GROUND BASED WIRELESS AND WIRED POWER TRANSMISSION
COST
COMPARISONRichard
M
Dickinson, JPL’Jet Propulsion Laboratory
MIS
238-5284800 Oak Grove DrivePasadena, CA 9
1
109
(8
18)
3.54-2359, Fax:
(8
18)
393-0207
Abstract:
Recent interest
in
applicationsof wireless power transmission has raisedthe question of the cost comparison ofwireless vs wired transmission of powerfrom ground site to ground site. Costs interms of $NW-km, a figure of merit, forpast demonstrations and stimates forfuture suggested activities are given
s
afunction of power delivery distance.Wired power types uch
as
open wirehigh voltage ac and
C
lines, direct burialand submarinecables, appliance cordsand circular waveguide transmission linesare discussed. Beamed
RF
powerapplications in laboratory tests, fieldtests, proposed direct power deliveryapplications and power via reflector relayare presented. Although not ground-to-ground, but ground-to-air, additionalhistorical and proposed electric aircraftdemonstrations and airship applicationsare also discussed for comparison.A graph of the first approximation of
installed.
system
costsas
a
function
of
distance, in then year for the figure ofmerit is generated. Except for the veryhighest power, longest range cases, thewireless power cost is orders ofmagnitude more costly than wired powerfor delivery at the same distance on
r
near the Earth’s surface.INTRODUCTIONRecent studiesby NASA of Space(Based) Solar Power
[I]
have rekindledinterest
in
transmitting electric power byRF microwave beam
[2].
Owen Maynard, Aerospace Engr. Ret.
5.5
Blue Springs Drive, Apt.
70.5
Waterloo, Ontario Canada N2J-4T3Fax
1.5
19)
888-9176
(5 19)
888-996.5
Those unfamiliar with the lements ofcosts want to find an Earth based locationfeature such as a canyon[3], rivercrossing, lake
[4],
strait,
or
otherdifficult crossing to apply
F
wirelesspower transmission (WPT) thinkingtmust be cheaper than wired ower.To our knowledge, such relative (albeitimprecise) cost comparison has not beenpublished. The purpose of this paper isto attempt to document a firstapproximation comparisonof wired andwireless
RF
power transmission,reception and control systems
s
afunction of power magnitude anddistance.Because the field f
WPT
is rather new,we will be mostly comparing laboratorytests, first-of-a-kind demonstrations andprojected application’s costs withestablished utility costs and their longrange projections. Thus, a scatter plot ofthese disparate costs will be usedor thereader to compare, with the caveat that thevasious entries have
markedly
diifferentprovenance. Only general trends shouldbe derived from such data.Economical applicationsof beamedpower
in
the near future will probably befor powering high altitude telecom andobservation platforms,stratospherictourism airships and other ff-Earthvenues. Although not strictly ground-to-ground, they are point-to-pointand wheredata exist hey will be included forcomparison.
I
contract with the National Aeronautics
and
Space Administration.The research described in this paper was carried out by the Jet Propulsion Laboratory. California lnstltute
of
Technology, under a
5/27/99
1
2:
19
PM
As
the two key aspects of WPT arepower magnitude delivered and thedistance over which
t
is delivered,we
will
use as a figure of merit
(FOM)
theinstalled system cost scaled by thedistance
in
km and the power magnitudedelivered
in
MW.
The WPT installedcosts will include the electric powerconditioning at both endsf the link
as
well as the beam safety equipment. Costconsiderations for the source, prime-generator
or
load will not be included,only transmission costs.We will explore the safety and economicsof wired and wireless power, where bothends of the transmission link are on thesurface of theEarth, by briefly reviewingprevious activities. This paper will thendiscuss some current investigations ostestimates and compare themo pastwireless and wired power cost estimatesor projections. Then conclusions willbedrawn.GROUNDWPTBACKGROUNDOne of the authors (Dickinson) recallsbeing asked in 1976 by Sam Fordyce
his
NASA Headquarters sponsor to look ntodelivering electric power via microwavesto Manhattan Island fromNew Jerseyover a particular
16
km
range,
as
undenvater cables were being overloadedand were expensive to install. This wasafter his successful managingf ademonstration in 1975
[5]
of deliveringover 34 kW of DC at 1.6
m
atGoldstone, CA under the ContractProgram Management by the otherauthor (Maynard) when he was atRaytheon. We both thank Bill Brownvery much for developing the rectenna [6]
that
was key to that project, and being itsTechnical Director.The calculation performed for Fordycedetermined that
it
would require atransmitting may and a rectenna array,both larger
in
diameter than the height of
an
18 story building,
in
order to achieveover 90% beam coupling efficiency at
2.35
GHz
over the distance required.After that shochng result, we lookedelsewhere for potential applications, butfound no critical
or
desirable sites. Theend of the oil crisis and the NationalResearch Council report [7] soon took thewind out of SpaceSolar Power (SPS)sails, and interest
n
WPT waned.Thus WPT activities were mostly quietuntil 1987 when
the
Canadians powered
a
model aircraft with rectennas near 2.45GHz [8].A contact from Exxon in Houston wantedto know
if
the energy from Alaska northslope oil could be converted tomicrowaves and delivered o Houston.
An
Alaska entrepreneur wanted to knowif
WPT
could be used to supply electricpower to native American strandedvillages
in
his
state. DOE began lookinginto delivering Alaska crude energy toJapan via beamed power. NASA’sCenter for Space Power at Texas&M
[4]
along with
DOE
also began aninvestigation of conducting ademonstration of beaming power acrosslake toa village in Alaska.
S.
Bharj and colleagues at Sarnoff,powered the “Moonstruck” rover at
5.8
GHz with 450W output at 200ft range(61m)
in
1992 [9].The Japanese conducted demonstration
of
ground-to-ground power transmissionat
Yamasaki
with the Kansai ElectricCompany
[IO]
in 1994-95. Theyinvestigated beaming power from theirmainland to islands in the Inlandea.However, subsequently they ncounteredsevere environmental concerns from theisland residents and abandoned theinvestigation.Ralph Nansen
[
1
11
endeavored to puttogether and conduct an approximately
$50
M,
3-year demonstration
n
Texas ofground based photovoltaic powerconverted to 50-250kW microwave RFand beamed
1-5
km
to a rectenna to betied into the local utility’s grid. Theproject did not materialize due to lack offunding.5/27/992 2: 19
PM
provide the short term storage to bridge tomedium term flywheels and
o
longerterm natural gas turbines
or
pumpedhydro storage
for
example.The second potential WPT venue whichsurfaced in 1997-8 was
in
EasternCanada across the Straitf Bell Isle thatseparates Newfoundland island fromLabrador mainland. Here the distance isabout
40
km
and the desired power levelwas postulated to be about 800MW. Inaddition to cruise ships, there areoccasional
60
m (several hundred
ft.)
tallicebergs that scour the strait bottom,giving underwater cable systemsproblems.Again, due to weather andQOSconsiderations, 2.45 GHz is the ISM-Band frequency of choice. Thus thediameter of transmitters and rectennasor90
%
beam coupling efficiency must beover 97 m in diameter(29 stories tall)each if equal in size. In this case, thebulge of the Earth adds
n
additional
3 1.25
m height at each end for the beamto just be tangent to the Earth at the mid-point. Realistically, the ends
of
theterminal need to be sited at differentelevations to minimize multipathpropagation and
for
safety reasons thereis a preference
for
the rectenna o
be
higher.Terminals
of
the
RF
link should be basedon mountain tops
or
hillsides inorder forthe beam to clear cruise ships.
An
investigation by Maynard revealed thatsuch sites in the topography
f
the BellIsle strait would require
RF
transmissiondistances from
65
to 88 km, not 40 km,thus increasing the required transmitterand rectenna diameters to over
130
m.Similar beam safety and energy storageconsiderations apply as n the Singaporecase, but due to the large quantity ofpower (0.8 GW), the prudent engineeringdesign would be to have parallel pairs toprovide redundancy
for
accommodatingiceberg blockage and to permit down time
for
maintenance. Perhaps
four
200MWlinks spaced more than an iceberg lengthapart should be
used.
WIREDPOWER
SYSTEM COSTSThe least costly wired powertransmission system s a
6
ft
appliancecord at
10
A,
1
1OV
and
$5
cost, and
so
its installed cost figure of meritFOM) asshown
in
the comparison graph
is
over
10**3$/MW-km.
The United States Department ofAgriculture Rural ElectrificationAdministration (REA) Bulletin 151shows the results of typical economicconductor analysis or a high voltagetransmission line in support ofa
200MW
load at $160,000/mile, r for example at
100
km
long, the FOM is over 10**2$/MW-km,
an
order of magnitudecheaper than the appliance ord. There areeconomies of scale at work.Obtained from theAsea Brown Boveri(ABB) web site is the costf a 65
km
undersea 200MW ine between Finlandand Estonia at FOM of nearly 10**4
$/MW-km.
Data for a
1.2
GWunderwater link or the New York PowerAuthority to Long Island at 13
km
rangehas a FOM that is about 80% of the
BB
link, probably due o the larger powerlevel that is carried. However, both arean order of magnitude over the REAine.The approximately
1
GW,
+/-
500kV
DC
line from the Dalles in Oregon toSylmar in California around
000
km
has a FOM of installed cost of around
10**3
$fMW-km.Based upon Bechtel studies in 1968
[
161,
a
TEOl mode circular waveguide at GWand
500
km
length has a FOM of over10**2
$/MW-km.
An
8
GW nearlylo00
km
system would havea FOM about halfthat, the lowest cost
(
in then-year
$)
ground-to-ground wired systemencountered.Michael Klemke has estimated the costfmoving large quantitiesof electric powerover intercontinental distances with5/27/994
2:
19
PM
|