1

Abstract-- Stand alone wind-diesel power systems can be an

attractive solution for the supply of remotely located consumers or the provision of enhanced energy independence. This study presents the dynamic behavior of a hybrid power system containing a prior wind generator and a supplementary generator set driven by a combustion engine, thus independently and economically supplying electrical load demand for the main hospital in Gaza-Strip – Palestine. Accumulator bank and electrolyzer plus hydrogen storage tank are added as short term and long term storage devices – respectively – in order to balance power in the system. To analyze the behavior of the above mentioned power system, the measurements for wind speed were taken from a meteorological station located in the relevant coastal area, for relatively more than one year; correspondingly, the electrical load shape of the hospital had been recorded in high temporal resolution. Stored hydrogen is injected to the generator set under a prescribed scenario of operation. The simulation results prove that continual electricity supply at low emissions and high degree of energy independence is possible.

Index Terms-- battery-bank; generator set; hydrogen storage tank; renewable energy; stand-alone system; wind turbine.

I. INTRODUCTION ver the past few years, the total electricity needs have steadily been on the rise. Recent studies are expecting

that conventional sources of primary energy will be rare, if not exhausted, thus encouraging scientists and engineers to find alternative and sustainable sources of energy. Moreover, environmental aspects play an important role to limit the green house effect gases exhausted by the fossil fuelled power plants. To achieve this, many projects have been supported to develop technologies to exploit, as efficiently as possible, clean energy sources. The technology with the greatest impact in this area is the wind power generation. Such alternative sources of energy are not required to fully replace conventional types everywhere; rather, they are useful to be a supplementary factor besides traditional sources, especially in developing public utilities – hospitals and water pumps – and rural areas where strong conventional energy grids do not yet exist or some external restrictions prevent demands to be supported. In these places, many diesel generators are used to

M. Mushtaha and G. Krost are with the Department of Electrical

Engineering, University Duisburg-Essen, Duisburg 47057, Germany (e-mail:{mohammed.mushtaha,Gerhard.krost}@uni-due.de).

meet the load demands. Nevertheless, several problems occur such as the extra cost to transport the fuel to the target area.

Transportation system breakdown is another problem to support the continuous need of fuel to the electric power generator. Furthermore, continuing escalation in the fuel price make it cost effective to use the renewable energy sources. Thus, they can play an important role in the future.

Wind-diesel power systems are the most appropriate hybrid systems utilized in weak or isolated power systems. The first commercial application of High-Penetration, No Storage, Wind-Diesel (HPNSWD) technology was installed in 1999 by Northern Power System (Vermont, USA) on St. Paul Island, Alaska [1]. One of the main challenges with these types of power systems is the reliability and quality of the produced power [2]. As a matter of fact, fluctuations in the output power result from variations in the wind speed. One way to overcome this problem is to include different kinds of energy storage (e.g., accumulator batteries for small appliances, or generation and storage of hydrogen gas by electrolysis and pressure tank – respectively – and use of this gas for a combustion motor driven generator).

Storage technologies which are based on hydrogen production and utilization – that is expected to be used for very different applications – constitute some interesting advantages in terms of cost, independency and environmental effects [3].

Combination of short term and long term energy storage devices has many advantages. A battery-bank can be used to meet the peak power demands as well as eliminate the fluctuations to the input power of the electrolyzer; hence, the size of the electrolyzer can be minimized while its life-time expectancy will evolute. A large electrolyzer is inefficient to respond to the fast changes and high peak power excess [4]. Thus, the combined energy storage system can be operated more efficiently than a single hydrogen storage system.

The Near-East region – especially Palestine – was selected as the area of study. Availability of appreciable wind speeds and solar radiations drew the attention of engineers to use the clean sources of energy to face the continuous lack of the produced power. Incredible increase of populations together with growth of societal development forced the responsibles to think in detached power systems. Unstable political situations and the age of the local power grid are also strong reasons

Performance Study of Self-sufficient and Renewables Based Electricity Supply of a

Hospital in the Near East Region M. Mushtaha, and G. Krost, Member IEEE

O

978-1-4673-2729-9/12/$31.00 ©2012 IEEE

staindapareprosouGahomo

dieapdeforstofolhysecdiethesimpodissecsum

sycoindenas acanresimtheStostralsma

A. Asof sypo

Fig

anding behinddependent from

pplications beceas within thesoved to be curces, i.e. a daza-Strip, Palesospital is locateore than 40% oThe objective

esel hybrid ppplication. Theemand, as wellr storing and orages. To achllowed by a

ydrogen – resction, the planesel power syse mechanism omulated modelossible real apscussion of soction V. Lastlymmary, and ou

A typical stanstem is illustr

ontained in the duction gener

ngine driven sywell as su

cumulator bannd storage tankspectively). FL

mportant role ine battery-bankored hydrogenrategy – in the so leading to eain target.

Wind Turbines indicated earf energy as it stems [1]. A

ower range (1

g. 1. Stand-alone w

d such kind om the local powcame more anse last years. A

competitive rediesel generatostine, is taken aed near the coaof Gazian peope of this work ower system system shoull as have an e

re-using the hieve that, a time estimato

spectively – ant description stem is briefly of storage and of the Alshiefpplication is come of the oby, section VI utlook.

II. PLANT CO

nd-alone hybrirated in Fig. proposed syste

rator (main soynchronous genupplementary nk and hydrogk (short-term LC together wn controlling thk and the elecn is injected combustion en

extensive energ

e rlier, the wind

is often the variety of w

100 kW up to

wind-diesel power

of systems totwer grid. Thusd more populActually, windgarding the trr set [5]. The as case study hastal area in Gple. is to develop afor the previ

ld be able to dffective and insurplus powe

Fuzzy Logic or and a presare applied. Iof a stand-alooutlined. Sect

d re-use of the fa hospital elecclarified in setained results is dedicated t

ONFIGURATION

d wind-diesel 1. Four main

em which are: ource of enernerator set (freand backup

gen path incluand long-term

with the time ehe flow of excectrolyzer as w

– according ngine driving thgy independen

turbine acts ascase in such

ind generatorso 5 MW) ba

r system (main com

ally isolated a, the ‘small wilar in the remd turbine usageraditional enemain hospital

here. The Alshiaza city. It ser

an efficient wiiously mentiondeliver the powntelligent strateer in the bi-fController (FL

scribed re-use In the followone hybrid wition III introdu

excess powertricity supply aection IV, whwill be foundo the conclusi

power generat

n components wind turbine wrgy), combustequency gover

power sourcuding electrolym energy storaestimator playess power towa

will be seen lato a pre-definhe generator, tcy – which is

s the main soudetached pow

s in the relevased on differ

mponents).

and ind’

mote e is

ergy l in iefa rves

ind-ned wer egy fold LC)

of wing ind-uces r. A as a hile d in ion,

tion are

with tion rnor ce), yzer age, an

ards ater. ned thus the

urce wer vant rent

aerodynavailablfollows the rotor

PWT=wher

PWT :cp : ρ : A : TuVwind: Wα : Tipβ : BlaV, Vc, Vffurling s

The mand α =

B. GeneDual

environmit can bcircumslong lowgeneratosystems

The iis not oestimatefuel is rrated cothen assthere shThis is generatoconsumapproximfull loadalmost consumpdependeengines.a reductamount above. Ton such

Detaimodes o

C. EnerOne

is choosenergy. play an

namic and elle. As the withe variations r speed as indi

0.5·cp α , βPnom 0

re Mechanical o Performance cAir density (k

urbine swept aWind speed (m/p speed ratio oade pitch angle Vf, Vr : wind spspeed, and ratemaximum valu0.81. For mor

erator Set l-fuel engine mental-sustain

be committed fstances, for insw wind periodor set plays an s since it regulaimpact of the w

obvious as it me fuel saving required to proonditions of thsume that for ehould be a pronot the case,

or decreases amed for a given

mation, fuel cod value, and tlinear or qu

mption to rateent, with the la. A decrease intion in fuel usethat would be

The actual fuefactors as the iled representaof operation wi

rgy storage sysof the main ch

sing the most aNot only the important rol

lectrical princind speed fluc

of the wind spicated in (1) [6

·ρ·A·Vwind 3 V

utput power ofcoefficient of t

kg/m³), area (m²), /s), f the rotor blad(deg.)

peed estimated ed speed in (m/ue of cp (0.48) e details see [8

as source onable – especiaflexibly and instance running s which are ve

n important roleates the frequenwind turbine o

might at first swould be to doduce a kWh

he diesel modeevery kWh proo rata drop inhowever, bec

at low load, amount of pro

onsumption atthe relationshiuadratic. Noteed fuel consuarger values con power generae, but it is typie calculated in el saving at desize, type, andations of the ill be discussed

stem hallenges for suappropriate stra

economical ale in selecting

ciples are coctuates, the oupeed as a cube, 7]. Vc ≤ V ≤ V Vr ≤ V ≤

V ≤ Vc and V ≥

f the turbine (Wthe turbine,

de tip speed to

at height h, cu/s), respectively is achieved fo

8].

of energy isally with dieselndependently fon diesel fuel

ery rare; furthee in the stand-ancy in the systeon diesel fuel ceem. The simpdetermine firstof electrical e

e generator setoduced by the wn diesel fuel cause the efficicausing more

oduced energyt no load is 15p for interven

e that the noumption ratio

orresponding toation will actuaically closer tothe simple wa

ecreased loadind the age of the

dual-fuel engd later.

uch isolated poategy for storinand environme

the most suit

2

ommercially utput power e function of

Vr

Vf

Vf

(1)

Watt),

wind speed,

ut-off speed, y. or β = 0 deg.

s not fully l mode – but from natural only during

ermore, such alone hybrid em. consumption plest way to t how much

energy under t. One might wind turbine onsumption. iency of the

e fuel to be . To a rough

5-30% of the ning loads is o load fuel o is engine o the smaller ally result in

o 2/3rd of the ay described ngs depends

e engine [9]. gine and the

ower systems ng the excess ental aspects table storage

dewe

ondaThmaacmode

e.gdesuran

eleaneffpelimin hyof

exhighyilluof

D.

knrulmaoucohaproim

Figfibval

evices, but the ell as cost consFor instance,

nly appropriatay/night cycleshey can also aain function incumulator banodel which is r

efinite (full) chaLonger term

g., electrolysisecomposed intorplus electric

n aqueous electr

H2OUnder ideal

ectricity and 8nd 1.0133 Bar)ficiency can n

erfectly ideal dmitations. The

the range of ydrogen storagf internal pressu

Shape of thexternal surface ghest challeng

ydrogen storagustrates a verti

f storage vessel

Fuzzy Logic Fuzzy logic

nowledge and les. It does noath equations g

utputs. It proviontrol because andled by ruleocess which r

mplementation.

g. 2. A schematicer and epoxy resinlve, (5) tube for hy

functionality osiderations can

customary acte for small s of smaller phact as short ten our work. Thnk can be detrun in parallel arging statusesand larger sca

s and hydrogeo its elementacurrent betweerolyte [13]:

O+electrical ecircumstance

.9 liters of wato produce 1

ever be reachedue to thermocurrent electro

f 52% to 82%e can be simpure. e hydrogen stand the type o

ges concerningge system. Forical section forls.

Controller andallows descrexperience in

ot require anygoverning the rides also an alit is closer to

es, membershipresults in impr

c view of the hybrn, (2) thin aluminuydrogen [15].

of each individalso determine

ccumulators arscale storag

hotovoltaic baserm energy buhe actual charermined by anand tuned from

s, e.g. [11]. ale storage canen storage [12ary componenten two electro

energy→ H2+es it requirester at normal ckg of hydroge

ed because theodynamics as olyzer efficienc

% [14]. The chply determined

torage vessel, of the storage mg the design ar more details,r one of the new

d Time Estimatribing complexn simple natu

y system moderelationship belternative solutthe real worldp functions, aroved performa

rid hydrogen storaum liner, (3) hydro

dual technologye the proper onre expensive a

ge, i.e. coversed systems [1

uffer which is rging status ofn observer bam time to time

n be achieved 2]. Water can ts by passing odes separated

12

O2 s 39.4 kWh conditions (25en. The maxime process is ne

well as matecies generally harging status

d by measurem

the liner of media are now and build-up o, see [15]. Figwest technolog

tor x systems usural-language-leling or competween inputs ation to non-lin

d. Non-linearityand the infereance and simp

age vessel: (1) carogen storage alloy

y as nes. and ring 10]. the

f an ased

for

by, be

the by

(2) of

° C mum ever erial

are s of

ment

the the

of a g. 2 gies

sing like plex and

near y is nce pler

rbon y, (4)

Regashould befficientto deterwhile thchargingfluctuatiset-poinallowedbefore sreduce expectanthe stratconsider

III. EX

As mstored tstrategieand the storage

A. FuzzThe

containiapproprcontrollefficienttargetingrelates tTHEN) of the hyto the Fexcess smethod

Thersplit pepresentssystem d

As wcontrollto drivethe life burnish

Fig. 3. Stestimator,electrolyz

IF (SOC iIF (SOC iIF (SOC THEN (TaIF (SOC i

arding the propbe managed intly. Fuzzy logirmine the actuthe battery-bang level in ordions. The time

nts of the electd time duratioswitching ON/

material strncy. The follow

ategy of the enrations.

XCESS POWER:mentioned befoto be used ines of the electdual-fuel gen

and re-use, res

zy control of elease of for

ing imprecise vriate means foer was used t split of the pg the operatiothe controller rules. Accum

ydrogen tank aFLC, while thesplit. Fig. 3 ill[16].

re are 23 fuzzercentage forws some exampdescribed in thwill be seen er is widely fl

e the electrolyztime of the

the portion

tructure of fuzzy c, which then pzer.

EXCERPT OF RUis Empty) THEN (is E-Medium) and is F-Medium) andank is High) is Full) and (Pressu

posed system, n the way that tic controller anual power set-nk has to be

der to cover the estimator cotrolyzer, but iton before cha/OFF the elec

ress and thewing section cnergy storage

: MECHANISM O

ore, excess pon the times trolyzer follow

nerator set conspectively.

lectrolyzer andrmulation of verbal expressor plant contrto achieve a

power excess ion set-point output to the

mulator State Oand the power e output is thelustrates the fu

zy logic rules warded to the ple rules of FLhis paper. later, the respluctuated. Suchzer. Thus, to inelectrolyzer, tn of exces

controller [16]; thprovides certain

TABLE I ULE BASE OF FUZZY(Tank is Zero) (Pressure is Dang

d (Pressure is Me

ure is not Dangero

storing the surthe electrolyzend time estima-point of the e kept on an he remaining rontrols not onlyt limits also thanging the sectrolyzer, both erefore enhancontains more dtaking into ac

OF STORAGE AN

wer has to beof deficiency.

wed by the timntrol the proces

d time estimatological oper

sions makes furol. Thus, a

situation depin the proposeof the electroinputs with a

Of Charge (SOexcess are the

e percentage ouzzy logic con

used to derive time estimat

LC that are ap

ponse of the h ripples are nncrease the effthe time estimss power en

e output is forwaroperational set-

Y LOGIC CONTROLL

gerous) THEN (Tadium) and (Exces

ous) THEN (Tank

3

rplus energy er is working ator are used electrolyzer, appropriate

rapid power y the power he minimum et-point and

in order to nce lifetime details about ccount some

ND RE-USE e efficientely . The work

me estimator ss of energy

r ration rules

uzzy logic an fuzzy logic

pendent and d system by olyzer. FLC a list of (IF-C), pressure three inputs

of the power ntrol analysis

e the power tor. Table I pplied in the

fuzzy logic not preferred fficiency and mator has to ntering the

rded to the time -points of the

LER

ank is Zero) ss is Abundant)

is High)

elechthe

B.

ento is igntot

priclealo • • wi• of • nutak • • • 80

en

cosyM

simPS

Figsetres

ectrolyzer by hanges, see Fige electrolyzer a

Dual-fuel GeA dual-fuel c

ngine that has butilize gaseoua true diesel

nition of the gtal energy of thThe dual-fue

imary benefit eaner cheaper one when nece1) Engine Cha a) HydrogRunning on gaAutomatic and

ithout loss of eTransfer to die

f engine power Automatic tri

umber of minutken in this stud b) Diesel MRunning on diTransfer to hyAutomatic tra

0% without lossFig. 5 illustra

ngine.

IVFor the inves

ontrol the applistem is requiratlab/Power SyPSB is a grap

mulation of poSB is used to

g. 4. Output signa-points finally gpectively [17].

constraining og. 4. For moreand the time es

nerator Set combustion enbeen fitted wits fuel as a suppand requires

gas as main fuehe fuel [18]. l engines hav

is that of fgaseous fuel

essary. aracteristics agen Gas Modeas and diesel pid instant trip tongine power anesel mode on rand speed. ip to diesel tes at engine lo

dy). Mode iesel fuel inject

ydrogen gas moansfer to hydros on engine poates the mode

V. MODELING A

stigation and iication of a simred. Regardingystem Blockset

phical tool that ower systems represent com

al of FLC and resgiven to the ele

operation perioe details aboutstimator, refer t

ngine – originath additional deplemental fuel.some level o

el, in practice a

ve a number ofuel flexibility

when availab

nd Operating me ilot fuel injectio diesel mode nd speed. request at any

mode after aoads below 15

tion. ode on request.ogen gas modwer and speeds of operation

AND SIMULATIO

implementationmulative modeg the case stut (PSB) is usedallows buildinin the Simul

mmon compon

sponse of the timectrolyzer), in bla

ods and set pot fuzzy controlto [17].

ally – is a dieevices allowin. This engine tyof diesel fuelaround 1% of

of advantages.y, operating wble and on die

modes [18]

ion. in alarm situat

load without l

a pre-determin5% (5 minutes

de at loads bel.

n of the dual-f

ONS n of proper plel of the compludy pursued hed for this issue.ng schematics alink environmenents and devi

e estimator (i.e. poack and red co

oint l of

esel ng it ype for the

. A with esel

tion

loss

ned are

low

fuel

lant lete ere, . and ent. ices

ower lors,

Fig. 5. M found iSimulin

Choosimulatemodel dLinearizoutput papplied describeto aerod

Thretraditionillustrateregulatospeed gsystem c

Fig. 6. Si

Fig. 7. Si

Modes of operation

in electrical pnk model of theosing the wind e the behaviodescriptions sized [19] or qupower of smalhere, the prod

es the power gdynamics. e main partsnally diesel ges the synchro

or (AVR), congovernor. Mocan be found in

imulink model of t

imulink model of d

of the dual-fuel en

power networke power system

turbine modelor of the renimulate the wiuadratic [20] eqller wind turb

duced power isgenerated by th

s constitute thgenerator as monous generato

nnected to the ore details abon [18].

the power system o

dual-fuel generator

ngine [18].

ks. Fig. 6 illm of the Alshiefl plays an impo

newable energyind turbine ouquations can r

bines. More prs represented bhe wind turbin

he dual-fuel mentioned earor with automdual-fuel engiout the dual-

of the Alshiefa ho

r [21].

4

lustrates the fa hospital. ortant role to y. Different utput power. represent the recisely, and by (1) which ne according

generator – rlier. Fig. 7

matic voltage ine with the -fuel engine

spital.

the– drawhter

(75thepropreeqPV

Z NH

R TVMH

prea sordtem(30hy

A.

indThneGaat apwisplocda(h

Fig

As shown froe rectified and for such dev

amatically thehich improve trm simulationsThe efficienc

5 %) includinge electrolyzer eoduced hydroessure of the h

quation [22]: where

Vi : is the init : is the com

H2 : is the hyd : is the ide

V,VV : are the vMH2. : is the mo

Note that the essure is lowerso called high pder to obviamperature is co00 K). Fig.

ydrogen storage

V

Wind and loaAs mentioned

dependent powhe Alshiefa hoear the sea. Thaza-Strip. The height hrof 5

pproximately 7ind speed is 5.4eed values fromcation. Equatio

ata recorded a= 56 m).

g. 8. Simulink mo

om Fig. 6, twocontrolled exc

vices given ine simulation spthe simulation s. cy of the batteg the inverter efficiency is ta

ogen is storedhydrogen tank

PV– PVi= Z

tial pressure ofmpressibility fadrogen flow in eal gas constantvessel temperatolecular mass o

compressibilitr than 140 Barpressure electr

ate an additioonsidered to b8 illustrates te tank.

V. RESULTS AN

ad profiles d before, the cwer supply of tspital is locate

he city has onoriginally mea

5 m – to be re-7 m/s, while t4 m/s. Fig. 9 shm August, 201on (4) is usef

at certain leve

del for hydrogen s

o main paths acess power. Thn Matlab/Simupeed. So, simspeed – were

ery-bank is conefficiency. On

aken constant hd in a vesselis obtained fr

Z·NH2·R·TV

MH2·VV

f the hydrogen actor of the hyd(moles/sec.), t, ture and volumof hydrogen. ty factor is equr [22], which isrolyzer that waonal compres

be constant at rthe Simulink

ND DISCUSSION

case study perthe main hospied at the west e of the best asured annual -calculated accthe minimum hows the mont0 to July, 2011ful to convert el to the estim

storage tank [22].

are introduced he detailed modulink – decre

mplified modele applied for lo

nsidered constn the other hahere at 75 %. Tl. The produ

rom the follow

in the vessel,drogen,

me; respectively

ual to 1 when s not exceededs assumed here

ssor. The vesroom temperatdiagram for

NS

rformed conceital in Gaza-Stpart of Gaza cwind potentialmean wind sp

cording to (4) –monthly aver

thly average w1 for the propo

the anemomemated hub cen

for dels ease s – ong

tant and, The

uced wing

(3)

y,

the d by e in ssel ture the

erns trip. city l in eed – is rage

wind osed eter nter

wherand α isto 1.4 at

A samillustratespeed reRecords

Fig.hospitalminuteselectricacycle; thweek. Afairly hiabout 11

B. OptiAs in

system storage relevantas descdriven inductioof energcable co

Fig. 9. Av

Fig. 10. Rthe small time as th

V

re Vr is the wins the ground sut the proposed mple of slightled in Fig. 10. ecords taken fs were measure11 shows the

l during a days in Decemberal load profilehis is evident sAccording to thigher than 800 1 MWh.

imization and Cndicated previconsists of a wsystem, and

t components (cribed in [23])

generator witon generator wgy are conneconnections.

verage monthly w

Recorded wind spexcerpt window

hose in Fig. 13 – 17

V = Vr · h h⁄nd speed measurface friction location.

ly more than oIt was collect

from August, 2ed every hour fmeasured elec

y. The load dar, 2009. It waes are almost since the clinic he consumptionkW, while the

Components Siiously, the exwind turbine, an electrical

(based on Part) has resultedth 800 kW,

was rated at 115cted to the va

wind speed values a

peed data (from Aushowing more de

7, so that they can

hr ured at referencoefficient wh

one year wind sted according 2010 to Septefor the whole tctrical load prata were taken

as identified ththe same duriservice contin

n records, the e daily consum

izing xample stand aa generator seload. The si

ticle Swarm Od in a combus

while the w50 kW. These ariable load th

at coastal area – G

ugust, 2010 to Sepetails is for the saimmediately be co

5(4)

nce height hr hich was set

speed data is to the wind

ember, 2011. time. rofile for the n every two hat the daily ing the year

nues 7 days a peak load is

med energy is

alone power et, an energy izing of the

Optimization, stion engine

wind turbine two sources

hrough short

aza.

ptember, 2011); ame periods of ompared.

Fig

MWtaneneleele

cotonthehy

Susim

C.

obSim

forof gethesppothe

ele

Fig

g. 11. Daily electr

The corresponWh, while thenk is 165 m³ –

nergy – and canectrolyzer outpectrolyzer is 10With this sizin

onsumption wons/year with pue produced hy

ydrogen re-use Fig. 12 repres

uzlon) wind tumulations. For

Simulation RThe simulatio

btained in mPowerSystemIt is clear from

r the wind turbf 1250 kW. Tenerator produce available wineed of 25 m/s.

ower that was pe wind turbine;In consequen

ectricity – eithe

g. 12. Characterist

ical load demand f

nding size of te volume of th

maximum capn withstand 14put. The nomi000 kW. ng of plant comould be only ure diesel base

ydrogen re-usedis much highersents the charaurbine – the dmore details, r

esults and Discon results show

Matlab-Simm toolbox. m Fig. 12 thatbine induction This means thces its maximund speed is abo

Fig. 13 illustrproduced durin; the simulationnce, the dualer by diesel fu

tics of S66-1.25MW

for Alshiefa hospit

the battery-banhe cylindrical pacity is 63 MW40 Bar as maxnal power of

mponents, the 530 tons/year d generation (id as fuel); ther, as to be seenacteristics of thata of which refer to [24].

cussions wn in the foll

mulink envir

t the rated windgenerator with

hat the wind tum power (125ove 14 m/s and ates the amoun

ng the whole pen resolution wal-fuel generat

uel or hydrogen

W wind turbine, S

tal.

nk should be 1hydrogen storWh of the surp

ximum pressurethe high-press

annual diesel finstead of 16

i.e. not yet have fuel saving wn later. he (S66-1.25Mwere used in

owing have bronment us

d speed is 14 mh a nominal powturbine induct0 kW) only whbelow the cut-

nt of the electrieriod of time fras set to 1 hourtor set produn gas – in orde

Suzlon.

1.44 rage plus e at sure

fuel 690

ving with

MW, the

een sing

m/s wer tion hen -off ical rom r.

uces r to

Fig. 13. D cover thillustratefrom thpower pMore prfrom booperatiofor diese

It is increasehydrogewater prpressurecase 12

Fig. 14. hydrogen gas modes

Delivered power b

he instantaneoes the shared a

he diesel fuel, produced durinrecisely, the proth modes ofon – would be el mode and hyobserved in Fi

es sharply. Inen pressure, a 2reparation is coe of the hydro0 Bar); this ca

Delivered power gas mode; for the

s, respectively, see

y the wind turbine

ous deficit eleamount of the while Fig. 14

ng the hydrogeroduced amou

f operation – 660 MWh an

ydrogen gas mig. 15 that the

n order to av200 kW additionnected to thegen tank excean be seen in

(a)

(b)

from dual-fuel gee ratio of maximae section III.

e.

ectrical powerelectrical pow

4(b) shows theen gas mode, runt of the elect

for the wholnd 1065 MWh mode. e electrical loadvoid dangerouional resistive e system direct

eeds a certain lFig. 15 in the

enerator set; (a) dal powers in diese

6

. Fig. 14(a) wer produced e amount of respectively. trical energy le period of sequentially

d sometimes us levels of

load for hot tly when the limit (in our e months of

iesel mode; (b) el and hydrogen

Minsexthe

theFigtimdifele[17

preIt (72

Fig

Fig

Fig

ay, June and stantaneous loa

xceeds the genee accumulator To investigat

e proposed eng. 16 – should

me estimator gfference betweectrolyzer is st7]. The SOC of

essure of the his observed th20 kWh) at t

g. 15. Electrical lo

g. 16. Power exces

g. 17. Response of

July. It conad increase witerator set nomibank. e the effectiveergy storage sbe considered

giving the set een excess powteadily compe

the battery-bahydrogen storaghat the SOC wthe beginning

oad demand for Al

ss to be stored.

f the time estimato

ntinues at leasthin these perioinal power – c

eness of the FLsystem, electricd. Fig. 17 show

points of the wer and powernsated by the

ank and the dege tank are illuwas specified t of simulatio

shiefa hospital.

or.

st 6 hours. Aods of time – t

can be covered

LC in controllc power exces

ws the evolutionelectrolyzer. T

r absorbed by accumulator,

evelopment of ustrated in Fig. to a certain len and is larg

Any that

d by

ling ss – n of The the see

the 18.

evel gely

maintainvariationit. The mto workpower fabsorb electrolycover th

It is cthe parconstraistoring overall sSelectioaccumupermanemaximutime in o

Tablefrom theset EdiesEhydrogen

Ehospital,

EN

Fig. 18. (

ned on averagens in the battermain role of thk as a buffer ofor the electrothe power a

yzer. It can alshe sudden peakclear from the rticular systemints and restriand releasing

system efficienon of the apulator-bank anent supply ofum calm spellone year. e II illustrates e wind turbineseland the hy

n, in addition t, respectively.

NERGY SHARING (FSymEwinEdiesEhydEhos

(a) Charging state

e. It is also notry-bank SOC –

he battery-bankf the excess enolyzer. Thus, according to so deliver powks and sharp va

obtained resulm componentctions and the

g – of the exncy and maximppropriate en

nd hydrogen tf energy takinl; i.e. the long

the sharing ameEwind, the die

ydrogen gas mto the energy

TABLE II FROM AUGUST, 20

mbol Value nd 6375 Msel 660 M

drogen 1065 Mspital 4715 M

(a)

(b)

of battery-bank; (b

ted the fast res– which is the

k in the proposnergy to burniit can readilythe input po

wer to the electariations. lts that the propts under pree efficient maxcess power imize the diesel nergy storage tank – yieldsng into considgest non-wind

mount of energesel mode of thmode of the g

consumed by

010 TO SEPTEMBER

MWh MWh MWh MWh

b) Pressure of hyd

7

sponse to the idea behind

sed system is ish the input y deliver or

ower of the trical load to

per sizing of e-determined anagement – improve the fuel saving. devices –

s also to a deration the

dy period of

gy produced he generator

generator set the hospital

R, 2011)

drogen tank.

8

It is noticed that the diesel fuel consumption is only 280 tons instead of 1975 tons with pure diesel based generation for the same period of time. More deeply, one can strongly say that the proposed hybrid wind-diesel power system performs significant cost advantages in comparison with diesel fuel only; and this even under inclusion of the investment cost as shown in [23]. The corresponding fuel cost is relatively higher than 0.4 Million Euro, being less than 15% of a diesel-only solution (about 2.96 Million Euro) to provide independence of the public grid.

These impressive results – especially economically ones – encourage the responsibles to use such clean sources of energy, achieve fuel cost saving and relative independence of fuel delivery, as mentioned before.

VI. CONCLUSION, SUMMARY, AND OUTLOOK A hybrid wind-diesel power system was investigated in this

paper, applied to a stand-alone electricity supply of the main hospital in Gaza-Strip as a relevant and realistic case study. MATLAB/Simulink models were introduced for the wind generator, dual-fuel generator set, hydrogen storage path and accumulator bank. Simulation results over one year, based on real wind harvest and load data, proved that the chosen rating of components in connection with the implemented operation strategy for the plant leads to efficient, continuous and independent supply with an amount of used diesel fuel below 15% of that needed for conventional independent electricity supply purely based on diesel generation.

Further work will aim in additionally investigating re-use of the waste heat of the combustion engine. Regarding the hospital example, surgical demands, air conditioning (by absorption coolers) as well as hot water supply are vigorous heat absorbents. In this way, the overall efficiency of the system can further be enhanced.

VII. ACKNOWLEDGMENT The authors thank the German Academic Exchange Service

– Deutscher Akademischer Austausch-Dienst – (DAAD) for providing scholarship for M. Mushtaha to pursue his PhD degree at University Duisburg-Essen.

VIII. REFERENCES [1] L. Mott (NPS) and B. Saulnier (IREQ), “Commercial wind-diesel

project, St. Paul Island, Alaska,” 14th Prime Power Diesel Inter-Utility Conference, Winnipeg, Manitoba, Canada, May 28-June 2.

[2] A. J. del Real, A. Arce, and C. Bordons, “Optimization strategy for element sizing in hybrid power systems,” Journal of Power Source, vol. 193, pp. 315-321, 2009.

[3] J. Kaldellis and D. Zafirakis, “Optimum energy storage techniques for the improvement of renewable energy sources-based electricity generation economic efficiency,” Energy, vol. 32, pp. 2295-2305, 2007.

[4] F. J. Pino, L. Valverde, and F. Rosa, “Influence of wind turbine power curve and electrolyzer operating temperature on hydrogen production in wind–hydrogen systems,” Journal of Power Sources, vol. 196, pp. 4418-4426, 2011.

[5] J. G. McGowan, W. Q. Jeffries, and J. F. Manwell, “Development of Dynamic Models for no Storage Wind-Diesel Systems,” Proceedings of the 17th British Wind Energy Association Conference, UK, July-1995, pp. 111-116.

[6] E. Muljadi, J. Sallan, M. Sanz, and C. P. Butterfield, “Investigation of self-excited induction generators for wind turbine applications,” IEEE

Industry Applications Society, Annual Meeting, Phoenix-Arizona, October, 1999.

[7] M. Druga, C. Nichita, G. Barakat and E. Ceanga, “Stand-Alone Wind Power System Operation with a Specific Storage Structure,” International Conference on Renewable Energies and Power Quality (ICREPQ’09), Valencia (Spain), Apr. 2009.

[8] S. Heier, “Grid Integration of Wind Energy Conversion Systems,” John Wiley & Sons Ltd, 1998, ISBN 0-471-97143-X.

[9] R. Hunter and G. Elliot, “Wind-Diesel Systems: A guide to the Technology and its Implementation,” Cambridge University Press, 1994.

[10] Press Release Saft-Batteries, New design of specialized rechargeable nickel-cadmium battery, http:/saftbatteries.com.

[11] P. Beckhaus, C. Hardt, G. Krost, and S. Souzani, “Ladezustandserkennung durch simulative Batteriebeobachtung” (Recognition of battery state of charge by simulative observer), 18th Symposium on Photovoltaic Solar Energie, Staffelstein (Germany), March 2003 (in German).

[12] P. Beckhaus, G Buchholz, A. Graw, G. Krost, and J. Matics, “Solar hydrogen based energy supply for residential development,” World Renewable Energy Congress (WREC-VII), Cologne (Germany), July 2002.

[13] M. J. Khan and M. T. Iqbal, “Dynamic Modeling and Simulation of a Small Wind-Fuel Cell Hybrid Energy System,” Renewable Energy, vol. 30, pp. 421-439, 2005.

[14] G. Saur, “Wind-To-Hydrogen Project: Electrolyzer Capital Cost Study,” National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401-3393, Tech. rep. NREL/TP-550-44103, Dec. 2008.

[15] N. Takeichi, H. Senoh, T. Yokota, H. Tsuruta, K. Hamada, H. T. Takeshita, H. Tanaka, T. Kiyobayashi, T. T. Takano, and N. Kuriyama, “Hybrid hydrogen storage vessel, a novel high-pressure hydrogen storage vessel combined with hydrogen storage material”, International Journal of Hydrogen Energy, vol. 28, pp. 1121-1129, 2003.

[16] K. Jeong, W. Lee, and C. Kim, “Energy management strategies for a fuel cell/battery hybrid system using fuzzy logic,” Jounal of Power Sources, vol. 145, pp. 319-326, 2005.

[17] M. Mushtaha, and G. Krost, “Independent and Clean Electricity Supply for a Hospital in Palestine Controlled by Fuzzy Rules and Time Estimator,” 16th International Conference on Intelligent System Applications to Power Systems (ISAP), Crete, Greece, Sept. 2011.

[18] B. Thijssen, “Dual-Fuel Electric LNG Carriers: LNG Shipping Operation,” Wärtsilä Finland Oy, Hamburg, Germany, Sept. 2006.

[19] Z. Wei, and Y. Hongxing, “One optimal sizing method for designing hybrid solar-wind-diesel power generation systems,” Proceedings of ISES Solar World Congress 2007, Solar energy and Human Settlements, vol. 4, pp. 1489-1494, 2007.

[20] R. Pallabazzer, and A. A. Gabow, “Wind generator potentiality in Somalia,” Renewable Energy, vol. 2, No. 4/5, pp. 353-361, 1992.

[21] K. Yeager, and J. Willis, “Modeling of Emergency Diesel Generators in an 800 Megawatt Nuclear Power Plant,” IEEE Transactions on Energy Conversion, vol. 8, No. 3, pp. 433-441, Sept. 1993.

[22] H. Görgün, “Dynamic modelling of a proton exchange membrane (PEM) electrolyzer,” International Journal of Hydrogen Energy, vol. 31, pp. 29-38, 2006.

[23] M. Mushtaha, and G. Krost, “Sizing a Self-sustaining Wind-Diesel Power Supply by Particle Swarm Optimization,” IEEE Symposium Series on Computational Intelligence, Paris, France, Apr. 2011.

[24] Suzlon, “Powering and Green Tomorrow, S66-1.25MW Technical Overview,” [online]. Available: http://www.suzlon.com.