC voltage ref
ge instead othe square o
bution is low e he devices it EC be the smoequest from thould supply the batteries sw seconds toing to the dyained based o Indeed the response tim
rolled by meanto obtain the dy to choose suerefore advisafor the three
d In order to chm is implemof the FEC opirect and quadted feedback
he currents as nals as outputive power as utputs In relat
amount han the V field
PVC MPPT) y from hold is able to he AC ms are voltage ntrol is e self-ccount
ZC and storage
to be ample higher
use the uld be ergy to
s when ference of the of the when
is also oothest he AC power should some
ynamic on the choice
mes for
ns of a desired uitable able to power
control mented perates drature chains inputs ts the inputs tion to
the If thcapathe
Fig
T
R2
wki arcaparepr
Asquareguof ththe
2
I
V
d
DC
Fglobresp
A
Tbidisidewhimeanestin tcomgensche
Fig
Tvoltof dcharslowcan tran
DC voltage che internal chaable of settingFEC can be dr
7 Control schem
The closed loop
dp
pd
VksC
skV
3
32
where Vd is there the constanacitance of thresenting the p
As shown in Fare of the DC ulator is easierhe DC voltagegrid current
2
3
RCs
RVd
d
C
From (3) it bally as a lowponse of the FE
Zebra conver
The ZC is reirectional stepe is connectedile the low voans of an induted feedback the inductance
mponents whierating the reeme of the ext
8 Control schem
The self-rechatage referencedroop action enrge around thw if compared
therefore bensfer function
control only tain is much fag the desired rawn as report
me of FEC
p transfer func
id
i
VksV
Rks
32
e direct compots of the PI reghe DC bus apower absorbeFig 7 the feevoltage In thr In fact in the directly depe
1C
R
is clear that w-pass filter wEC
rter
ealized with p-downstep-ud to the DC boltage side is
uctor In additichains The ine acting on thile the externaeference currternal loop is r
me of ZC
arge function e in function onsures that thehe desired vald to the dyname neglected in
the external chaster than the ecurrents the ted in fig 7
ction is
dV
onent of the ggulator C is thand R is a reed by the DC bedback chain is way indeed
he frequency dends on the di
the control whose bandwid
one inverter up converter bus with an os connected toion the ZC is nternal one cohe duty cycleal one regulaterent as an oureported in fig
is obtained of the state ofe battery tries lue In any ca
mic action of thn formulating
hain is interesexternal one ancontrol schem
rid voltage kp
he total equivaesistive parambus is realized on
d the tuning odomain the sqirect compone
system operdth limits the
leg acting The high volutput capacitao the ZEBRAoperated with
ontrols the cue of the switces the DC vol
utput The cog 8
changing the f charge Thisto keep its staase this actio
he control loopg the closed
sting nd is
me of
(3)
p and alent
meter
n the of the quare ent of
(4)
rates time
as a ltage ance A by h two urrent ching ltage ntrol
DC s sort ate of on is p and loop
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
wicoth
acse
A
tostepe
fapofalothpo
Fig
B
opto cothhat=steloex
C
anlocodr13
ith step changonnected to the fast regulati
In order to chieving the abet and analyzed
i Pulsii Faul
is coiii Reg
Pulsing load
A pulsing mography (CAep load varyineriod of 25 s a
The power rcility Howev
ower of the cilities Thusad has been prat of a real CA
ower absorbed
g 2 Pulsing load
Fault in AC
In the secondpening of the A the AC distr
onverter detecte converter
appens at tim=1635 s Meaeps simulatingads on the D
xperimental tes
Regenerativ
Usually elecnd braking opads For this r
onducted by aprive for a woo9 s with an a5 kW) consta
ges of 1 kWhe DC bus by on of the pow
III CA
test the capabove discussedd sing load lt of the AC nonnected to theenerative load
d
load is reprAT) scanner wng between aband a duty cycrequested by tver it can be s
different dev the DCDC crogrammed toAT although
d by the load d
d simulating powe
C network
d test an AC fAC breaker coribution netwots the AC-sidefrom the AC
me t=47 s wanwhile the lg the connectiDC bus In fst is reported
ve load
ctrical drives perations are reason the tespplying the reodworking plaacceleration pant speed pha
while the otmeans of a D
wer absorbed
ASE STUDY
abilities of thd goals three
network while e DC bus
d
resented by whose typicalbout 60 kW a
cle of 70 this load is toscaled by taki
vices connecteconverter conno absorb a pow
opportunely sduring the tests
er profile of CAT
fault is simulaonnecting the ork The contre lack of volta
C network Twhile the AC load demand ion and disconfig 3 the po
alternating bethe most di
t to simulate real power demant [17] The phase (maximuase (power ab
ther 30 kW bDCDC conver
he DC netwokinds of case
a step variabl
a computer l power diagraand 180 kW w
oo high for thing into accoued to the DCnected 30 kW
wer profile simscaled In fig s is reported
ated with the mfront-end con
rol algorithm age and disco
The simulatedreconnection
changes withnnection of difower profile
etween acceleiffused regeneregenerative lo
mand of an eleworking cyclum power abssorbed 700 W
bank is rter for
ork in s were
le load
axial am is a with a
he test unt the C test
W rated milar to
2 the
manual nverter of the nnects
d fault n is at h some fferent of the
eration erative oads is ectrical e lasts sorbed
W) and
brakmeaThecan insedrivconnexp
Fig
Fig
Indiffelecof toveopeapp
Insign
Agavdev
Tdynpartelec
king phase (man power absoerefore it is cl
lead to a sertion and discve has been nected at timeerimental pow
3 Load power du
4 Power diagram
IV M
n order to coferent convertctrical storage the model to rr the full dynrative workinlications are
an energy futhe average a peak shavabsorbdeliva certain ref
n general tnificantly loweAmong the vae attention toice behavior a
A ZEB
The ZEBRA banamic modelsticular in [18]ctrochemical i
maximum poworbed during olear how the ssignificant enconnection of simulated In
e t=32s and dwer diagram is
uring AC network
m of simulated reg
MODELING OF E
orrectly set uers of the DCdevices is re
represent the namic range ofng conditions
function in whpower requestving function
ver the requestference averagthe first funer charge and arious models
lumped modas seen from th
BRA battery m
attery model cs available in] and [19] leaimpedance spe
wer regeneratone cycle is lostorage action nergy recoveranother load in particular disconnected as reported in fi
k fault test
generative load
ELECTRICAL ST
up the controC system thequired In parbehavior of thf utilization iss of a storag
hich the deviceted by the app
n in which thted peak pow
ge power nction is chadischarge dyn available in
dels capable ohe terminals
model
comes from thn the literatuad acid batteryectroscopy (E
ted 17 kW) ower than 200of the DC sy
ry Moreoverin parallel witha 9 kW loa
at time t=47sig 4
TORAGE
l strategy fore modeling ofrticular the abhe electric stos needed Thege device in
e has to guaraplication he storage haer above or b
aracterized bnamic
the literaturef interpolating
he lead acid baure [18]-[21]y models base
EIS) are descri
The 0 W
ystem the h the ad is The
r the f the bility orage e two
grid
antee
as to elow
by a
e we g the
attery ] In ed on ibed
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
whco
Fig
(SthR1
thac
coeq
E
So
chnocu
A
dymin[2id
Fig
hile in [20] aompromise bet
The model us
g 5 ZEBRA batt
The ZEBRA SoC) represene state of char1 which takese device and
ccount the dynIn addition
onsidered to bquations descri
ESoCE )( 0
C
oCn
11
where
E0 is the opharged SoC isominal batteryurrent supplied
Supercapac
A supercapaynamics of th
mHz up to 10troduced in [2
24] where adentification w
g 6 Supercapacit
In this model Ri represen
available f C(u)=C0
RR
and [21] thirdtween complesed for our pur
tery model
battery is modnting the openrge (SoC) in ss into accountd the parallelnamic behavior
for our purpbe constant duibing the batte
SoCKe 1
dtib
pen circuit vos the state of
y capacity Ke d by the batter
citor model
acitor modele device in th
00 Hz is repr23] was intega simplified
was presented
tor model
l nts the high from the manu+KVyuml u wh
C(u)Ri
1
222
u
2
C
C(u)Ri
1
222
u
2
C
d-order modelxity and precirpose is repres
deled as an elen circuit voltaseries with thet the high freq branchR2-C2
r of the batterypose the batturing the workery device are
oltage when tf charge of this a voltage co
ry
l capable ofhe typical freqresented in f
grated in a moprocedure f
frequency resufacturerrsquos dathere C0 and
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
ls that show aision are descrsented in fig 5
ectromotive foage as a functe internal resisquency resista2 which takey tery temperatk cycle so th
the battery ishe battery Cn
onstant and ib
f representinquency range fig 6 This more general mofor the para
sistance whictasheet
KV are con
)(
)
uC vsc
isc
)(
)
uC vsc
isc
a good ribed 5[22]
orce E tion of stance
ance of es into
ture is hat the
(1)
(2)
s fully is the
b is the
ng the of 10
model odel in ameter
ch was
nstants
TreprFor lowsign
Adefiachithe 1
2
34
Itconconand
Tthe Howprefcapaotheconbe aand worquathessuitmodstorautovaluhas at thvoltto cooconconothevolt
Tassibetwwouano
evaluated byin [24] τ(u)=3yuml(Rdc-Rresistance re
The supercaparesent the redour purposes
w frequencies nificant error
A control strained In ordeieving all ofcontrol strateg
The stabilizaat a steady sThe automaone or moreThe self-recThe optimal
t is evident ttrol has to bverters that ca
d PVC The possibility
use of only owever in ordeferable to havable of workier units The trol laws for tadapted to the
d have to be rking conditiolity on the DCse laws can btably tuned difications conrage system onomously itue The tuningto be coordina
he same timetage Thus theoptimally exprdinated withverters On thtrols of the foer because thtage and curre
The duty of imigned only toween the voluld cause a cother even in
y a constant c
Ri)yumlC(u) wheported in the acitor and battistribution and
s these phenoalmost close t
V CONTRO
ategy for the er to make the aims discgy should ensuation of the D
state for differeatic configurabe devices are uharge of the stl utilization of that to achiebe implemenan regulate th
y of splitting one controller
er to realize a mve separate coing independeproposed con
the four convedifferent devintegrated in
ons for the difC bus As willbe obtained uand modifie
nsists in the shas to im
ts own state g of the dynamated to make e to guarantee e dynamic respploit the conh the dynamhe contrary frour converters hey are basedent at their conmposing a stao one converltages measurontinuous pown the absenc
current charge
here Rdc repmanufacturertery models ud self-discharmena which to DC can be
OL STRATEGY
network conthe DC net
cussed in the ure
DC voltage durent kind of loability of the
unavailable torage system
f all of the devve the first g
nted Howevehe DC voltage
the voltage rr or of coordmore robust coontrollers for tently from thntrol strategy erters The coices they conn
order to ensfferent systeml be shown in using traditioed One ofself-recharge f
mplement in of charge ar
mic responseseach device wthe stabilizati
ponse of each ntrolled devic
mic performanrom an operati
are independd only on thennection nodeationary DC brter otherwisred at the diwer flow frome of loads
e test as descr
presents the rsquos datasheet used are unabrge of the devtake place at neglected wit
nverters has twork capableprevious sect
ring transientsads control schem
ms vices goal DC volr there are e FEC ZC S
regulation requinated controlontrol it woulthe different u
he presence ofis to design
ontrol laws havnect to the DCsure both opt
ms and high pothe next Sect
onal control lf the innovafunction that
order to kround the des of the conve
work at its best ion of the DCconverter is tu
ce but it is nces of the oion viewpoint
dent one from e measuremen
bus voltage case the differfferent conve
m one convertThe stiff vol
ribed
DC
le to vices
very thout
been e of tions
s and
me if
ltage four
SCC
uires llers ld be units f the four
ve to C bus timal ower ions
loops ative each
keep sired
erters and
C bus uned also
other t the each
nt of
an be rence erters er to ltage
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
reofenit imalgthreabnefureimchthSCsydethdufacoth
this suvoth
adpogrduwompebaofth
febavafinun
thAfoncoaranexan
gulation has bf energy in thnergy stored in
is not depmplements a gorithm in orde PV field It ached Indeed
bsorb the eneretwork is not ully charged)gulation but
mplemented Inharging actione state of charCC are depenystems respecefined as a fune optimal stat
uring the nightilure of the
overed by the e loads duringIn order to aey are close t
tuned to thupercapacitorsoltage) In thiey are far fromIn order to ac
dvisable that thossible feedinrid Moreoveruring transientork at time p
minutes The serformances oandwidths of tf different bane different devAs shown beledback chain
andwidth for aalues for the rend the closednits
A Fr
The FEC is ae FEC a phafter locking thn the rotating omponents of re implementend the switchixternal one hand the direct an
been committehe AC networkn the storage spendent on w
maximum pder to draw thshuts down if
d this happensrgy produced available and
The ZC ain order to enn particular i
n the droop isrge Thus the ndent on the ctively An opnction of the gte of charge fttime and loweAC network
PV field whig the nighttimeamplify the acto the optimalhe supercaparsquo energy (pros way their e
m the optimal chieve the opthe power suppng only the mr the supercts of a few seperiods ranginsplit of the pof each devicthe convertersndwidths impvices low each conand PI regula
a power unit egulator const
d loop transfe
ront end conve
a traditional acase lock loop he grid phasereference conthe current [2
ed The internaing componen
as the DC voltnd quadrature
ed to the FEC k is consistensystems and uweather condpower point e maximum pf a maximum s only if the sby the PV fi
d all of the sand SCC connsure stabilityin order to alss realized by voltage refereSoC values fptimal state ogoal of the netfor the batterer during the d
k during the le the PVC cae ction of the sul SoC the SCacitorsrsquo voltaoportional to energy contribstate of chargtimal use of thplied by the FE
mean power recapacitors shoeconds while ng from a fewpower accordie can be obtasrsquo controllers
plies different
nverter is contrator In order tit is necessarytants It is theer functions f
erter
ctive front end(PLL) algorith the control o
ntrolling the di25] Two nestal one uses thnt control signtage and react currents as ou
because the antly greater thunlike the PVditions The
tracking (Mpossible energy
voltage threshsystem is not aield (ie if thstorage systemntribute to vy a droop conso perform thtaking into ac
ences for the Zfor the two sof charge has twork For exaies could be daytime becau
daytime couannot give ene
upercapacitorsCC voltage ref
ge instead othe square o
bution is low e he devices it EC be the smoequest from thould supply the batteries sw seconds toing to the dyained based o Indeed the response tim
rolled by meanto obtain the dy to choose suerefore advisafor the three
d In order to chm is implemof the FEC opirect and quadted feedback
he currents as nals as outputive power as utputs In relat
amount han the V field
PVC MPPT) y from hold is able to he AC ms are voltage ntrol is e self-ccount
ZC and storage
to be ample higher
use the uld be ergy to
s when ference of the of the when
is also oothest he AC power should some
ynamic on the choice
mes for
ns of a desired uitable able to power
control mented perates drature chains inputs ts the inputs tion to
the If thcapathe
Fig
T
R2
wki arcaparepr
Asquareguof ththe
2
I
V
d
DC
Fglobresp
A
Tbidisidewhimeanestin tcomgensche
Fig
Tvoltof dcharslowcan tran
DC voltage che internal chaable of settingFEC can be dr
7 Control schem
The closed loop
dp
pd
VksC
skV
3
32
where Vd is there the constanacitance of thresenting the p
As shown in Fare of the DC ulator is easierhe DC voltagegrid current
2
3
RCs
RVd
d
C
From (3) it bally as a lowponse of the FE
Zebra conver
The ZC is reirectional stepe is connectedile the low voans of an induted feedback the inductance
mponents whierating the reeme of the ext
8 Control schem
The self-rechatage referencedroop action enrge around thw if compared
therefore bensfer function
control only tain is much fag the desired rawn as report
me of FEC
p transfer func
id
i
VksV
Rks
32
e direct compots of the PI reghe DC bus apower absorbeFig 7 the feevoltage In thr In fact in the directly depe
1C
R
is clear that w-pass filter wEC
rter
ealized with p-downstep-ud to the DC boltage side is
uctor In additichains The ine acting on thile the externaeference currternal loop is r
me of ZC
arge function e in function onsures that thehe desired vald to the dyname neglected in
the external chaster than the ecurrents the ted in fig 7
ction is
dV
onent of the ggulator C is thand R is a reed by the DC bedback chain is way indeed
he frequency dends on the di
the control whose bandwid
one inverter up converter bus with an os connected toion the ZC is nternal one cohe duty cycleal one regulaterent as an oureported in fig
is obtained of the state ofe battery tries lue In any ca
mic action of thn formulating
hain is interesexternal one ancontrol schem
rid voltage kp
he total equivaesistive parambus is realized on
d the tuning odomain the sqirect compone
system operdth limits the
leg acting The high volutput capacitao the ZEBRAoperated with
ontrols the cue of the switces the DC vol
utput The cog 8
changing the f charge Thisto keep its staase this actio
he control loopg the closed
sting nd is
me of
(3)
p and alent
meter
n the of the quare ent of
(4)
rates time
as a ltage ance A by h two urrent ching ltage ntrol
DC s sort ate of on is p and loop
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
whco
Fig
(SthR1
thac
coeq
E
So
chnocu
A
dymin[2id
Fig
hile in [20] aompromise bet
The model us
g 5 ZEBRA batt
The ZEBRA SoC) represene state of char1 which takese device and
ccount the dynIn addition
onsidered to bquations descri
ESoCE )( 0
C
oCn
11
where
E0 is the opharged SoC isominal batteryurrent supplied
Supercapac
A supercapaynamics of th
mHz up to 10troduced in [2
24] where adentification w
g 6 Supercapacit
In this model Ri represen
available f C(u)=C0
RR
and [21] thirdtween complesed for our pur
tery model
battery is modnting the openrge (SoC) in ss into accountd the parallelnamic behavior
for our purpbe constant duibing the batte
SoCKe 1
dtib
pen circuit vos the state of
y capacity Ke d by the batter
citor model
acitor modele device in th
00 Hz is repr23] was intega simplified
was presented
tor model
l nts the high from the manu+KVyuml u wh
C(u)Ri
1
222
u
2
C
C(u)Ri
1
222
u
2
C
d-order modelxity and precirpose is repres
deled as an elen circuit voltaseries with thet the high freq branchR2-C2
r of the batterypose the batturing the workery device are
oltage when tf charge of this a voltage co
ry
l capable ofhe typical freqresented in f
grated in a moprocedure f
frequency resufacturerrsquos dathere C0 and
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
(222 Cn
u
)(
)(2 uC
u
2
)(u
2
)(uC
ls that show aision are descrsented in fig 5
ectromotive foage as a functe internal resisquency resista2 which takey tery temperatk cycle so th
the battery ishe battery Cn
onstant and ib
f representinquency range fig 6 This more general mofor the para
sistance whictasheet
KV are con
)(
)
uC vsc
isc
)(
)
uC vsc
isc
a good ribed 5[22]
orce E tion of stance
ance of es into
ture is hat the
(1)
(2)
s fully is the
b is the
ng the of 10
model odel in ameter
ch was
nstants
TreprFor lowsign
Adefiachithe 1
2
34
Itconconand
Tthe Howprefcapaotheconbe aand worquathessuitmodstorautovaluhas at thvoltto cooconconothevolt
Tassibetwwouano
evaluated byin [24] τ(u)=3yuml(Rdc-Rresistance re
The supercaparesent the redour purposes
w frequencies nificant error
A control strained In ordeieving all ofcontrol strateg
The stabilizaat a steady sThe automaone or moreThe self-recThe optimal
t is evident ttrol has to bverters that ca
d PVC The possibility
use of only owever in ordeferable to havable of workier units The trol laws for tadapted to the
d have to be rking conditiolity on the DCse laws can btably tuned difications conrage system onomously itue The tuningto be coordina
he same timetage Thus theoptimally exprdinated withverters On thtrols of the foer because thtage and curre
The duty of imigned only toween the voluld cause a cother even in
y a constant c
Ri)yumlC(u) wheported in the acitor and battistribution and
s these phenoalmost close t
V CONTRO
ategy for the er to make the aims discgy should ensuation of the D
state for differeatic configurabe devices are uharge of the stl utilization of that to achiebe implemenan regulate th
y of splitting one controller
er to realize a mve separate coing independeproposed con
the four convedifferent devintegrated in
ons for the difC bus As willbe obtained uand modifie
nsists in the shas to im
ts own state g of the dynamated to make e to guarantee e dynamic respploit the conh the dynamhe contrary frour converters hey are basedent at their conmposing a stao one converltages measurontinuous pown the absenc
current charge
here Rdc repmanufacturertery models ud self-discharmena which to DC can be
OL STRATEGY
network conthe DC net
cussed in the ure
DC voltage durent kind of loability of the
unavailable torage system
f all of the devve the first g
nted Howevehe DC voltage
the voltage rr or of coordmore robust coontrollers for tently from thntrol strategy erters The coices they conn
order to ensfferent systeml be shown in using traditioed One ofself-recharge f
mplement in of charge ar
mic responseseach device wthe stabilizati
ponse of each ntrolled devic
mic performanrom an operati
are independd only on thennection nodeationary DC brter otherwisred at the diwer flow frome of loads
e test as descr
presents the rsquos datasheet used are unabrge of the devtake place at neglected wit
nverters has twork capableprevious sect
ring transientsads control schem
ms vices goal DC volr there are e FEC ZC S
regulation requinated controlontrol it woulthe different u
he presence ofis to design
ontrol laws havnect to the DCsure both opt
ms and high pothe next Sect
onal control lf the innovafunction that
order to kround the des of the conve
work at its best ion of the DCconverter is tu
ce but it is nces of the oion viewpoint
dent one from e measuremen
bus voltage case the differfferent conve
m one convertThe stiff vol
ribed
DC
le to vices
very thout
been e of tions
s and
me if
ltage four
SCC
uires llers ld be units f the four
ve to C bus timal ower ions
loops ative each
keep sired
erters and
C bus uned also
other t the each
nt of
an be rence erters er to ltage
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
reofenit imalgthreabnefureimchthSCsydethdufacoth
this suvoth
adpogrduwompebaofth
febavafinun
thAfoncoaranexan
gulation has bf energy in thnergy stored in
is not depmplements a gorithm in orde PV field It ached Indeed
bsorb the eneretwork is not ully charged)gulation but
mplemented Inharging actione state of charCC are depenystems respecefined as a fune optimal stat
uring the nightilure of the
overed by the e loads duringIn order to aey are close t
tuned to thupercapacitorsoltage) In thiey are far fromIn order to ac
dvisable that thossible feedinrid Moreoveruring transientork at time p
minutes The serformances oandwidths of tf different bane different devAs shown beledback chain
andwidth for aalues for the rend the closednits
A Fr
The FEC is ae FEC a phafter locking thn the rotating omponents of re implementend the switchixternal one hand the direct an
been committehe AC networkn the storage spendent on w
maximum pder to draw thshuts down if
d this happensrgy produced available and
The ZC ain order to enn particular i
n the droop isrge Thus the ndent on the ctively An opnction of the gte of charge fttime and loweAC network
PV field whig the nighttimeamplify the acto the optimalhe supercaparsquo energy (pros way their e
m the optimal chieve the opthe power suppng only the mr the supercts of a few seperiods ranginsplit of the pof each devicthe convertersndwidths impvices low each conand PI regula
a power unit egulator const
d loop transfe
ront end conve
a traditional acase lock loop he grid phasereference conthe current [2
ed The internaing componen
as the DC voltnd quadrature
ed to the FEC k is consistensystems and uweather condpower point e maximum pf a maximum s only if the sby the PV fi
d all of the sand SCC connsure stabilityin order to alss realized by voltage refereSoC values fptimal state ogoal of the netfor the batterer during the d
k during the le the PVC cae ction of the sul SoC the SCacitorsrsquo voltaoportional to energy contribstate of chargtimal use of thplied by the FE
mean power recapacitors shoeconds while ng from a fewpower accordie can be obtasrsquo controllers
plies different
nverter is contrator In order tit is necessarytants It is theer functions f
erter
ctive front end(PLL) algorith the control o
ntrolling the di25] Two nestal one uses thnt control signtage and react currents as ou
because the antly greater thunlike the PVditions The
tracking (Mpossible energy
voltage threshsystem is not aield (ie if thstorage systemntribute to vy a droop conso perform thtaking into ac
ences for the Zfor the two sof charge has twork For exaies could be daytime becau
daytime couannot give ene
upercapacitorsCC voltage ref
ge instead othe square o
bution is low e he devices it EC be the smoequest from thould supply the batteries sw seconds toing to the dyained based o Indeed the response tim
rolled by meanto obtain the dy to choose suerefore advisafor the three
d In order to chm is implemof the FEC opirect and quadted feedback
he currents as nals as outputive power as utputs In relat
amount han the V field
PVC MPPT) y from hold is able to he AC ms are voltage ntrol is e self-ccount
ZC and storage
to be ample higher
use the uld be ergy to
s when ference of the of the when
is also oothest he AC power should some
ynamic on the choice
mes for
ns of a desired uitable able to power
control mented perates drature chains inputs ts the inputs tion to
the If thcapathe
Fig
T
R2
wki arcaparepr
Asquareguof ththe
2
I
V
d
DC
Fglobresp
A
Tbidisidewhimeanestin tcomgensche
Fig
Tvoltof dcharslowcan tran
DC voltage che internal chaable of settingFEC can be dr
7 Control schem
The closed loop
dp
pd
VksC
skV
3
32
where Vd is there the constanacitance of thresenting the p
As shown in Fare of the DC ulator is easierhe DC voltagegrid current
2
3
RCs
RVd
d
C
From (3) it bally as a lowponse of the FE
Zebra conver
The ZC is reirectional stepe is connectedile the low voans of an induted feedback the inductance
mponents whierating the reeme of the ext
8 Control schem
The self-rechatage referencedroop action enrge around thw if compared
therefore bensfer function
control only tain is much fag the desired rawn as report
me of FEC
p transfer func
id
i
VksV
Rks
32
e direct compots of the PI reghe DC bus apower absorbeFig 7 the feevoltage In thr In fact in the directly depe
1C
R
is clear that w-pass filter wEC
rter
ealized with p-downstep-ud to the DC boltage side is
uctor In additichains The ine acting on thile the externaeference currternal loop is r
me of ZC
arge function e in function onsures that thehe desired vald to the dyname neglected in
the external chaster than the ecurrents the ted in fig 7
ction is
dV
onent of the ggulator C is thand R is a reed by the DC bedback chain is way indeed
he frequency dends on the di
the control whose bandwid
one inverter up converter bus with an os connected toion the ZC is nternal one cohe duty cycleal one regulaterent as an oureported in fig
is obtained of the state ofe battery tries lue In any ca
mic action of thn formulating
hain is interesexternal one ancontrol schem
rid voltage kp
he total equivaesistive parambus is realized on
d the tuning odomain the sqirect compone
system operdth limits the
leg acting The high volutput capacitao the ZEBRAoperated with
ontrols the cue of the switces the DC vol
utput The cog 8
changing the f charge Thisto keep its staase this actio
he control loopg the closed
sting nd is
me of
(3)
p and alent
meter
n the of the quare ent of
(4)
rates time
as a ltage ance A by h two urrent ching ltage ntrol
DC s sort ate of on is p and loop
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
reofenit imalgthreabnefureimchthSCsydethdufacoth
this suvoth
adpogrduwompebaofth
febavafinun
thAfoncoaranexan
gulation has bf energy in thnergy stored in
is not depmplements a gorithm in orde PV field It ached Indeed
bsorb the eneretwork is not ully charged)gulation but
mplemented Inharging actione state of charCC are depenystems respecefined as a fune optimal stat
uring the nightilure of the
overed by the e loads duringIn order to aey are close t
tuned to thupercapacitorsoltage) In thiey are far fromIn order to ac
dvisable that thossible feedinrid Moreoveruring transientork at time p
minutes The serformances oandwidths of tf different bane different devAs shown beledback chain
andwidth for aalues for the rend the closednits
A Fr
The FEC is ae FEC a phafter locking thn the rotating omponents of re implementend the switchixternal one hand the direct an
been committehe AC networkn the storage spendent on w
maximum pder to draw thshuts down if
d this happensrgy produced available and
The ZC ain order to enn particular i
n the droop isrge Thus the ndent on the ctively An opnction of the gte of charge fttime and loweAC network
PV field whig the nighttimeamplify the acto the optimalhe supercaparsquo energy (pros way their e
m the optimal chieve the opthe power suppng only the mr the supercts of a few seperiods ranginsplit of the pof each devicthe convertersndwidths impvices low each conand PI regula
a power unit egulator const
d loop transfe
ront end conve
a traditional acase lock loop he grid phasereference conthe current [2
ed The internaing componen
as the DC voltnd quadrature
ed to the FEC k is consistensystems and uweather condpower point e maximum pf a maximum s only if the sby the PV fi
d all of the sand SCC connsure stabilityin order to alss realized by voltage refereSoC values fptimal state ogoal of the netfor the batterer during the d
k during the le the PVC cae ction of the sul SoC the SCacitorsrsquo voltaoportional to energy contribstate of chargtimal use of thplied by the FE
mean power recapacitors shoeconds while ng from a fewpower accordie can be obtasrsquo controllers
plies different
nverter is contrator In order tit is necessarytants It is theer functions f
erter
ctive front end(PLL) algorith the control o
ntrolling the di25] Two nestal one uses thnt control signtage and react currents as ou
because the antly greater thunlike the PVditions The
tracking (Mpossible energy
voltage threshsystem is not aield (ie if thstorage systemntribute to vy a droop conso perform thtaking into ac
ences for the Zfor the two sof charge has twork For exaies could be daytime becau
daytime couannot give ene
upercapacitorsCC voltage ref
ge instead othe square o
bution is low e he devices it EC be the smoequest from thould supply the batteries sw seconds toing to the dyained based o Indeed the response tim
rolled by meanto obtain the dy to choose suerefore advisafor the three
d In order to chm is implemof the FEC opirect and quadted feedback
he currents as nals as outputive power as utputs In relat
amount han the V field
PVC MPPT) y from hold is able to he AC ms are voltage ntrol is e self-ccount
ZC and storage
to be ample higher
use the uld be ergy to
s when ference of the of the when
is also oothest he AC power should some
ynamic on the choice
mes for
ns of a desired uitable able to power
control mented perates drature chains inputs ts the inputs tion to
the If thcapathe
Fig
T
R2
wki arcaparepr
Asquareguof ththe
2
I
V
d
DC
Fglobresp
A
Tbidisidewhimeanestin tcomgensche
Fig
Tvoltof dcharslowcan tran
DC voltage che internal chaable of settingFEC can be dr
7 Control schem
The closed loop
dp
pd
VksC
skV
3
32
where Vd is there the constanacitance of thresenting the p
As shown in Fare of the DC ulator is easierhe DC voltagegrid current
2
3
RCs
RVd
d
C
From (3) it bally as a lowponse of the FE
Zebra conver
The ZC is reirectional stepe is connectedile the low voans of an induted feedback the inductance
mponents whierating the reeme of the ext
8 Control schem
The self-rechatage referencedroop action enrge around thw if compared
therefore bensfer function
control only tain is much fag the desired rawn as report
me of FEC
p transfer func
id
i
VksV
Rks
32
e direct compots of the PI reghe DC bus apower absorbeFig 7 the feevoltage In thr In fact in the directly depe
1C
R
is clear that w-pass filter wEC
rter
ealized with p-downstep-ud to the DC boltage side is
uctor In additichains The ine acting on thile the externaeference currternal loop is r
me of ZC
arge function e in function onsures that thehe desired vald to the dyname neglected in
the external chaster than the ecurrents the ted in fig 7
ction is
dV
onent of the ggulator C is thand R is a reed by the DC bedback chain is way indeed
he frequency dends on the di
the control whose bandwid
one inverter up converter bus with an os connected toion the ZC is nternal one cohe duty cycleal one regulaterent as an oureported in fig
is obtained of the state ofe battery tries lue In any ca
mic action of thn formulating
hain is interesexternal one ancontrol schem
rid voltage kp
he total equivaesistive parambus is realized on
d the tuning odomain the sqirect compone
system operdth limits the
leg acting The high volutput capacitao the ZEBRAoperated with
ontrols the cue of the switces the DC vol
utput The cog 8
changing the f charge Thisto keep its staase this actio
he control loopg the closed
sting nd is
me of
(3)
p and alent
meter
n the of the quare ent of
(4)
rates time
as a ltage ance A by h two urrent ching ltage ntrol
DC s sort ate of on is p and loop
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
R
B
ThouthSCchcyreoufigto
Fig
vososuAnis an
R
nufohachre
cousanthcoDC
k
00
bp
pb
VksRC
skV
2
where Vb is th
Supercapac
The SCC is rhe high voltautput capacitane supercapaciCC is operatedhain controls tycle of the swgulates the DC
utput The cong 9 It differs the feedback
g 9 Control sche
The self-recholtage referencort of droop upercapacitors nyway if kSC slow if comp
nd the closed l
Sp
pSC
VksRC
kV
2
where VSC is
VI NU
In order to umerical and eollowing somave been choshanging the tusults for the dIn order to ac
ontrollers of thsing the paramnd (6) the spee load For
onverters chanC bus as repo
TABLE
FEC
kp ki k
001 0002 00
bib
i
Vks
Rk
1
he battery volt
citors converte
ealized using age side is conce while theitors by meansd with two nesthe current in witching comC voltage genentrol scheme o from the conon the superc
eme of SCC
harge functionce in function
action is instate of ch
is small enoupared to the dyoop transfer fu
SiSC
i
Vks
Rks
1
the supercapa
UMERICAL AND
test the prexperimental te results will sen to show uning of the
different applicchieve the requhe converters
meters reportedeed response o
this reasonnge as a functorted in fig 10
I PARAMETERS O
ZC
kp ki kb
005 001 05
tage
er
the same confonnected to the low voltage s of an inductsted feedback the inductance
mponents whilerating the refof the externalntrol scheme oapacitorsrsquo vol
n is obtainedn of supercapantroduce in arge around
ugh the actionynamic action function is
SC
acitorsrsquo voltage
D EXPERIMENT
roposed contrtests have beebe reported
how the samcontrollers c
cations discussuired bandwidthe PI regulat
d in Table I Aof the control the bandwition of the po0
OF CONVERTER RE
b SOCref
[] kp
5 90 001
figuration as thhe DC bus wside is connector In additionchains The ine acting on thle the externaference currenl loop is repor
of the ZC in retage
d changing thacitors voltageorder to keethe desired
n of the droopof the control
e
TAL RESULTS
rol strategy en performed
In particularme network wcan give verysed in Sectiondth separation tors have been
According to (3answer depen
idths of the ower supplied
EGULATORS
SC
ki kSC
1 01 0143
(5)
he ZC with an cted to n the nternal he duty al one
nt as an rted in elation
he DC e This ep the value
p chain l loop
(6)
many In the
these without y good n III
of the n tuned 3) (5) nds on
three to the
VSCref [V]
220
Fig
Afig somwhisecoaccoand the varichoTheconbeenof appan ais wthe reasquic
Inload
Inand The
Insimu
A
Tto drequThegridand
10 Bandwidths v
According to t10 the actio
me hundreds oile the interveonds The coount the maxi
d of the SoC obatteries con
iation of 10sen maximum
e desired SoCcerns the supn set to 220 V110 V (in
roximately to admissible DCworth noting tSCC has been
son why the cker than that n the followinds presented inn particular t
d experimentaley are summar
in the first tesystem A between aboin the secopower requcase a netwalmost lasts in the third to the systet1=32 s and t
n the followmulations and e
Pulsing loads
The pulsing lodevelop the nuested by the e test lasts 3 md is simulated d 11b the sim
vs power of three
the bandwidthon of the SCCof millisecondention of the onstants kb animum desirabof the two stonsidering a m equal to 3m depth of disC has been chpercapacitors V Considering
fact half frac34 of the stor
C voltage variathat the admisn chosen lower
supercapacitof the batterie
ng the results n Section III athree cases hally tested All rized in Table est the pulsindisconnection
out minutes 1 nd test the rest reported iwork failure for all three mtest the regenem A step lt2 = 46 s wing the reexperimental te
s
ad of fig 2 isnumerical simload is used i
min during wbetween abou
mulated and
e converters
hs of the convC will have ads the ZC wiFEC will be nd kSC are chble variation oorage devices
maximum allow38V in correscharge (76)hosen equal t
the desired vg the possibleof the volt
red energy) anation of 4 it ssible DC volr than that of ttors action ises obtained by
are analyzed ave been num
of these testsII and reporte
ng load of fig n of the AC and 2 response of tin fig 3 is poccurs at the
minutes nerative load oload variation
esults obtaineests are report
s applied to thmulations thein simulating hich a disconnut minutes 1 aexperimental
verters reportea time constanill act in secoslowed to tenhosen taking of the DC vol In particularwable DC volespondence of) it results kb=to 90 For wvoltage valuee voltage variatage corresp
nd associating t results kSC=1ltage variationthe ZC This is expected to
applying the t
merically simuls lasted 3 mined below 2 is applied togrid is simul
the system toperformed In e first instant
of fig 4 is appn occurs betw
ed in numeted and discus
he system In o measured pothe load behanection of theand 2 In figs
DC voltages
ed in nt of onds ns of
into ltage r for ltage f the =05 what
e has ation onds it to
7 It n for is the o be
three
lated nutes
o the lated
o the this and
plied ween
erical sed
order ower
avior e AC 11a s are
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
rep
C
P
F
Re
ACThexsim
diad
subasuthAfsusu
ported
TA
Case study
Pulsing load Pm
T =
Fault of AC network
ste
egenerative load
Fig 11 DC vo
The voltage eC grid is dische power suppxperimental remulations Even if a cenfferent conver
dapting it to thIndeed as i
upplied essenandwidth Thupplies only the first instantfter some sec
upplies the upercapacitors
ABLE II TEST CAS
Key feature
max = 19 kW Pmin = 25 s duty cycl
Pmax = 45 kWPmin = -45 kWTdisconn = 47
Treconn = 1635ep variations of D
Pmax =35 kWPmin = ndash17 kWPmean = 200 W
a
b
oltages
excursion is vconnected theplied by each esults match
ntralized contrrters are able
he characteristiis clear fromntially by thehe FEC prehe mean powts the SCC sconds as allo
mean pow
SES ANALYZED IN
es
= 8 kWe = 66
AC gr
Dreconn
iW W s 5 s
DC loads W W W
a 9 kW
a) simulated
b) measured
very limited (plusmne voltage dropdevice is repothose obtaine
rol action is noto split the l
ics of the diffem fig 19 the
e SCC becaesenting the
wer At the grisupplies the wowed by its
wer and als
N PAPER Notes
rid is disconnectet = 1 s to t = 2 sDisconnection annection times are in experimental te
W load is connectet = 32 s to t = 47
plusmn2) and whp is lower thaorted in fig 12ed in the num
ot implementeload power dierent devicese pulsing powause of its
lower bandid disconnecti
whole load debandwidth thso recharges
ed from s nd
lagged est
ed from s
hen the an 3 2 The
merical
ed the iagram
wer is higher
dwidth ion in emand he ZC s the
Fig
Aenerreaspowdroofig havdecrreferechWhtwoSimdesikeep
B
Inqualoadexp
Tcausbattto thacti
HDC
Tdiff
12 Power supplie
At the grid rrgy from the son why the Fwer request op action intro8 Indeed dure been dischreased The d
erence and theharging proceen the SoC o
o DC referenmilarly the supired value of 2p the desired v
Fault of AC n
n order to sholity energy sud condition prerimentally re
The voltage flused by the cteries change the load their Don necessary t
However the cvoltage is alw
The powers eferent devices
a)
b)
ed by different de
reconnection DC bus to recFEC supplies This self-rechoduced in the ring the main harged and t
difference betwe DC referencess occurring f the batteries
nces are equapercapacitors 220 V Fig 13value of the su
network
ow the ability upply even wheesented in fig
ealized In fig uctuation arouchanging SoCtheir SoC absoDC voltage reto realize the schoice of a lo
ways inside a vexchanged beare reported in
simulated
measured
evices
the ZC autcharge the bata power grea
harge action icontrol of thegrid disconnetheir DC refween the ratedce of the ZC i
at the netws reaches the al and the se
keeps their v3 shows the abupercapacitors
of the systemen a main grid 3 is numeric 14 the DC vound the rated C of the batorbing from oeference is chaself-recharge aow value for kvery tight bandetween the Dn fig 15
tomatically drtteries This isater than the is obtained bye ZC and showction the battference has d value of theis the cause o
work reconnecdesired value
elf-recharge evoltage aroundbility of the SCsrsquo SoC
m to ensure ad faults occursally simulatedoltage is reporvalue of 380
tteries Whiler supplying po
anged by the daction kb ensures thad DC bus and
rains s the load
y the wn in teries been
e DC f the
ction e the ends d the CC to
high s the d and rted V is
e the ower
droop
at the
d the
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
Fig
Fig
powois fro
g 13 Supercapac
g 14 DC voltage
For the diffeower change orth noting tha load becau
om the DC bu
a
b
citorsrsquo voltages
a
b
e
ferent bandwidthe first devic
hat during the se it supplies
us
a) simulated
b) measured
a) simulated
b) measured
dths of the cce to take actiabsence of thits auxiliaries
controllers ation is the SCC
he AC grid ths by drawing
t each C It is e FEC power
Fig
C
Fsimuamopredrealworis re
InIt iesselowthe assoSCCcon
Itbefois lo
Tconnthe avaiencovoltshavprofas lobtamic
15 Simulated po
Regenerative
Finally the cmulated and ex
ount of energdominant actiolized by the hrk as the main eported n fig 17 the pis clear that entially by the
wer than 1 secoSCC (see f
ociated with tC keeps suptribution to tht is worth notiore the time inower than the r
The wide diffunected to distr
regulation ilability of ourages their tage stabilizativing and the files In this scleaves to an ain a high qcro-grid not on
a)
b)
wer supplied by d
loads
case of a regxperimentally gy is exchanon of the supehigher bandwpower source
powers supplithe alternativ
e supercapaciond where thefig 10) In the load step
pplying the ahe transient powing that the ZCnstant of 32 s breference one
VII CON
usion of distribribution netwoof distributionew techno
use in power ion and regulamatching of
cenario the usAC backbone
quality distribnly allows the
simulated
measured
different devices
generative loatested In th
nged and forercapacitors is
width of the Se In fig 16 th
ied by each deve power deitors because e predominantaddition durvariation (at
alternative power request
C drains powebecause the So
NCLUSIONS
buted renewaborks presents aon networksologies for systems to ac
ation continuigenerated and
se of DC micre is an effecbution networ
easier connec
ad is numerichis case no lr this reasons advisable ThSCC that makhe DC bus vol
evice are repoemand is covit occurs in tt answer is thring the trantime t=32 s)
ower beyond
er from the DCoC of the batt
ble energy soua new scenario Moreover storage sys
chieve the goaity of service d requested poo-grids connetive possibilit
rk Indeed action of renew
cally large the his is kes it ltage
orted vered times hat of nsient the
the
C bus teries
urces o for
the stems als of peak ower ected ty to
DC wable
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
soelidrACse
Fig
Fig
kinbe
ources and stoiminate all ofrives and suppC+DC distribu
ervice but also
g 16 DC voltage
g 17 Power supp
In order to cnd of distribu
een realized a
orage systemsf the input repliers Thus tution system ithe efficiency
a
b
es
a
b
plied by different
conduct some ution schemeat RSE in Mi
s but also mctifiers installthe advantageinvolve not ony of the overal
a) simulated
b) measured
a) simulated
b) measured
devices
preliminary e a 100-kW Dilan Italy Th
makes it possiled inside ele
es of using a hnly the qualityll system
experiments oDC test facilihis facility has
ible to ectrical hybrid of the
on this ty has s been
usedinteunit
Incona DpowstraconIn osimuconncarrbeenconand effecondynworstoreffe
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
d to set up aerconnecting dt storage systen this paper atrol strategy h
DC micro-gridwer quality levategy does noverter control
order to set up mulations were
nected to the nried out in ordn achieved Ttinuity of serv
d voltage staectively perfortrol of the c
namic responserk at its best rage systems wectively achiev
Hadjsaid N Caelectrical disresourcesPoweIEEE pp1-4 26
Soni A Ozvere42nd InternationUPEC 2007 pp
Song Yun-weiinspiration InEngineering (IC
Sarikprueck PDeveloping imPower and Ener
Carrasco JMGuisado RCPPower-ElectronEnergy SourceElectronics vol
Hopkins MDDistributed Renvol3 no2 pp1
Mamo X MallThe cornerstoneSociety General
Standard for IntSystems IEEE S
Reference technconsumers to tCompany CEI 0
Reference technto the LV electri
Baran ME Mopportunities aApplications v
a control stradevices to a Dems loads) after a descriphas been propod in terms of vel and energot need a cen
acts only on tthe proposed performed afnetwork Then
der to verify thThe reported evice when a abilization anrmed by the converters ales of the diffepotential Fin
was integratedved as shown
REFER
aire R Raison Bstribution systeramp Energy Soci6-30 July 2009
en CS Renewnal Universities 717-720 4-6 Sep
American pnternational ConfCMSE) 2010 pp
P Korkua SKmportant renewabgy Society Gener
Franquelo LP Prats MaAnic Systems for es A Survey53 no4 pp1002
Pahwa A Enewable Resource1047-1054 June 2
let S Coste T e for smart grid d Meeting 2009 P
terconnecting DisStandard 1547 20
nical rules for tthe HV and M0-16 fasc 9251 2
nical rules for theical Utilities CEI
Mahajan NR and challengesvol39 no6 pp 1
ategy for all DC bus (AC ne
tion of the tesosed to show
f the continuigy saving Thentral intelligenthe basis of thstrategy prel
fter modeling n some exper
hat all of the reexperimental rfault occurred
nd regeneratiDC micro-gr
llowed the derent units mnally the self-d into the contby the experim
RENCES B Decentralizetems with iety General Me
wable energy marPower Engineeri
pt 2007
policy on renewference on Man1513-1519 24-26
K Wei-Jen Leble energies in Tral Meeting pp1-
G BialasiewicM Leon JI
r the Grid Integ IEEE Transa2-1016 June 2006
Easton T Intes IEEE Transa2012
Grenard S Ddevelopment stratePES 09 IEEE p
stributed Resourc003
the connection oV electrical net
2008
e connection of acI 0-21 fasc 11666
DC distribution IEEE Trans
1596- 1601 Nov-
of the conveetwork genera
st facility a rothe capabilitiety of the sere proposed conce because he DC bus voliminary numeall of the dev
rimental tests wequested goalsresults showedd on the AC ive actions wrid Moreoverdecoupling of
making each de-recharging oftrol logic andmental results
ed operating moddistributed e
eeting 2009 PES
rket potential in Uing Conference
wable energy anagement Science
6 Nov 2010
ee Lumyong Thailand 2011 -8 24-29 July 201
cz JT GalvanMoreno-Alfonso
gration of Reneactions on Indu6
telligent Dispatcactions on Smart
Distribution automegy Power amp Ep1-6 26-30 July
es with Electric P
of active and patworks of distrib
ctive and passive6 2011
for industrial syssactions on Ind-Dec 2003
erters ation
obust es of
rvice ntrol each tage
erical vices were s had d the grid were the
f the evice f the
d was s
des for energy S 09
UK 2007
nd its e and
P IEEE
11
n E o N wable ustrial
h for Grid
mation Energy y 2009
Power
assive bution
users
stems dustry
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity
Copyright (c) 2013 IEEE Personal use is permitted For any other purposes permission must be obtained from the IEEE by emailing pubs-permissionsieeeorg
This article has been accepted for publication in a future issue of this journal but has not been fully edited Content may change prior to final publication
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
[24
[25
2] Kwasinski AStabilization Loads IEEE 834 March 20
3] Lago J Heldof a Power Coof DC ActiveElectronics vo
4] Xu CD ChAC versus DConference opp1-12 8-10
5] Musolino VTornelli C applications o14th Internati(ICHQP) 201
6] I Stadler ldquoStSmall DistribhttpwwwamStandards-for-University-GI
7] Musolino VQuality level International C2012 pp1-7
8] P Mauracherusing impedanJournal of Pow
9] H Anderssonimpedance daElectrochemis
0] N Moubayedidentification Specialists Co
1] MCeraolo ldquoNTransaction on
2] Einhorn M Cand ParameteApplicationspp1429-1437
3] S Buller Edynamic behaIndustry Appl
4] Musolino VSupercapacitoTransactions 2013
5] Piegari L Tsmart-grids fInternational C2010 pp1-7
A Onwuchekwof DC MicrogriTransactions on
011
dwein ML Oonverter for Curre Distribution Nol26 no3 pp87
eng KWE ADC distributed pon Power Electro
June 2011
Piegari L T Simulations
of LV DC distribional Conference0 pp1-6 26-29
tudy about Internbuted Generatormericadosolorgw-the-connection-oIZ-2011pdf
Piegari L TiroDC distribution
Conference on HaJune 2012
r E Karden ldquoDnce spectroscopy wer Sources 67 6
n I Petersson Eata for semi-bipolastry 31 1-11 (200
d J Kouta A EI-of the leadacid
onference 2008 P
New Dynamical n Power Systems
Conte F V Kraerization of Elec IEEE Transacti March 2013
E Karden D Koavior of supercaplications Conferen
Piegari L Tior Model With on Industrial Ele
Tricoli P A cofor providing uConference on Ha26-29 Sept 2010
S GrildquoLaureaengineepower Genoarespect
Currwith DE
wa CN Dids With InstantPower Electronic
Operation and Corent Balancing an
Networks IEEE 77-885 March 20
A survey of distrpower system
onics Systems an
Tironi E Marts and field tesbution network te on Harmonics Sept 2010
national Standardrs to the powewp-contentuploadof-Small-Distri-G
oni Tornelli C n system for woarmonics and Qu
Dynamic modellinfor parameter ide
69-84 (1997)
E Ahlberg ldquoMoar lead acid batte01)
-AIi H Dernaykabattery modelrdquo
PVSC 08 pp 16
Models of Lea VOL 15 NO 4
al C Fleig J ctrical Battery Mions on Power E
ok R W De Dpacitors using imnce Vol 4 2001
ironi E NewEasy Identifica
ectronics vol60
ontrol algorithm uninterruptible anarmonics and Qu
0
illo (Srsquo05ndashMardquo degree ering and thesystems from Genoa Italyively
rently he is AEIB Politecni
Dynamic Behavitaneous Constancs vol26 no3
ontrol-Oriented Mnd Stability Impro
Transactions on11
ributed power sy2011 4th Inter
nd Applications (
tini L Brambist results for pto reduce flicker
and Quality of
ds for the connecer gridrdquo Availds201107Intern
Generators-Cologn
Villa A Highoodworking plantuality of Power (I
ng of leadacid bentificationrdquo ELS
odelling electroceriesrdquo Journal of
a R Outbib ldquoPa33rd IEEE Phot (2008)
adAcid Batteriesrdquo4 November 2000
Comparison SeModels for AutoElectronics vol2
Doncker ldquo Modelmpedance spectro pp 2500 ndash 2504
w Full-Frequencyation Procedure no1 pp112-12
of power convencillary services
uality of Power (I
Mrsquo09) receivein elec
e PhD degrm the Univers in 2004 and
Assistant Proico di Milano
ior and nt-Power
pp822-
Modeling ovement n Power
stem mdash rnational (PESA)
illa A potential
effect f Power
ction of able at
national-ne-
h Power ts 15th CHQP)
batteries SEVIER
chemical Applied
arameter tovoltaic
rdquo IEEE 0
election omotive 8 no3
ling the oscopyrdquo 4
y-Range IEEE 20 Jan
erters in 14th CHQP)
ed the ctronic ree in sity of
2008
ofessor Italy
His conto pgen
syst
Italymacrene
powP
(CENatSys
Smaresothe
research intetrol techniquepower systemeration manag
tems in transp
y His currchines high-ewable energy
wer quality andProf Tironi iEI) Italian tional Researchtem
art Grids incluources informelectrical pow
erests includees and neural
ms (ie securigement)
V Musoreceived engineeriPolitecni
ActualDimac Rcompanyinterests devices managem
ortation and st
L Piegareceived electricalUniversitNaples respectiv
CurrenDEIB P
rent research-efficiency p
y sources and s
E TironElectricaPolitecni1972 heElectricadi Milano
At preDEIB Pof resear
d distributed gs a member Electrical Ash Council (CN
C TorneElectroniPolitecniHe workDistributof RSElocated several pcontrol oon techn
uding power emation and cower system and
smart grids networks andity assessment
olino was borthe PhD de
ing in 20ico di Milano lly he is a projRed Srl (a
y in Biassoinclude e
and power ement and intetationary appl
ri (Mrsquo04) wasthe MS and
l engineerity of Napl
Italy in 1vely ntly he is AssPolitecnico dih interests ipower electrstorage system
ni received thal Engineeriico di Milanoe joined theal Engineeringo esent he is olitecnico di rch include p
generation of Italian St
ssociation (ANR) group o
elli received tic Engineeringico di Milano (ks for the tion Technolo a National in Milan (Itaprojects abouof microgrids ologies for aclectronics for ommunicationd standards for
optimization d their applicat local and s
rn in 1982 anegree in elect012 from Milano Italyject coordinata private owono Italy) electrical stoelectronics foregrations of tlications
s born in 1975d PhD degreeng from les Federico 1999 and 2
sistant Professi Milano Minclude electronic conver
ms modeling
he MS degreing from Italy in 197e Department of the Politec
Full ProfessoMilano His a
power electro
tandard AuthAEI) and Itaof Electrical Po
the MS degreg in 1986 from(Italy) ldquoTransmission
ogies DepartmResearch Ce
aly) Involveut automation
he actually wctive networksdistributed en
n technologiesr interoperabil
n and ation small
nd he trical
the tor in wned
His orage r the these
5 He es in
the II
2003
sor at Milan
trical rters
ee in the
2 In t of cnico
or at areas
onics
hority alian ower
ee in m the
n amp mentrdquo entre d in and
works s and nergy s for lity