power sources for remote arctic applications

Power Sources for Remote ArcticApplications

June 1994

OTA-BP-ETI-129NTIS order #PB94-193232

Recommended Citation: U.S. Congress, Office of Technology Assessment,Power Sources for Remote Arctic Applications, OTA-BP-ETI 129(Washington, DC: June 1994).

Foreword

The U.S. Air Force operates a seismic observatory at Burnt Mountain, Alaska to assist in nucleartreaty verification. This unattended station --consisting of five sites clustered within a 1.5 mileradius--is located in a remote area north of the Arctic Circle, approximately 50 miles from thenearest villages. The seismic monitoring and data communications equipment at the station re-

quire low, but very reliable, power. Currently, the power comes from 10 radioisotope thermoelectric gen-erators (RTGs), each containing between 1.2 and 3.9 pounds of Strontium 90, a highly radioactive material.

In August and September 1992, a tundra fire at Burnt Mountain damaged some of the data cablesconnecting the sites. Though the fire did not impinge on the monitoring, communications, and powerequipment at the sites, it raised public concern among the nearby populations about the safety of usingradioactive material as the power source at the sites. Senators Stevens and Murkowski of Alaska askedOTA to evaluate alternative power technologies for the site. Senator Stevens is a member of the SenateCommittee on Commerce, Science, and Technology, among others; Senator Murkowski is a member ofthe Senate Committee on Energy and Natural Resources, among others. OTA examined the safety of theRTGs at Burnt Mountain and assessed the viability and risks of two alternative power sources--thermoelectric generators and photovoltaic (PV) systems--for the station.

This background paper concludes that continued use of the RTGs at Burnt Mountain entails low riskfor the safety of maintenance workers and local populations and for the environment. Installation of light-ning protection devices and intrusion detection devices at the station would lower the risk further still. Ifthe RTGs were required to be removed immediately, the only viable replacement power source would bea propane-fueled thermoelectric generator system. The principle risk of this type of system is the transportand storage of the roughly 5,000 pounds of propane that would be needed each year. If the RTGs could betolerated at site for three or four more years or longer, other power technologies may prove feasible. Atpresent, a PV system appears to be the most viable nonfuel power source. A PV prototype system wouldneed to be tested at the site to prove the technology’s cold weather and low sunlight performance. The safety and environmental risk of using PV system at the site is possible release of toxic fumes and heavymetals from the batteries.

OTA appreciates the assistance received from several individuals and organizations in the course ofthis study. To all of them goes the gratitude of OTA and the personal thanks of the project staff.

ROGER C. HERDMANDirector

. . .Ill

Preject Staff

Peter D. BlairAssistant DirectorIndustry, Commerce, and Inter-

national Security Division

Emilia L. GovanProgram DirectorEnergy, Transportation, and

Infrastructure Program

PRINCIPAL STAFF CONTRACTOR

John E. Newman Future Resources Associates, Inc.

Project Director (beginningAugust 1993)

Joe Raguso ADMINISTRATIVE STAFFProject Director (until August Lillian Chapman

1993) Division Administrator

Marsha FennTechnical Editor

Tina AikensAdministrative Secretary

Gay JacksonPC Specialist

iv

Mohamad Ali AzarmBrookhaven National Laboratory

Steven L. BaggettU.S. Nuclear Regulatory Commission

Samuel BaldwinOffice of Technology Assessment

Thomas CochranNatural Resources Defense Council

Gerald EpsteinOffice of Technology Asscessment

Tony Fainbergoffice of Technology Assessment

Lois JoellenbeckOffice of Technology Assessment

Peter JohnsonOffice of Technology Assessment

Reviewers

Thomas LampWright LaboratoryU.S. Air Force

Charles E. MayerSchool of EngineeringUniversity of Alaska - Fairbanks

Stan ReadAlaska Department of

Environmental Conservation

German ReyesOffice of Technology Assessment

George H. RotheU.S. Air Force Technical

Applications Center

John F. VogtTeledyne Brown Engineering-

Energy Systems

Matthew WeinbergOffice of Technology Assessment

contents

Chapter 1- Summary 1Air Force Findings 2OTA Findings 3Conclusions 5

Chapter 2- Background 7Equipment and Operations 8Radioisotope Thermoelectric Generators 11Strontium-90 Fuel 12Location and Climate 13

Chapter 3- Evaluation Criteria 17Reliability 17Safety and Environment 18Cost 18

Chapter 4- Power Source Equipment: Cost and Reliability 19Radioisotope Thermoelectric Generators 19Propane-Fueled Thermoelectric Generators 20Photovoltaics 23

Stand-Alone PV Power System (PV/Battery) 26Hybrid Power System (PV/TEG) 27Reliability 27

Summary of Costs and Reliability 28

Chapter 5- Power Source Equipment: Safety andEnvironmental Assessment 31

Radioisotope Thermoelectric Generators 31Licenses and Emergency Plans 32Accident Scenarios 34

Propane-Fueled Thermoelectric Generators 37Accident Scenarios 38

Photovoltaics 39Conclusions 39

(Continued on page VII)

vi

(Continued from page vi)

Figures

Figure 1-1

Figure 2-1

Figure 2-2

Figure 2-3

Figure 4-1Figure 4-2

Tables

Table 2-1Table 2-2

Table 4-1Table 4-2Table 4-3

Timeline of Power Options for the Burnt Mountain Seismic Observatory 4

Layout of Burnt Mountain Seismic Observatory Showing Distances BetweenEquipment 9

Configuration and Measured Power Requirements of Equipment at BurntMountain Seismic Observatory 10

Map of Alaska Showing the Location of the Burnt Mountain SeismicObservatory and Nearby Communities 14

Schematic of a Typical RTG Used for Terrestrial Applications 21Solar Balance for a Hypothetical PV Array in an Arctic Location 25

RTG Models, Power Demand, and Estimated Replacement Dates 12Climate Conditions at Fort Yukon 15

Characteristics of the RTGs at Burnt Mountain 22Examples of Photovoltaic Power Systems Used in Cold Regions 24Summary of Costs of TEG and PV Power Sources as Estimated by the

Air Force 29

Appendix A - Acronyms 42

vii

Summary 1

T he U.S. Air Force operates a seismic observatory on Burnt Mountain in Alaska to helpverify compliance with nuclear test ban treaties. Data from the unattended station, locatedin a remote area about 60 miles north of the Arctic Circle, are used to ascertain whether or -

not seismic activity has been caused by nuclear explosions. The data collection and com-munications equipment at the station is powered by 10 nuclear batteries, called radioisotope thermo-electric generators (RTGs). Each RTG is fueled with between 1.2 and 3.9 pounds of strontium-90(Sr-90), a highly radioactive material. RTGs are used because of their high reliability and low mainte-nance requirements.

In August and September 1992, a tundra fire encroached on the Burnt Mountain site. It damagedsome data cables, but did not disturb the monitoring, communications, and power equipment. Thefire raised public concern among nearby inhabitants about the safety of using a radioactive materialas the power source at the station. To address this concern, Senator Murkowski of Alaska asked theAir Force to inspect the site, conduct public meetings to discuss the risks and advantages of RTGs,and analyze alternative potential power sources for the station. Additionally, Senator Stevens ofAlaska along with Senator Murkowski requested that the Office of Technology Assessment (OTA)undertake an independent evaluation of alternative power technologies for the site. The objective ofthe assessment was to identify a remote power source technology that presents the lowest health andsafety risk to nearby populations and equipment technicians at an acceptable life-cycle cost. TheSenators’ letter stated specifically that the health, safety, and environmental aspects of the systemwere to be given precedence over cost considerations. This background paper examines the safety ofusing RTGs at Burnt Mountain and assesses the viability of using alternative power sources at thestation.

There are three principal issues that must be resolved with regard to the Burnt Mountain SeismicObservatory and its power system. First, should the observatory continue to operate; i.e., is the sta-tion still necessary given the changed face of world security coming with the end of the Cold War?The Air Force has recently been given the responsibility y for monitoring compliance with a worldwidecomprehensive test ban treaty in addition to its previous treaty monitoring duties. Fully monitoredstations such as the one at Burnt Mountain are important to this new assignment. The Air Force

1

2 I Power Sources for Remote Arctic Applications

considers the data from Burnt Mountain to becritical. Thus, this background paper assumesthat the station will remain operational. Second,assuming that the station continues to operate,when should the RTGs be replaced? Should theRTGs remain in place until the end of their use-ful power production at the station or should theybe replaced at an earlier time? The current RTGscould conceivably fully power the observatoryuntil 2009. Several of the units could power theirassociated equipment until 2018 or later. Third,what power system could be used to eventuallyreplace the RTGs? Leading candidate technolo-gies include: modified RTGs, propane-fueledthermoelectric generators (TEGs), and photo-voltaic (PV) systems.

AIR FORCE FINDINGSThe Air Force, at the request of Senator

Murkowski, conducted a study of RTGs and al-ternative power technologies for the BurntMountain station.1 The study concluded that:

. . . continued use of the RTGs is clearly thesafest, most reliable, and most economicalapproach to supplying electrical power tothe Burnt Mountain Seismic Observatory.. . . [The RTGs] should continue to be op-erated until the end of their useful powerlife. The first unit falls below the requiredpower level in 2009. For an added marginof safety it is recommended that com-bustible materials be cleared annually fromthe equipment sites.

A logical plan would be to phase out theRTGs as they reach the end of their usefullifetimes. This approach would also pro-vide the opportunity to field test replace-ment systems without jeopardizing the reli-ability of the observatory operations. . . .[A]t this time, propane-fueled TEGs ap-pear to be the best candidate for immediatereplacement of the RTGs. However, by theend of the projected useful lifetime of theRTGs other, emerging technologies mayprove more economical and safe than theTEGs.

The Air Force’s preference for TEGs stems fromtheir proven track record in applications with cli-mate conditions and energy requirements similarto those at Burnt Mountain. In addition, TEGscould be deployed in a dispersed configurationsimilar to that used by the RTGs now.

A PV system was found to be the next mostviable option. A major design issue with such asystem is how to deliver adequate power duringthe dark winter months in the Arctic. The AirForce examined PVS with two different powerbackup systems for the winter--batteries andTEGs. The stand-alone PV/battery system wasjudged less desirable, because of the expense ofthe large number of batteries required. The highcost covers not only the initial purchase of thebatteries, but also their transport to Burnt Moun-tain. Several other power technologies were ex-amined for the application, but were consideredtoo costly, too unreliable, or unproven.2

1 Wright Laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, “Power System Assess-ment for the Burnt Mountain Seismic Observatory, ” report prepared for the Air Force Technical Applications Center,Patrick Air Force Base, FL, May 1994.

2Considered too costly or too unreliable by the Air Force were: fuel cells, aluminum-air batteries, gasoline-poweredcombustion-engine-driven generators, wind turbine with battery storage, commercial power with a land-line connection.Considered too unproven were: combustion thermionic generators, thermal photovoltaic generators, combustion-drivenstirling generators, microwave power beaming, hydrogen thermoelectric converters, and alkali metal thermoelectric con-verters.

Chapter 1 Summary I 3

OTA FINDINGSOTA, with help from Future Resources

Associates, examined the safety and environ-mental characteristics of RTGs and alternativepower technologies under several accident sce-narios.3 The use of RTGs presents risks to peo-ple and the environment through the possiblerelease of Sr-90. However, the probability ofany accident--with the exception of dedicatedvandalism--causing a release of radioactive ma-terial to the environment is very low. No naturaldisaster presents much risk of causing a releaseof radioactive material to the environment, andmost accidents associated with human activitiespresent little risk of contamination. In the eventthat radioisotope material is released, therewould probably be minimal long-range disper-sal, so that cleanup activities would be able toremove the bulk of the material in the units.Residual radioisotopes in the environmentwould remain embedded in a fairly inert ceramicmaterial, with minimal uptake by plants and in-corporation into the food chain. It appears rea-sonable to conclude that continued operation ofthe RTGs at Burnt Mountain presents minimalrisk to the surrounding area and population.

The use of TEG power systems at BurntMountain would introduce different risks to thefacility. In the event of an accident, TEGs aremore likely than RTGs to damage the station’sequipment and less likely to harm people andthe environment. Propane fuel is flammable andexplosive (in certain mixtures with air), and itsuse would subject the seismic equipment to avariety of risks due to fires and explosions. Ac-cidents could arise in delivering propane fuel tothe remote Burnt Mountain site, and in distribut-ing fuel on the ground at the site, Propane acci-

dents during unattended operation of the obser-vatory can be caused either by natural events,like offsite fires and earthquakes, or by vandal-ism. The TEG power systems would not presentany substantial risks to nearby populations, ex-cept in the event that a propane fuel accidentignites a fire that spreads offsite--an unlike] y oc-currence given the cleared area around the seis-mic facilities,

PV energy systems present minimal risks forthe environment during routine operation, main-tenance, and transportation. There are, however,potential safety and environmental problems as-sociated with PVs--particularly the batteries--inaccident situations. Releases of toxic heavy met -als into the environment is a potential problemAnnual maintenance visits are recommended forPV/battery systems, but no annual fuel deliver-ies are necessary, Of course, if TEGs were usedas the winter backup for the PV system, therewould be additional risks of the sort mentionedearlier. However, since the fuel requirementswould be smaller, the risk would be somewhatlower than for a stand-alone TEG system.

There are many timing variations associatedwith the implementation of altternative powersources at the Burnt Mountain observatory. Fig-ure 1-1 illustrates several timing possibilities fordeploying the three major candidate power sys-tems at the station. The simplest from a logisticsviewpoint is continuing to operate the RTGs(Option 1). Without any changes whatsoever.RTGs could fully power the station until 2009--another 15 years. There are also methods for ex-tending the life of the RTGs. In this vein, the AirForce has considered: discontinuing the use ofnoncrucial communications equipment--specifi-cally the voice frequency responder--at the sta-

3 Future Resources Associates, “Power System Assessment for the Burnt Mountain Seismic Observatory, ” OTAcontractor report, January 1994.

4 I Power Sources For Remote Arctic Applications

Option 1:RTG: continue

Option 2:RTG: remove VF

Option 3:RTG: 4 sites

Option 4:RTG: TEM retrofit

Option 5:TEG*

Option 6:PV/battery* I ,

I , I1990 2000 2010 2020 2030 2040

11 9 9 4 Full RTG Partial RTG

s Transition period ❑ Alternative technology

KEY:RTG = radioisotope thermoelectric generatorsVF = voice frequency responderTEM = thermoelectric moduleTEG = propane-fueled thermoelectric generatorsPV/battery = photovoltaic array with battery backup

*Earliest implementation of replacement technology. Later deployment is also viable.

The black region ■ indicates the time period in which RTGs fully power the observatory. The patternedregion indicates the time period when RTGs can only partially power the observatory. The patternedregion ❑ indicates the transition and/or testing period between RTGs and the rep lacement powertechnology. Note that in this period the RTGs are still onsite. The white region indicates the periodin which replacement technologies fully power the observatory and the RTGs have been removed. TheRTGs are removed at the end of the patterned regions.

SOURCE: Office of Technology Assessment, 1994.

Chapter 1 Summary I 5

tion (Option 2); allowing one of the station’s fivemonitoring devices to shut down (Option 3); andrefitting the RTGs with improved thermoelectricmodules (Option 4).4 These changes could ex-tend the life of the RTGs by three to 23 yearsand could be implemented any time up until2009 with minimal degradation of effectiveness.Propane-fueled TEGs could be implementedvery quickly, possibly by next year (Option 5).The RTGs could not be removed as soon, be-cause they would be needed as backup duringthe startup and associated troubleshooting of theTEG systems. PV/battery systems (Option 6)would require more time to put in place at BurntMountain, because more extensive testing andtroubleshooting is needed. However, if testingwere started in the near future, a reliablePV/battery system could probably be opera-tional in three to four years. Assuming that ar-rangements could be made for the long-termstorage of the RTGs, they could be removedshortly after the PV testing was complete.

CONCLUSIONSContinued use of RTGs at Burnt Mountain

bears low risk for the safety of maintenanceworkers and local populations and for the envi-ronment. In addition, it minimizes costs and fur-ther environmental disruption to the site. TheAir Force’s recommendations for the clearing ofcombustible materials from the equipment siteson a yearly basis are sound. Other useful precau-tions that should be considered are:

. Installing equipment to protect the sites fromlightning strikes.

. Performing a periodic check of the structuralintegrity of the RTG units, assuming that use-ful nondestructive testing can be performedonsite. Such testing would monitor any degra-dation of materials within the RTGs due tolong-term exposure to radiation.

. Installing intrusion monitors at the station thatwould alert Air Force personnel at Fort Yukonand authorities in nearby villages to possibleproblems with vandals or terrorists. Thiswould help reduce the risk of radiation re-leases caused by dedicated vandalism by al-lowing quick response to the situation by AirForce personnel and civilian law enforcementauthorities.

Looking to the eventual replacement of theRTGs at Burnt Mountain, the interrelated factorsof substitute power technologies and replace-ment timing must both be considered. If theRTGs were required to be removed immedi-ately, the only viable replacement power sourcewould be propane-fueled TEG systems. TEGsare the only replacement technology that couldbe installed without extensive testing. The de-ployment of TEGs would introduce the risk ofdamage to the equipment at the station. In addi-tion, there is the high cost of installing TEGsand of transporting the fuel.

If use of the RTGs could be tolerated forthree or four more years or possibly until the endof their useful lives, other power technologiesmay prove viable replacements. Several years ofonsite testing would probably be adequate toprove the suitability of alternative power tech-nologies that do not require ongoing fuel deliv-eries. At present, PV/battery systems appear tobe the most viable nonfuel replacements for theRTGs. PV power generation, which is accom-plished without fuel or moving parts, is inher-ently more reliable than power generation withtechnologies that use conventional hydrocarbonfuels. PV systems currently provide reliablepower for remote, unattended applications in po-lar Alaska and Antarctica. However, only a sur-vey of the solar and weather conditions at BurntMountain and onsite testing of prototype PV de-signs can establish the viability of a PV powersystem for the observatory.

4l, Tew A, Schmidt, Technical operations Division, McClellan Air F“orce Base, ktkr tO the Air ~~r~e ‘rechni~~lApplications Center on the status of Burnt Mountain radioisotope thermoelectric generators, 1992.


The process of investigating the suitabilityof PV systems as an alternative could beginsoon, to develop adequate onsite experience toensure system reliability. The specific system tobe tested should be a stand-alone, decentralizedPV/battery system. In addition, onsite testing oflow-power seismic monitoring and data commu-nication electronics would be helpful. Systemelectronics with decreased power demand wouldfacilitate the use of alternative power systems.They would also extend the life of the RTGs iftheir continued use at Burnt Mountain weredeemed the proper course of action.

Background 2

Thc Air Force is responsible for verifying international compliance with nuclear weaponstesting treaties. The principal test ban agreements in force are the Limited Test Ban Treatyof 1963, the Threshold Test Ban Treaty of 1974, and the Peaceful Nuclear ExplosionsTreaty of 1976. To accomplish this mission, the Air Force Technical Applications Center

(AFTAC) operates and maintains the U.S. Atomic Energy Detection System, a worldwide system ofsensors to detect nuclear explosions underground, underwater, and in the atmosphere and space. Thesystem relies on seismic and hydroacoustic data, satellite information, and collected air and grounddebris. AFTAC is headquartered at Patrick Air Force Base (AFB), Florida, and has 14 detachments,six operating locations, and approximately 70 equipment locations. Burnt Mountain is one of sevensites in Alaska, and 19 sites worldwide, that collect seismic data. The system uses multiple monitor-ing sites and sophisticated signal enhancement (beamforming) data analysis techniques to detect verysmall seismic sources and to distinguish between ground disturbances caused by nuclear weaponstests and those caused by natural geological phenomena such as earthquakes. AFTAC Detachmcnt460 (Det-460), located at Eielson AFB (near Fairbanks, Alaska), is responsible for operating theBurnt Mountain observatory and several other seismic monitoring stations in Alaska.

Changes in world security following the Cold War call into question the necessity of the BurntMountain Seismic Observatory and other Air Force stations like it. The stations were installed tomonitor the Soviet Union during a time when there was no access to Soviet territory. Today, a world-wide network of open seismic stations--including several on the territory of the former SovietUnion--am providing data that vastly supplements the Air Force’s capabilities. The role of AFTAC’snetwork of seismic stations is different in this new environment, but it remains important. AFTAChas recently been assigned to improve its capability to meet the more stringent monitoring require-ments of the Comprehensive Test Ban Treaty currently being negotiated. This new assignment hasincreased the value of the data from Burnt Mountain. The Air Force views the observatory as one ofan integral handful of recording sites that will comprise the core event detection network of the newtreaty monitoring network. These closely monitored sites will be used to cue the open system towhich events need [o be examined more closely. Furthermore, guaranteed future access to data fromseismic stations in the former Soviet Union or near other regions where nuclear testing is likely to

7


occur cannot be relied on. In this background pa-per, the Burnt Mountain Seismic Observatory isassumed to operate for 30 more years.

The option of closing the Burnt Mountainsite is complicated by the possible loss of otherDet-460 seismic stations (or at least the prospectof greatly increased logistical expenses at otherstations). One seismic monitoring station, on theAleutian Island of Attu, can only be reached bycrossing two small bridges that need replacementor extensive repair. Two additional stations arelocated at Air Force radar stations, which providepower and some logistical support. If the radarfacilities were closed for some reason, the seis-mic equipment would have to be powered inde-pendently or shut down. Closing these seismicstations would make the Burnt Mountain site thatmuch more important, while keeping them openwould be facilitated by using whatever replace-ment power system is selected for Burnt Moun-tain.

EQUIPMENT AND OPERATIONSThe principal equipment at Burnt Mountain

consists of the borehole seismometers to collectthe seismic data and a signal multiplexer and aradio to communicate the data offsite for analy-sis. There are five seismometers clustered withina 1.5 mile radius and linked to a central commu-nications station via surface-laid data cable(figure 2-l). The remote terminals (RTs)--identified as sites U1, U2, U3, U4, and U5--eachconsist of a borehole, a seismic sensor, and ametal frame shelter for housing the power gener-ators and associated power conditioning equip-ment (figure 2-2). The remote operating facility(ROF), near site U3, houses the signal multi-plexer and the communications equipment. Datafrom the five sensors are fed to the ROF via datacables, multiplexed into a single data stream,transferred to Fort Yukon, Alaska, via line-of-sight UHF radio, transmitted to Det-460 at Eiel-son AFB via satellite link, and then sent on to

Shed containing the RTGs powering one of the remote seis-mometers in the Burnt Mountain observatory.

AFTAC headquarters and also to McClellanAFB near Sacramento, California. Multiple seis-mometers are used to increase the sensitivity ofthe readings. The five data signals are computerprocessed to diminish noise, allowing clearercharacterization of the seismic activity.

Each equipment shed is fenced in and thesurrounding area is cleared of trees and othervegetation to a distance of 50 feet. There are twoadditional shelters located near the ROF. Oneserves as lodging for maintenance crews and theother houses the station’s all-terrian service vehi-cle.

There are three site visits programmed eachyear: one scheduled maintenance, one scheduledinspection, and one unscheduled maintenance.However, since 1985--when most of the ra-dioisotope thermoelectric generators (RTGs)were installed--there has been an average of sixvisits a year for the purposes of maintenance, in-spection, and orientation. There has never been astation outage caused by the RTGs since the firstone was installed in 1973. There have been sig-

Chapter 2 Background I 9

N(2,480 ft)

(l,519 ft)U1

3.0 miles

/ \

(1,738 ft)

0.9 miles

/

ROF

, k U4

Y / / / / 2.6 miles

\(2,631 ft) (1,693 ft)

not to scale

KEY: ROF = remote operating facility.Numbers in parentheses are elevations given in feet above sea level.

SOURCE: Wright Laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, “Power System Assessmentfor the Burnt Mountain Seismic Observatory,” report prepared for the Air Force Technical Applications Center, March 1994.


R e m o t eoperating

facility nearSite U3

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -II#I Radio data :II link (5w) ::# 12VDC

:I I II Power II c

Site U3: conditioning

t25VDC

III Remote terminal (9w)I II II II It Power Io II It inverter II Im 1

I 129VAC II

I 1I I I 1

$II

: Data@ Voice link ~It multiplexero (2.2w)

(10.1w) :8 tI t- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

1

Sites Power 25VDCU1 , U2, conditioning — Remote terminal (9w)

U4, & U5

NOTES:● 25A/E, 25E, 25F and 10OF refer to the RTGs model numbers.● Remote terminals - Designed in the early 1980’s with Transistor-Transistor Logic (TTL) technology.● Power conditioning units - Designed in 1985 using discrete transistor technology for operation with RTGs.● Data multiplexer - Line Sharing Unit that multiplexes data from the five sites at Burnt Mountain. Designed using Small

Scale Integration-complementary metal-oxide silicon (CMOS), analog, TTL technology.● Inverter - Converts direct current (DC) to alternating current (AC) to provide power for the data multiplexer. Technology

is discrete and TTL,● Voice link - Voice frequency responder for testing the communications circuit. Since an inverter is used to power the

data multiplexer, AC power is available to accommodate and AC powered responder.● Radio data link - Commutronics radio with Small-Scale Integration technology, analog and discrete.

SOURCE: Wright Laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, “Power System Assessmentfor the Burnt Mountain Seismic Observatory,” draft report prepared for the Air Force Technical Applications Center, Patrick AirForce Base, FL, Oct. 29, 1993,


nal outages associated with problems with thepower conditioning electronics, radios, seismicinstruments, and data cable. 1

RADIOISOTOPE THERMOELECTRICGENERATORS

The seismometers, multiplexer, and radioequipment discussed in the previous section re-quire low, but very reliable power. These de-vices are currently powered by 10 RTGs. Thereare two Sentinel 25 RTGs each at sites Ul, U2,U4, and U5, and one Sentinel 25 and one Sen-tinel 100 RTG at the ROF near site U3. Cur-rently, the Sentinel model 25s generate about 9to 20 watts and the model 100 generates about54 watts. The larger RTG is used at site U3 topower the communications equipment as well asthe site’s monitoring device. The RTGs producelow voltage direct current (DC) output; powerconditioning and conversion equipment is usedto convert the power to the necessary voltagesand forms (DC or alternating current--AC). Eachof the five RTs uses electricity in the form of25V DC, The ROF uses electricity in two forms,129V AC and 12V DC. There currently is nopower distribution capability between the fiveRT sites at Burnt Mountain.

RTG technology was developed under theAtomic Energy Commission's “Beneficial Usesof Radioactive Material” program begun in1959, RTGs are used in remote applicationssuch as Arctic stations and space vehicles, wheresmall amounts of highly reliable, low-maintenance power are required. The first RTGwas installed at Burnt Mountain in 1973 andnine additional ones were installed in 1985. Theunit installed in 1973 was purchased directly

from Teledyne. Some of the units installed in1985 came from the Federal Aviation Adminis-tration; others were Navy surplus. They havenever failed to produce power.

The RTGs at, Burnt Mountain differ slightlyfrom one another in terms of their date of fuelingand power capacity. Thus their useful life at thestation also varies. Table 2-1 shows when theRTGs will cease to provide enough power fortheir associated equipment. The RTGs, as cur-rently configured, could fully power the obser-vatory until March 2009. Several of the unitscould be used longer, but all would need to bereplaced or modified by June 2019 at the latest.These are the earliest dates that the generatorswould have to be replaced. Advances in the sys-tem electronics that reduce power requirementsshould outpace RTG power degradation, TheAir Force has considered several options for ex-tending the life of the RTGs. 2

. Discontinuing the use of the voice link used totest the communications circuit. This wouldextend the life of the RTGs at site U3 by 10years, making the RTGs at site US the first toneed replacement (in 201 2).

. Allowing the power to cut off at site US.When coupled with the previous option, thiswould extend the life of the RTGs at the sta-tion until 2018.

. Installing improved thermoelectric modules(TEMs), the internal RTG components thatconvert the heat into power.3 The currentTEMs have an efficiency of about 5 percent:improved ones would be about 15 percent ef-ficient, effectively tripling the power outtput ofthe RTGs,4 With improved TEMs, the RTGscould power the observatory equipment until

1Lt. Col. Terry Fout, letter to Alaska Porcupine Caribou Commission, Nov. 3, 1992.2C01. Terry A. Schmidt. Technical Operations Division, McClellan AFB. letter to the Air Force Technical Applica-

tions Center on the status of Burnt Mountain radioisotope thermoelectric generators.3Refitting the RTGs with new TEMS would require an additional Certificate of Compliance from the Nuclear

Regulatory Commission, See section on Licenses and Emergency Plans in chapter 5 of this report.4Improvements in thermoelectric conversion efficiency depend on the timing of the installation. Were new modules

fabricated today, their efficiency would be on the order of 1() percent. If they were fabricated in 20 years or So, 15 percentefficiency may be possible given the potential advances in thermoelectric material technology.

121 Power Sources for Remote Arctic Applications

Site RTGnumber serial number

U1 817

U2 920

U3 114

U4 1018

U5 419

RTGmodel number

Sentinel 25ESentinel 25E


Sentinel lOOFSentinel 25F


Sentinel 25ASentinel 25E

Present power Site power Estimatedoutput (watts)’ demand _(watts)b replacement dates

10.4 9.0 June 201814.8

9.8 9.2 January 201814.4

42.3 26.2 March 200910.4

9.5 9.0 June 201915.1

6.7 9.0 August 201214.6

a power losses of 20 to percent due to heat losses have been factored in.b Measured loads of the equipment as recorded by the Air Force Technical Applications Center/Technical Operations

Division engineers during a June 1992 site visit.

SOURCE: Office of Technology Assessment based on data from Wright Laboratory, Aeropropulsion and Power Directorate,Aerospace Power Division, ‘Power System Assessment for the Burnt Mountain Seismic Observatory,’ report prepared for theAir Force Technical Applications Center, March, 1994.

2032 to 2047 depending on the particular site.The TEMs do not have to be installed imme-diately to realize these gains. Putting them inanytime before the RTGs otherwise becometoo weak to power their equipment is suffi-cient.

There are other power management techniquesthat could extend the operating life of the RTGs.For example, swapping generators number 14and number 17 would extend the lifetime of theBurnt Mountain RTGs by 21/2A years (to Septem-ber 2011). Also, installing future generations ofseismic and power conditioning electronics,which are likely to require less power, would

extend the useful life of the RTGs. With thesekinds of power management methods and likelyimprovements in system electronics, it is en-tirely possible that RTGs could power the BurntMountain station well into the second if notthird decade of the next century.5

STRONTIUM-90 FUELEach of the RTGs at Burnt Mountain con-

tains between 1.2 and 3.9 pounds of Strontium-90 (Sr-90) fuel. Sr-90 produces the heat neededin the RTGs via radioactive decay, not throughfission or fusion. It is manufactured as a byprod-uct of spent nuclear fuel reprocessing. G Sr-90 has

5John F. Vogt, Sentinel Program Manager, Teledyne Brown Engineering-Energy Systems, personal communication,Mar. 18, 1994.


a half-life of 28 years. This makes it suitable forlong-term power uses. RTGs do not require con-stant refueling.

In the RTGs, the Sr-90 fuel is in the form ofhockey puck-sized disks of strontium titanate(SrTiO3 or SrTiO4), a solid ceramic material.This material was selected in large part for itsstrength, fire-resistance, and low water-volubility.

Sr-90 is the main source of environmentalrisk associated with the RTGs at Burnt Moun-tain. It is not an explosive material, but is haz-ardous if dispersed by other means. Sr-90 and itsonly radioactive decay product, Yttrium-90 (Y-90), are both beta emitters. Beta particles fromSr-90 travel about 10 inches in air and those ofY-90 travel about 100 inches in air. The parti-cles travel shorter distances in liquids andsolids. Close contact with the radioisotopethrough ingestion or inhalation is necessary todeliver a long-term beta dose of radiation to thebody. Beta particles, such as those from decayof Sr-90, also create x-rays (Bremsstrahlung ra-diation) as they are slowed down in the sur-rounding fuel and shielding material. X-rays arepenetrating radiation that can deliver a radiationdose from outside the body.

Sr-90 follows calcium metabolically, and ifingested or inhaled in a biologically availableform (i.e., as a substance that is soluble in thebloodstream or other fluids of a living organism)accumulates in the bones. From there it contin-ues to deliver radiation to associated tissues overits entire radioactive life. This has been found toinduce bone cancer and in some casesleukemia.’ If Sr-90 is ingested or inhaled in arelatively nonbiologically available form--suchas strontium titanate--the material passes fromthe body naturally, delivering primarily a short-term dose of radiation.

LOCATION AND CLIMATEBurnt Mountain is located at 67°25’ north

latitude, 144°36’ west longitude, approximately62 miles north of the Arctic Circle (figure 2-3).It is a remote area about 50 miles from the clos-est villages (Venetie, Arctic Village, andChalkyitsik) and 56 miles from the nearest AirForce facility (Fort Yukon Air Force Station),

The site was chosen for a seismic observa-tory because of its geologic structure and longdistance from human sources of noise. Its hard-rock surface surrounded with soft soil yieldsvery clear seismic signals with minimum noise.

The Burnt Mountain observatory lies in thefoothills of the Brooks Range, so the terrain inthe immediate vicinity is mountainous. There isa 1,100-foot difference in elevation between thehighest and lowest of the sensor sites. The landis tundra that varies from barren rocky ground toareas of considerable cover, mostly white andblack spruce and some aspen, birch, and wil-

View of the ROF site at Burnt Mountain. The RTGs poweringthis site are located inside the structure within the fence.Several kilometers in the distance is a small clearing contain-ing one of the remote seismographs.

csr-go is ~l~o formed in nuclew wea~ns explosions and is a component of radioactive fallout.

TJohn ~]. }]ar]ey, “Radioactive ~a]]out, ” McGraw-Hill Encyclopedia on Science and Technology, 7th ed. (New York,NY: McGraw-Hill, 1992).


) ”

.

SOURCE: Office of Technology Assessment, 1994.

lows. Vegetation has been cleared to a distanceof 50 feet from each of the sensor sites. Discon-tinuous permafrost exists in the area, There isalso wildlife in the area, including barren-ground caribou, grizzly and black bears, moose,and a variety of fur bearers and ptarmigan.8

The climatic conditions at Burnt Mountainare not known precisely; the nearest weather ob-servatory is at Fort Yukon. Climatic conditionsfor Fort Yukon are summarized in table 2-2. Av-erage daily temperatures range from -22°F in the

winter to 63°F in the summer, and extremes of-71° to 1000F have been recorded. Humidity isnil to low year round. Precipitation is moderate.Wind speeds are low year round. Monthly aver-age wind speeds range from 3 to 7 knots. Gustsas high as 35 knots were recorded during the1980s.

The extreme northern location of the sitelimits daylight that might be useful for solarpower. Average daily daylight ranges from 21 to24 hours during the summer and 2 to 3 hours

8c orp~ of Engineer5, Alaska District, Environmental Impacf Assessment, AITAC Project FI1081 (Anchorage, AK:

May 1972).


Temperature OtherLow (“F) High (“F)

Winter (average) -22 -2 Wind (average) less than 5 knots (5.8mph)

Spring (average) 29 54 Wind (gusts) 35 knots (40 mph)Summer (average) 42 63 Precipitation 17 inches annuallyFall (average) -9 6 Humidity lowExtremes -71 100

SOURCES: Wright Laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, “Power System Assessmentfor the Burnt Mountain Seismic Observatory, " draft report prepared for the Air Force Technical Applications Center, Oct. 29,1993; and Letter from Senators Stevens and Murkowski to the Office of Technology Assessment, Oct. 27, 1993.

during the winter. Certain days in the winter,though, receive no sunlight at all. The averagedaily insolation, a measure of the intensity ofsunlight, varies between a low of 0.1 MJ/m2 inDecember to a high of 18.5 MJ/m2 in May.9

There are no roads in the area. All personneland materials are flown in from Fort Yukon viahelicopter. Fort Yukon is the principal stagingarea for Burnt Mountain, but is only marginallymore accessible. Personnel and incidental mate-rials are flown in and bulk supplies are deliveredvia barge. Bulk transport from Fairbanks is byroad to Nenana (50 miles southwest of Fair-

banks) and then by barge on the Yukon River toFort Yukon. There are three barge trips peryear--June, July, and September.

The Alaska District Corps of Engineers con-ducted an Environmental Impact Assessment forsiting the observatory at Burnt Mountain andtwo other locations in 1972. The assessment fo-cused almost exclusively on the environmentalimpact of the construction of the facilities, noton their operation. Impacts discussed clearingvegetation and soil erosion as a result of vegeta-tion removal and road access.l0

9Solar Energy Research Institute (now the National Renewable Energy Laboratory), Solar Raciialion Energy Resource

Atlas of the LJnired Srufes, SERUSP-642-1037 (Golden, CO: October 1981). Data is for Big Delta, AK, with panels tiltedat 64° (latitude) from horizontal.

I°Corps of Engineers, op. cit., footnote 8.

T

EvaluationCriteria 3

here are three principal performance criteria that the power source used for the Burnt Moun-tain Observatory must meet: high reliability, safe and environmentally benign operation,and reasonable life-cycle cost. An overview of these evaluation criteria is presented in thischapter, and the discussion of radioisotope thermoelectric generators (RTGs), propane-

fueled thermoelectric generators (TEGs), and photovoltaic (PV) systems and their ability to meetthese requirements appears in subsequent chapters. The reliability and cost characteristics of thepower systems are assessed in chapter 4; the safety and environmental characteristics are examinedin chapter 5.

RELIABILITYThe importance of the Burnt Mountain station’s mission and the difficult access to the site require

that the monitoring and communications equipment and their power sources be very reliable. It wasthis need for high reliability that led to the use of RTGs as the power source for the site in the firstplace. The reliability of RTGs derives from their lack of moving parts and their ability to providecontinuous power over a very long period without refueling. The RTGs at Burnt Mountain have allbeen in service at the station for over eight years and could conceivably power the seismic sensingequipment for another 15 to 25 years (see table 2-1). In addition, RTGs require little maintenance.

Though a highly reliable power system is certainly very important and very desirable, a certainrisk of power outages can be tolerated. The station collects and transmits signals from five seismicdevices. Loss of one of the five signals should not deteriorate the data signal a great deal, and proba-bly can be endured for a short time. As mentioned earlier, the Air Force has considered letting one ofthe signals cut off permanently as a means of extending the life of the station.1 This indicates some

ICO1. Te~ A. Schrnid[, Technica] operations Division, McClellan Air Force Base, letter to the Air Force Technical

Applications Center on the status of Burnt Mountain radioisotope thermoelectric generators, 1992.


willingness to accept a slightly degraded signalcaused by a power problem, but new Compre-hensive Test Ban Treaty mission requirements(described in chapter 2) may force a revision ofthe data requirements toward higher reliability.Note, however, that the system powering thecommunications equipment must meet a higherstandard of reliability than the other power sys-tems at the station. An outage at the communica-tions link means the loss of all five of the signals,not just one.

When considering alternative power tech-nologies, the need for high reliability puts a highpremium on proven technologies. The field expe-rience of technologies in Arctic conditions is ofgreat importance in establishing their reliability.

SAFETY AND ENVIRONMENTThe safety and environmental characteristics

of the RTGs and alternative power systems forthe Burnt Mountain Seismic Observatory are ofparamount importance. They were the major im-petus for this study. Risks to the local populationand to Air Force personnel that maintain thestation are both of concern. The principal safetyand environmental risks of the power systemsunder consideration stem from:

. exposure to radioactive material (Strontium-90) from RTGs,

. fires and/or explosions connected with propanefrom TEGs, and

. exposure to toxic fumes and heavy metalsfrom the batteries in PV systems.

Each of the power systems would no doubt bedesigned to keep the risks associated with theseevents low. However, the safety and environ-mental risks would not be eliminated; small riskswould remain. The level of risk that can be at-tained is primarily a question of engineering andeconomics. The level of risk that can be toleratedis determined socially and politically.

COSTOne of the most important determinants of

cost is the configuration of the power sources. Ina distributed configuration, each remote terminal(RT) has its own power source. In a central con-figuration, there is one main power source andthe power is transmitted to the RTs via cable.Because of power losses in the transmissionstage, the generator in a centralized system mustbeat least 50 percent larger than the sum of thosein a distributed system.

The present energy system at Burnt Moun-tain uses a distributed configuration, with twogenerators located at each of the five monitor-ing/communications sites of the station. A cen-tralized power system for the observatory, pre-sumably located near the remote operating facil-ity (ROF), would require a network to distributepower to the five RT sites. Transmission cablescould be strung along the rights-of-way alreadyestablished for the data cables, either surface laidor buried. The Air Force study recommends thatif a centralized system were implemented, thepower be transmitted at 120V direct current (DC)to minimize resistive losses. Even so, therewould be significant losses associated with step-ping the voltage up to 120V and then stepping itback down after transmission. The voltage con-version at each end is on the order of 75 percentefficient. The Air Force is investigating more ef-ficient conversion technologies for this applica-tion.

The principal constraints on the operation ofpower sources are the site’s climate and remote-ness and its lack of sunlight in the winter. Theremoteness and the inclement weather makemaintenance of the site costly. There is a pre-mium placed on keeping site visits, especiallythose with large cargo (e.g., fuel or batteries), toa minimum. This is not only for cost reasons butalso for safety reasons. The difficult y of transportalso makes extensive construction of power sys-tem apparatus at the site very costly.

—

Power SourceEquipment:

Cost andReliability 4

RADIOISOTOPE THERMOELECTRIC GENERATORSRadioisotope thermoelectric generators (RTGs) are a type of nuclear battery that uses the Seebeck

thermoelectric effect to generate electric power from the heat of decay of a radioactive material. TheSeebeck effect generates a small electric potential in a thermocouple that spans a temperature gradi-ent. In an RTG, many thermocouples--made of semiconductors--traverse the distance from the hotzone near the nuclear center to the device’s cool outer surface. Decay of radioactive material(strontium-90--Sr-9O--in the RTGs at Burnt Mountain) provides the heat. The thermocouples areconnected together as a thermopile to boost the electrical output to useful magnitude. Because ra-dioactive decay occurs as an intrinsic property of radioisotopes, the fuel charge in an RTG can befixed in place and last many years. RTGs are attractive power systems for remote applications be-cause they have no moving parts, the fuel supply is integral to the system, and the units are sealed,operate passively, and require very little maintenance. This is the basis of their high reliability. Thelevel of power generated depends on the amount of radioactive material the device was fueled with,the length of time since the fueling, and how well the heat flow is focused across the thermoelectricconversion module.

RTGs were developed as part of the effort begun in the late 1950s to find peaceful uses fornuclear materials. The first unit was for a weather station. The Sr-90 used in the Sentinel models atBurnt Mountain is a byproduct of weapons manufacture at Hanford, Washington. It was processedinto strontium titanate at Oak Ridge, Tennessee. Teledyne Isotopes Inc., Nuclear Systems (now Tele-dyne Isotopes Inc., Energy Systems Division trading as Teledyne Brown Engineering-Energy Sys-tems) produced the Sentinel series.

The radioisotope fuel in the RTGs at Burnt Mountain is surrounded by shielding and insulatingmaterials, A schematic cross section of a typical RTG is shown in figure 4-1. The Sr-90 fuel isfabricated as strontium titanate: SrTiOJ in the Sentinel 25 models and as Sr2Ti04 in the Sentinel 100Fmodel. The hockey puck-size material is encased in fuel cladding consisting of a stainless steel linerand a superalloy (Hastelloy C) fuel capsule. This fuel capsule is surrounded by a radiation shieldfabricated from tungsten. This in turn is enclosed in thermal insulation to direct the heat upward tothe thermocouples. Lastly, there is the exterior housing of the unit. Seven of the RTGs at Burnt

19


One of the RTGs powering the ROF site at Burnt Mountain,where data are collected from remote seismographs andrelayed by radio to Ft. Yukon, some 100 km to the south.

Mountain are housed in steel, two in aluminum,and one in cast iron.

The fuel capsules of RTGs have passed strin-gent heat, thermal shock, impact, and projectilestriking tests without developing any detectableleaks of the radioisotope material. The tungstenradiation shield and the housing together are de-signed to reduce radiation levels to a maximumof 10 millirem per hour at a distance of 1 meterfrom the RTG surface. As points of reference, atypical chest x-ray is about 45 millirem, and theaverage annual whole body dose to a person in

the United States from natural and manmadesources is about 360 millirems.1 The RTGs ascomplete units have not been tested. Instead, en-gineering analyses were conducted to demon-strate that the RTG designs met the applicableNuclear Regulatory Commission standards fortransportation packaging.

Nine of the RTGs at Burnt Mountain pro-duce continuous power of 9 to 20 watts. The onelarger unit produces continuous power of 53watts. The smaller units contain approximately1.2 pounds of Sr-90 and the larger one containsabout 3.9 pounds, Each RTG weighs approxi-mately 1 to 2 tons. The units are housed inwooden utility sheds, There are actually fourmodels of RTGs at Burnt Mountain; from small-est to largest (in terms of amount of nuclear ma-terial) they are: one Sentinel 25A at 50,000curies (Ci); seven Sentinel 25Es at 56,000 to 61,000Ci; one Sentinel 25F at 60,000 Ci; and one Sen-tinel 100F at 189,000 Ci. These quantities repre-sent the estimated activities in April 1994. Othercharacteristics of the Sentinel RTGs are shown intable 4-1.

Since RTGs are already in place, the cost oftheir continued operation is far lower than anypossible alternative that would have to be con-structed and installed there. However, modifica-tions to the generators and/or the electronics sys-tems would be needed to enable the RTGs to pro-vide enough power over the projected life ofBurnt Mountain station.

PROPANE-FUELED THERMO-ELECTRIC GENERATORS

Propane-fueled thermoelectric generators(TEGs) generate electricity on the same principleas RTGs, except that propane or some other hy-drocarbon rather than a radioisotope provides theheat. However, unlike RTGs, which can provide

Iwfight ~~ratov, Aeropropulsion and power Dir~torate, Aerospace Power Division, “power System Assessment

for the Burnt Mountain Seismic Observatory,” report prepared for the Air Force Technical Applications Center, PatrickAir Force Base, FL, May 1994; and National Council on RadiatioIl Protection and Measurements, Ionizing RadiafionExposures of the Population of the United States, NCRP report 93 (Bethesda, MD: 1987).

Chapter 4 Power Source Equipment: Cost and Reliability I 21

* 4... THERMOELECTRICCONVERTER MODULE/

8... FINNEDCOOLING HEAD

\

\7... PRESSURE VESSELHOUSING LID

] 3... RADIATION SHIELD1...RADiOISOOPEHEAT SOURCE

(BIO-SHIELD)

2...FUEL CLA

5...THERMAL6... HOUSING INSULATION

FIGURE 2. Schematic Diagram of a TypicalRTG Used for Terrestrial Applications. The diagramillustrates the major features of an R T G .

SOURCE: Wright Laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, ‘Power system Assessmentfor the Burnt Mountain Seismic Observatory, ” draft report prepared for the Air Force Technical Applications Center, Patrick AirForce Base, FL, Oct. 29, 1993.

30 years or more on a single fueling, TEGs re-quire periodic refueling. The Air Force estimatesthat using TEGs in a centralized configuration atBurnt Mountain would consume 6,300 poundsof propane per year.2 In a distributed configura-tion, TEGs would consume 5,000 pounds ofpropane annually, This fuel would have tostored in large tanks and flown in probably twicea year. The storage tanks would most likely beburied, and the containment would need to bedesigned to accommodate shifts in the per-mafrost. In addition, there would need to besome mechanical linkages to control the flow of

propane from the tanks to the TEGs and to pre-vent problems during periods of extreme coldweather.

TEGs are available in essentially the samepower and voltage output ranges as the existingRTGs, and therefore could be directly substi-tuted for the RTGs in the existing shelters anddistributed configuration. Alternative)) TEGscould be installed in a centralized mode, in orderto simplify fuel handling and storage operations.

It is expected that the propane would betransported in bundled 100-pound capacitycylinders via helicopter. Depending on the ac-

2Wright Laboratory, op. cit., footnote 1.


Activity of initialcharge (curies)

Date of initial charge

Activity in April 1994(curies)

Exposure rate athousing surface(millirem/M)

Voltage (V)

Housing material

Weight (lbs)

Dimensions, height xdiameter (inches)

Housing pressurerating (PSI)

Design applications

Sentinel 25A Sentinel 25E Sentinel 25F Sentinel 100F94,000 1o5,000-

109,OOO

1968 1969-71

50,000 56,000-61,000

55 65

3.3

Cast iron

3,000

35x26

500

Tailored to landand shallowwater (300meter depth)applications

3.5

Steel

4,170

42x26

10,OOO

Tailored to deepsea (6,700 meterdepth) applica-tions, or landapplicationswith coolinghead

108,000 329,000

1970 1972

60,000 189,000

75 125

3.5 9

Aluminum Aluminum

1,400 2,720

36x20 46x28

500 500

Tailored to land Tailored to landand shallow and shallowwater (300 water (300meter depth) meter depth)applications applications

SOURCE: Product brochures from Teledyne Energy Systems (now Teledyne Brown Engineering-Energy Systems), 1986


tual power configuration and the lift capacity ofthe helicopters used, four to five helicopter tripswould be needed per year. The handling andtransport of this much propane raises safetyquestions that are discussed in the followingchapter.

TEGs are inherently less reliable than theexisting RTGs because of the need for a moreelaborate fuel delivery system, whereas the heatsource for RTGs is dependent on radioactive de-cay and is therefore completely passive and im-mobile. Nevertheless, TEGs are designed for re-mote operation, including in severe climates,and their performance and reliability have beendemonstrated. 3

The Air Force report recommends deployingTEGs in a distributed configuration in order toenhance overall system reliability, take maxi-mum advantage of existing equipment and facil-ities, minimize fuel use, and avoid the expenseand environmental disruption of installing trans-mission cables that would be required for a cen-tralized configuration.4 However, the study doesnot appear to give adequate consideration to theincreased risks and environmental impacts of theextra fuel distribution operations that will berequired to service the TEGs in a distributedconfiguration. The distributed configuration alsowould have greater environmental impact due tothe need for fuel tank installation at five differ-ent sites, instead of just one.

The Air Force report also fails to give ade-quate consideration to the issue of power supplyreliability regarding TEGs. The report cites dataon the catastrophic failure rate for the machines,but does not develop any data to indicate theirreliability with regard to the far more probable

noncatastrophic failure modes, which includeproblems of flame stability, fuel delivery, andignition.

PHOTOVOLTAICSSolar photovoltaic (PV) panels could be

used to power the Burnt Mountain equipment.PV generators share some of the most desirablecharacteristics of the existing RTGs: no movingparts, no need for fuel deliveries, passive opera-tion, and minimal maintenance that can be per-formed in conjunction with regular service visitsto the site for general maintenance procedures.PV systems present minimal health and environ-mental risks under normal operating conditions.There are, however, risks associated with PVsystems in fires or transportation accidents. Pos-sible releases of corrosive or toxic fumes and/ortoxic metals in such events could create healthand environmental problems. Annual mainte-nance visits are recommended for PV power andbattery systems, but no annual fuel deliveries arenecessary.

PV generators also produce a form of powerthat is very similar to that produced by RTGswith respect to alternating/direct current(AC/DC) characteristics, voltage, and wattage,and PV systems are ideally suited for distributedoperating configurations. Indeed, the majority ofPV installations in current operation are remotepower applications, many in severe environ-ments.5 PV systems have been proven as reliablepower sources for remote, unattended, low-power applications in sites all over the world,including many sites in polar regions (table 4-2).

The most difficult aspect of designing a PVsystem for use at the Burnt Mountain Seismic

3J.H. Doolittle, Development of an Automatic Geophysical Observatory for Use in Antarctica (Palo Alto, CA:Lockheed Missiles and Space Co., Inc., Research and Development Division, May 1986).

4 Wfight Laboratory, op. cit., footnote 1.

me Navy has an installation using RTGs at Fairway Reek, west of Wales, Alaska. Stan Read, EnvironmentalEngineering Assistant for Rural Health, Alaska Department of Environmental Conservation, Fairbanks, personal com-munication, Apr. 18, 1994.


Site

Antarctica

Antarctica

Alaska:8 sites

Canada:Labrador, Newfoundland2 sites

Description

Black Island uplink satellite power system, a hybrid that consistsof three HR3 wind turbines, 8 kW photovoltaic array, three on-demand 1.2 kW closed cycle vapor turbines.

Energy system for the National Science Foundation portable SeaIce Laboratory, consisting of a photovoltaic array and heated withpassive solar collectors.

Hybrid power systems that consist of 720 peak watt photovoltaicarrays, TEGs, and batteries for an Air Force installation.

Solar-powered Obstruction Lighting Systems (SOLSTM) toilluminate power transmission lines.

SOURCE: Compiled by Future Resources Associates, Inc. from information provided by Northern Power Systems.

Observatory is the extreme northerly location ofthe site. For approximately a three-monthperiod, November through January, the sitereceives very little solar resource (insolation).Figure 4-2 shows the solar resource deficitduring the winter months for a possible PV arrayat Burnt Mountain. The observatory mustfunction during this prolonged dark period,when the PV system provides almost noelectrical power.

Two approaches are available for poweringthe observatory during the three-month darkperiod using a PV power system: 1) a PV stand-alone system using a battery storage systemlarge enough to allow operations throughout thedark period and a PV array large enough toperform long-term charging of the batteriesduring the summer, and 2) a hybrid systemcombining PV power in the summer with analternative, supplemental source of powerduring the dark period.

A PV power system for the Burnt MountainSeismic Observatory could be built in either acentral or a distributed configuration. In the cen-

tral configuration the PVwould all be installed atpower distribution system

panels and batteriesa single site, and awould have to be in-

stalled at the site to provide power to each of thefive remote terminal (RT) sites, and to the re-mote operating facility (ROF). In the distributedcofiguration, five different PV generation in-stallations would be developed near each of thefive RT sites, and no electricity distribution sys-tem would be required. The distributed configu-ration is equivalent to the one used with the ex-isting RTG power system.

The Air Force assessment of alternativepower technologies for Burnt Mountain sug-gested that, if a PV system were deployed at theobservatory, it should be designed using a cen-tralized configuration, sited near the ROF site.The report recommends a centralized configura-tion over a distributed configuration because ofpossible solar access problems at some of theRT sites. In fact, however, PV systems are ide-ally suited for distributed configurations, andcould easily be integrated into the existing sys-tem. The centralized configuration greatly in-


5 0 ,I

40- - PV output

[

30- -

20-“

1 0! r Power load

0 I – )

J an, Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Based on insolation data for Beffles, AK (66°55’N, 151°31’VV). The stepped line is the power output averaged over the givenmonth for a PV array consisting of two serial and two parallel PC-4 modules. The straight horizontal line is the power demand ofone of the remote terminal sites at Burnt Mountain. The figure shows that there is power surplus in nine months and a deficit inthree months (November, December, and January).

SOURCE Siemens Solar

creases the total system cost because of the need Mountain if it is at all feasible to do so. PVfor more PV panels and battery capacity, as well arrays for several of the RT sites may need to beas for the installation of a power distribution located a short distance away from the actualsystem. monitoring and communications equipment in

OTA’s contractor, Future Resources order to avoid excessive terrestrial shading.Associates, Inc. contacted the three major PV Only a proper site survey can determine whethersystem packagers in the United States: adequate sites are available near each of the RTIntegrated Power Corporation, Photocomm, sites. Optimal layout of a distributed PV systemInc., and Northern Power Systems. All three should lead to substantial cost savings inrecommended that a distributed configuration be comparison with the centralized system con-used for a PV power system installation at Burnt sidered in the report. The cost savings include:


fewer total solar panels needed, smaller powerinverters and power conditioning equipmentneeded, and much less transmission cableneeded. For the purposes of this backgroundpaper, it is assumed that a distributedconfiguration could be used for a PV energy sys-tem installation at Burnt Mountain.

Stand-Alone PV Power System(PV/Battery)

A stand-alone PV power system would re-quire large battery banks for storage through thedark months of November through February,and PV arrays large enough to both run theequipment and recharge the batteries during theperiod when maximum solar resources are avail-able. These facts make battery design a criticalcomponent of a successful stand-alone PV sys-tem, but there are no technical barriers to thedesign of such a system. Due to the northernlocation of the installation, batteries become amuch more dominant component of the overallsystem than in most remote PV applications.

A battery backup system would entail deliv-ery of a large volume of batteries to the site. Theactual amount would depend on the type of bat-tery used. In addition, the PV panels would needto be sized large enough to recharge the batteriesin the summer in addition to providing thepower needed directly to seismic equipment.The Air Force estimates that 95,000 pounds ofbatteries would be needed for this service. Theestimate is based on the use of lead-acid batter-ies. In addition, it assumes that these batteriescannot be discharged any more than 20 percentwithout the risk of freezing. This means that thebattery storage must be five times the amount ofenergy to be used from the batteries. Under theseassumptions, approximately 40 helicopter tripswould be needed to transport the fully chargedbatteries to the site for the initial installation.

It is recommended by Northern PowerSystems--one of the leading companies involvedin the design and implementation of PV energysystems for cold-weather applications--thatnickel-cadmium (NiCd) batteries be given seri-ous consideration for use at Burnt Mountain.NiCd batteries cost more than conventionallead-acid batteries, but they have a longer life-time, and deliver much higher performance un-der cold-weather conditions with less mainte-nance requirements. They have a higher powerdensity (power per pound of battery) and theycan be discharged more deeply (up to 80 per-cent). This would yield substantial savings inthe transport of the batteries to the site. Anotheradvantage of NiCd batteries is that they requiremuch less maintenance than conventional lead-acid batteries, Over the 30-year expected operat-ing life of the Burnt Mountain Observatory, thebatteries will probably have to be replaced onetime, instead of twice for the conventional lead-acid batteries.6

NiCd batteries are not without disadvan-tages, though. An accidental release of cadmiumwould present potential environmental prob-lems. Cadmium is classified by the Environ-mental Protection Agency as one of the 17 mostdangerous substances if released into the envi-ronment.

The batteries probably could be housed inthe existing shelters, and with adequate insula-tion it should be possible to maintain the batter-ies without a requirement for external heating. Itis possible that the battery cost could representas much as 50 percent of the total installed costof a PV energy system for Burnt Mountain.

Overall system reliability could be enhancedby making, when necessary, a service call to thesite in late October to booster charge the batter-ies with portable generators for the winter haul.This could be performed in years for which

Csealed lead-acid batteries with ~()-year lifetimes are now available. Like NiCds, these would only have to be replaced

once.


cloudier than normal conditions persist duringthe summer months. Remote monitoring of bat-tery charge levels should be easy to accomplish.

A stand-alone PV power system for theBurnt Mountain Seismic Observatory wouldconsist of five separate, isolated installations.Each of the installations would include two tothree PV panels, support structures, power con-ditioning equipment, and NiCd storage batteriesdesigned for seasonal power storage. Each PVpanel, rated at 50 to 60 watts peak, would mea-sure less than 1 square meter in surface area.Four of the installations would be designed toserve continuous loads of 10 watts and peakloads of 150 watts each. The fifth installationwould be designed to serve a continuous load ofabout 30 watts and a peak load of 150 watts.

Actual PV system costs for such an installa-tion can only be determined by performing a sitesurvey, resource assessment, and preliminaryengineering. The regular rule of thumb used inthe industry for estimating PV system installedcosts for complete systems including batteries is$10 to $15 per installed peak watt of capacity.This rule of thumb covers systems designed forconventional applications, and for sites that donot experience prolonged dark periods. Themuch larger and more sophisticated batteriesneeded for an application like Burnt Mountainmean that actual system costs would be muchhigher than conventional PV systems.

Hybrid Power System (PV/TEG)The alternative to designing a system with

adequate battery storage to allow for 100 percentsolar energy production is to utilize a hybridsystem, in which power during the dark periodand other prolonged periods of unavailability ofsolar insolation is provided by an alternativepower source. Propane-fueled TEG power sys-tems are the primary candidate to act as the sup-plemental power source. The use of a hybridsystem, in comparison with a PV/battery sys-tem, allows substantial savings in terms of re-

duced battery and PV panel size requirements,as seasonal storage is no longer necessary, How-ever, a more complicated control system is re-quired.

However, it also means transportingpropane, though a smaller amount than in theTEG alone option. In addition, there would needto be an automatic control system to start theTEG when PV power output was too low. Thereliability of such a control system is a concern,especially its cold starting performance.

In a hybrid PV/TEG system for BurntMountain, the TEG component of the systemwould be sized large enough to carry the loadfully during the dark period. Each of the remoteRT sites (Ul, U2, U4, and U5) would requireabout a 15 watt TEG, while the U3 site, whichwould carry the ROF load as well as the RTload, would require about a 40 watt unit. Thefive TEG units would be expected to operate forapproximately 2,200 hours per year (one-quaterof the year), using approximately 1,250 poundsof fuel per year, which is about 250 gallons peryear. This is one-quarter as much fuel as wouldbe used by a stand-alone TEG power system.

The use of a PV/TEG hybrid power systemin a distributed configuration would require theinstallation of storage tanks for propane fuel andcompressed nitrogen at each of the five RT sitesat Burnt Mountain. Nitrogen would be requiredfor forcing fuel to the TEG units under the verycold weather conditions that are known to occurregularly during the dark period at the BurntMountain site. Electric ignition and flame stabi-lization would also be required for the TEGs,which would be depended on for operation ofthe observatory during the most severe weatherconditions of the year, when site maintenance isalmost impossible.

I ReliabilityPVS are able to provide the required level of

electric service reliability because they are en-tirely passive in their operation and have no


moving parts. In addition, with a PV system, theobservatory’s equipment would be powered di-rectly by the battery system, not the PV moduleitself. The battery system is passive in operationas well, and does not employ any moving parts,It should be easy to install a remote monitoringcapability for the battery charge levels as part ofa PV power system, in order to allow monitor-ing of battery performance by the Air ForceTechnical Applications Center.

Of the three PV system packagers contactedfor this background paper, two favor the use ofa stand-alone PV system for Burnt Mountain,and the other favors the use of a hybrid PV/TEGsystem. OTA’s analysis indicates that a stand-alone PV/battery system has lower cost, greateroperational simplicity, and lower environmentalimpact. For these reasons, this background paperconcludes that a stand-alone system should begiven priority in testing and installation.

After the initial period of insolation andweather data collection, several PVS should beoperated side-by-side with the RTGs in order toestablish their operating reliability and their re-sistance to windloading and snow and icebuildup. Also, the annual service calls to theBurnt Mountain Observatory site should includefull service for the PV energy system. PV ser-vice should include annual maintenance on thebattery system, as well as cleaning of the PVmodule and checking of the module’s supportstructure, wiring, and control systems.

In cases where a module has not receivedsufficient insolation during a given summer pe-riod to store up enough charge for the winterrun, a portable generator could be brought in tobooster charge the battery for continued reliableoperation of the system. However, since trans-porting materials and equipment to Burnt Moun-tain is costly and logistically challenging, the

“ PV system should be designed (sized) so thatbooster charging is required only in extremelyrare instances. Under these conditions, a PVpower system should be able to deliver the de-sired level of reliability.

SUMMARY OF COSTS ANDRELIABILITY

RTGs have proven very reliable powersources for the Burnt Mountain Observatory.The costs of keeping them operating include an-nual leak testing trips (which are accomplishedin conjunction with electronics maintenancetrips) and the $1,500 annual license fees. Therewill be substantial costs to moving the RTGsfrom Burnt Mountain whenever they reach theend of their lifetime.

TEGs are also reliable. There is some con-cern about their cold starting capabilities, butinsulation and line burial should minimize prob-lems in this area. The Air Force estimates thatthe installation costs would be between$430,000 to $880,000 depending on the config-uration (table 4-3).

PV power systems are commercially proventechnologies for supplying the type of powerneeded at Burnt Mountain. PV systems currentlyprovide power for remote, unattended applica-tions in polar Alaska and Antarctica. The AirForce estimates that the installation costs wouldbe about $1 million, owing to the cost of layingthe power distribution system and the large vol-ume of batteries required. These costs could beconsiderably lower if a distributed configurationand NiCd batteries were used. Moreover, thereare no refueling requirements. Annual mainte-nance requirements would be low, but periodicreplacement of batteries and possibly PV panelswould be necessary.

It should be relatively easy to integrate a PVpower system into the existing equipment andelectrical configuration at Burnt Mountain, asPV modules and battery systems produce elec-tricity in similar form--DC--and voltage to theexisting RTG power system. It may be desirableto electrify one RT site with a PV system andoperate it for a year or so while the existing RTGsystem is in place. This approach would help todemonstrate the reliability of the technologyprior to removal of the RTGs from Burnt Moun-tain.


TEG TEG PV[central) (distributed) (central)

InstallationEquipment $49 $69 $203Airlift of equipment and fuel 90 104 177Installation of equipment 192 181 66Installation of power lines 393 393Mangement and engineering 156 76 180

Subtotal $880 $429 $1.020

Replacement (once)Equipment (TEGs PV arrays) 49 69 7Airlift of equipment 26Management and engineering 7

Subtotal 49 69 41

Replacement (twice)

Equipment (batteries) $114Airlift of equipment 98Management and engineering 46

Subtotal o 0 $257

AnnualFuel $10 $8Airlift of fuel 23 21Miscellaneous supplies (29 years) 15Management and engineering 7 6 3

Subtotal $39 $3-I $18

Total present value a $1,110 $632 $1,207

KEY: TEG = thermoelectric generator, PV = photovoltaic.

@ Calculated at a discount rate of 15 percent

NOTE: Cost estimates based on 30-year lifetime of service

SOURCE: Wright laboratory, Aeropropulsion and Power Directorate, Aerospace Power Division, ‘Power System Assessment forthe Burnt Mountain Seismic Observatory, ” report prepared for the Air Force Technical Applications Center, May 1994. Data is fromTables 2.1.7-1, 2.2.7-1, and 227-2

Power SourceEquipment:

Safety andEnvironmental

Assessment 5

The safety and environmental characteristics of the radioisotope thermoelectric generators(RTGs) and alternative power systems for the Burnt Mountain Seismic Observatory are ofparamount importance. In their letter to the Office of Technology Assessment (OTA),Senators Stevens and Murkowski identified the following as the primary evaluation criteria

of the power system:

. . . the health and safety of the nearby population, the health and safety of Air Force Tech-nical Applications Center (AFTAC) maintenance and transportation technicians, andthe environmental impacts to the surrounding area, including anticipated or potential emis-sion of effluents to the environment. (Areas surrounding the site are important wildlifehabitat and subsistence hunting and trapping areas for local populations.)

This chapter examines the three most viable power generating systems under consideration for theBurnt Mountain Seismic Observatory: the existing RTGs, propane-fueled thermoelectric generators(TEGs), and photovoltaic (PV) power systems. The focus is on the safety and environmental impactsassociated with potential accidents at the site. The risks connected with the routine deployment andoperation of the candidate power systems are smaller and therefore given little attention. The accidentscenarios that were examined include offsite fire, earthquake, vandalism, aircraft crash, and transporta-tion of the RTGs out of the site or propane fuel into the site.

The risk analysis in this background paper is limited in scope. The accidents’ initiating events andthe associated critical pathways to environmental and safety problems are analyzed qualitatively. O TAdid not attempt to calculate actual probabilities for the initiating events or the various potential path-ways in the accident chain of events. Thus, this paper can only comment on possible events that couldcause damage, but cannot compare actual risks.

RADIOISOTOPE THERMOELECTRIC GENERATORSThe “fuel” for RTGs, strontium-90 (Sr-90), is the main source of environmental risk associated

with the RTGs. The radiation and toxicological characteristics of Sr-90 are discussed in chapter 2. Itshould be reiterated that Sr-90 is not an explosive material. The fuel, Sr-90, is present in the form of a

31


ceramic material: strontium-titanate (SrTiO3 orSr2TiO4). This material was selected in large partfor its fire resistance and low water-volubilityproperties. In an RTG, the fuel is encased incladding and a capsule, and then surrounded bya radiation shield made of tungsten. The radia-tion shield is surrounded by thermal insulationand encased in a metal housing. RTG fuel cap-sules have passed stringent heat, thermal shock,impact, and projectile striking tests without de-veloping any detectable leaks of the radioisotopematerial. Engineering analyses have been con-ducted on RTG “packages” to demonstrate theiraccident resistance during transportation.

~ Licenses and Emergency PlansRTGs are covered by various federal regula-

tions governing the use, transport, and disposalof radioactive materials, The Air Force TechnicalApplications Center (AFTAC) maintains theQuality Assurance Program (QAP) No. 0772,Revision O for the RTGs at Burnt Mountain. TheQAP is necessary to keep (or obtain) Certificatesof Compliance (COCs) from the Nuclear Regula-tory Commission (NRC). COCs are necessaryfor transport and upgrade modifications ofRTGs. In the future, AFTAC will pay approxi-mately $1,500 per year for administration of theQAP. ’

The RTGs used at Burnt Mountain wereoriginally licensed via the NRC licensing pro-cess. In 1981, the Alaska Department of Healthand Social Services confirmed its awareness ofthe RTGs at the station, When additional RTGswere requested in 1985, a public meeting washeld in Fort Yukon.

Current use of the Burnt Mountain RTGs iscovered by U.S. Air Force (USAF) RadioactiveMaterial Permit No. 09-30272 -lAFP issued bythe USAF Radioisotope Committee, whose au-thority is granted by NRC Master Materials Li-cense No. 42-23539 -0IAF. This permit, whichwas renewed for the period January 29, 1992through October 31, 1994, covers a wide rangeof nuclear materials, not just the RTGs, used byAFTAC. The permit stipulates that the RTGsmust be tested for radioactive leakage and/orcontamination at least once every 12 months.The previous permit required that the RTGs beleak tested once every six months. This waschanged because of the remoteness and difficultyof getting to the site. A threshold of 0.005 mi-croCurie per 100 cm2 is used to indicate a leakingsource, The RTGs at Burnt Mountain have al-ways tested at levels below 0.00005 microCurieper 100 cm2, the detection threshold of the labo-ratory procedure,2 The permittee must reporteach year to the USAF Radioisotope Committeeon the containment (leak test) status of the RTGsand present evidence that the manufacturer’s rec-ommended operating temperature has not beenexceeded.

All RTGs are designed to comply with thefollowing standards for radiation dose rates dur-ing routine operation and transportation and forradiation containment during potential trans-portation accidents:

During operation, allowable radiation doserates are stipulated in Title 10 of the Code ofFederal Regulations (CFR), Part 20, which isenforced by NRC. The threshold amounts of

1until recently, Teledyne Isotopes, Inc., the manufacturer of the Sentinel RTGs, maintained the OAP and held theCOCS for the RTGs at Burnt Mountain. Those certificates and the associated responsibilities have been transferred to theAir Force.

~Though Sr-90 emits only beta particles, these standard assays measure alpha and g aroma levels as well as betalevels. Typical readings in 1991 were: gross alpha activity at less than 0.000002 microCurie per 100 cmz; gross betaactivity at less than 0.000001 microCurie per 100 cm2; and gross gamma activity at less than 0.00005 microCurie per 100cm2. Over the years, the RTGs have on occasion tested higher than these levels, but have always been well below thethreshold indicating leakage. In 1992, the gamma assays were discontinued.

Chapter 5 Power Source Equipment: Safety and Environmental Assessment I 33

radiation in unrestricted areas are levels which,if a member of the general public were contin-uously present in the area, could result in hisreceiving a dose in excess of: 2 millirem in anyone hour, 100 millirem in any seven consecu-tive days, or 0,5 rem in a year.3

. During transportation, allowable radiationlevels are prescribed in Title 49 of the CFR,Part 173, Subpart I, which is enforced by theDepartment of Transportation (DOT). If theRTG “package” emits radiation in excess of200 millirem per hour at any point on its exte-rior and emits more than 10 millirem per hourat 3.3 feet (1 meter) from any accessible exter-nal surface, it must be shipped in a transportvehicle (except aircraft) assigned for the soleuse of the consignee and must meet other re-strictions.

. For potential accidents, radioactive mate-rial containment is covered by Title 10 of theCFR, Part 71 and Title 49 of the CFR, Part173, Subpart I, and is enforced by NRC andDOT, respectively. These provisions set per-formance criteria that packages used for thetransportation of radioactive material mustmeet for conditions of heat (fire), impact, per-cussion, thermal shock, pressure, and leakageresistance. Similar standards are required bythe International Atomic Energy Agency(IAEA) Safety Series 33 guidelines for the safedesign, construction, and use of RTGs. Underthese provisions, the fuel _capsule must retainits original leak tightness in the following: 1)heat (fire) test in which the fuel capsule isheated to 800oC (1472oF) for 30 minutes; 2)impact test in which the fuel capsule isdropped 9 meters (29.5 feet) onto a flat, con-

crete supported steel plate; 3) percussion testin which the fuel capsule is struck by a steelbillet with an impact equivalent to 7 kg (15.4pounds) falling a distance of 1 meter (3.3 feet);4) thermal shock test in which the fuel capsuleis heated to its maximum operating tempera-ture and then plunged into O°C (32°F) waterand submerged for 10 minutes; and 5) pressuretest in which the fuel capsule is subjected to anexternal pressure of 14,500 psi.

An assessment of the potential environmen-tal impact of the use of RTGs was completed in1973 and revised in 1977 by Weiner Associatesfor the Naval Facilities Engineering Command.4

The document examined normal transportationand operation of RTGs on land as well as oceanbottom and surface/near surface locations. It con-cluded there were no adverse environmental im-pacts associated with the transportation and oper-ation of RTGs covered by the Navy’s NRClicense. It found that the radiation levels for theRTGs as packaged for shipment did not exceedthe 200 millirem per hour at the surface and the10 millirem per hour at 3 feet from the surfacecriteria, Based on tests of a Sentinel model 100Ffuel capsule, the RTGs were found to complywith all IAEA Safety Series 33 tests for resis-tance to impact, percussion, heat, thermal shock,pressure, and leakages An extensive engineeringevaluation of the resistance of RTG housings totransportation accidents was also performed (ona different model RTG, a SNAP-21). Accidents,such as a head-on collision with another truck,total burial in earth, chemical attack, truck-traincollisions, and fires were evaluated. It was con-cluded that the impact energy from a head-on

JAn unrestricted ~ea is any area that is not controlled by the licensee for purposes of protection of individuals from

exposure to radiation and radioactive materials, and any area used for residential quarters.AWeiner Associates, Inc., “An Environmental Assessment for the Use of Radioisotope Thermoelectric Generators, ”

project WAI- 104. report prepared for the Department of the Navy, Naval Facilities Engineering Command, NuclearPower Division, May 1973 and revised May 1977.

‘The tests were performed by the U.S. Naval Ordnance Laboratory.


collision could rupture the RTG housing and bio-logical shield, but that the fuel capsule wouldsurvive.

The methods of removal of the RTGs fromBurnt Mountain and their disposal are uncertain.Since the RTGs contain high-level nuclear mate-rial, there is currently no permanent disposal sitethat will accept them. They would have to bekept at a temporary storage site such as the Han-ford facility in Washington until a permanentdisposal site could be located. AFTAC has therequired Certificates of Compliance and the AirForce has an NRC license that allows them to“permit” the receipt, storage, use, and transporta-tion of the RTGs under specific conditions. Noadditional approvals from NRC are required tomove the RTGs from Burnt Mountain.

The Air Force has an emergency proceduresplan in case of a suspected radiological accidentat Burnt Mountain. The procedures are executedwhenever there is an interruption in the datacoming from the site. The first step is to checkthe integrity of the communications electronics,because loss of data mayor may not be caused byproblems with the RTGs. If needed, a helicoptercan be sent to the site or flyovers can be sched-uled for quick damage assessment. The plan in-cludes procedures for tending to any people atthe site and notification of Air Force comman-ders, rescue teams, and radiation safety person-nel.

1 Accident ScenariosOTA has examined five accident scenarios

for RTG power systems. The scenarios are iden-tified by their initiating events: offsitc fire, van-dalism, earthquake, aircraft crash onto the BurntMountain site, and transportation of the RTGsout of the Burnt Mountain site. The first fourscenarios cover accidents possible during thecontinued operation of the RTGs, while the fifthscenario encompasses the risks of removing theunits from Burnt Mountain. It must be noted thatthe units will need to be removed and disposedof at some time--at the latest when the observa-tory itself is taken out of service. Of course, oper-ating RTGs at Burnt Mountain for another 20

years or so, as envisioned by the Air Force, re-duces by nearly half the radioactivity of the Sr-90that must be transported. The amount of radioac-tive material, however, does not decrease overthis period.

Four accident scenarios--those initiated byoffsite fires, earthquakes, vandalism, and thecrash of an aircraft directly into a power-systemenclosure at the Burnt Mountain site--have simi-lar event pathways. However, the relative proba-bilities of the various events for each initiator canvary. Following the initiating event, the potentialsequence of events leading to exposure of thenuclear material to humans is as follows: breachof the RTG housing, breach of the radiationshield, breach of the fuel capsule, air dispersal,water dispersal, and soil contamination. Plant up-take is highly unlikely because the strontium isbound in a ceramic with very low volubility. Un-less the accident compromises the radiationshield, there is no environmental or health im-pact from the event. If the radiation shield isbreached, but the fuel capsule remains intact, theproblem is principally one of worker exposure toradiation. Environmental contamination or expo-sure of people (such as hunters) unaffiliated withthe station to radiation could occur only if theexposed fuel capsule is not removed promptlyafter the radiation shield is breached, or both theradiation shield and fuel capsule are breached.Even if both of these containment layers werebreached, contamination would only occur in theevent of air dispersal or water dispersal. Giventhe stable nature of the strontium titanate and thecleared area around the sites, these events arevery unlikely. More probable in the event of acontainment failure is localized soil contamina-tion. This would be a problem primarily forworkers charged with cleaning up after the acci-dent.

FiresIt is known that forest fires can impinge on

the Burnt Mountain Seismic Observatory site. Afire in the summer of 1992 encroached on thesite, but did little damage to the power systemenclosures or any part of the power systems. A


future fire could sweep across the observatorysite, consuming one or more of the structureshousing the RTG power generating units. In theworst case, it might be possible that an RTG unitmight fall over, strike the ground, and then beexposed to fire.

RTGs have been designed to withstand bothshocks (e.g., by dropping) and fires without therelease of any of the radioisotope fuel. Theprobability of the RTG unit breaking upon a fallwithin the enclosure, or as a result of debrisfalling as the enclosure burns, is extremely small.At worst, the outer casing might crack, and theradiation shield might crack, but the probabilityof the fuel capsule itself being breached issmaller still. RTG fuel capsules have been testedby dropping them 9 meters onto a flat, very hardsurface, without any detectable radiation leak.No conceivable impact of that magnitude couldbe imagined in the event of a fire.

Temperatures in the hottest of forest fires canexceed 2,000oF. This is hot enough to melt somemetals, possibly including the ones in the hous-ing of the RTG units at Burnt Mountain. Thistemperature would have no effect, however, oneither the radiation shield or the fuel itself. Tung-sten has a melting point of 6, 179oF; SrTiO3 andSr2Tio4. have melting points of 3,704°F and3,380 +/- 36°F, respectively.b The heat of a2,000°F forest fire would not melt or volatilizeany of the radioisotope fuel contained in theRTGs at Burnt Mountain. Indeed, there is verylittle risk that any combination of impacts andheat exposure that could be experienced by anRTG in a worst-case forest fire would cause arelease of the radioisotope fuel from the units.Even a breach of the radiation shield is extremelyunlikely, and could only be the result of an im-pact. A breach of the radiation shield might pre-sent a risk of radiation exposure to the workersinvolved in the cleanup that would follow a fire,

but little risk to others. Worker exposure can beminimized using standard industry practices.

EarthquakesEarthquakes and most vandalism scenarios

for accidents to the RTGs at Burnt Mountain pre-sent mainly risks of various types of impacts tothe RTG energy systems. In the case of an earth-quake, the risks are that the RTG units will fallover, and that the power system enclosures willcollapse on them. Even in the worst case earth-quake, it would appear that the chances ofbreaching the biological shield are extremelysmall, and the chances of breaching the fuel cap-sule are smaller still. Cleanup from an earth-quake would probably be easier and safer thancleanup from the worst-case fire.

VandalismVandalism of the RTGs can be divided into

two categories: casual vandalism and dedicatedvandalism. Casual vandalism is defined as thetypes of acts that might be committed by peoplepassing through the site for other purposes, notprepared beforehand to assault the RTGs. Dedi-cated vandalism is defined as terrorist actsplanned and carried out against the RTGs. Actsof casual vandalism might include the shootingof hunting rifles at the RTGs, and attempts todislodge and move the RTGs from their fixcdpositions in the enclosures. Acts of dedicatedvandalism might include deliberate burning ofthe shelters and dynamiting of the RTGs.

Casual vandalismNo conceivable act of casual vandalism

would cause a breach of-the fuel capsule or a leakof the radioisotope from the RTGs. Penetrationof the radiation shield is also highly unlikely, andwould present only slight risks for maintenanceand repair workers. In fact, it is unlikely that the

6s J. Rlmshaw and E.E, Ketchen, Slron~ium. 9~ ~afa Sheef~, ORNL-4358 (()& Ridge, ~: OdC Ridge Nat ional

Laboratory, 1969).


shooting of a bullet into an RTG would have anyimpact on its operation. It is also unlikely thatcasual vandals would be able to dislodge theRTGs from their moorings, as the smallest of theRTG units at Burnt Mountain weighs approxi-mately three-quarters of a ton, Even if a unit wereknocked over, it is unlikely that any damage tothe housing or its contents would result. Thus,casual vandalism presents virtually no radiationrisk at the Burnt Mountain Observatory.

Dedicated vandalismFor RTG accidents caused by offsite fires,

earthquakes, and casual vandalism, breach ofboth the radiation shield and the fuel capsule ofan RTG and release of radioisotope material ishighly improbable. However, in the case of dedi-cated vandalism, the situation might be different.It might be possible for a sophisticated terroristto bring sufficiently high-powered explosives tothe Burnt Mountain site to damage and breachthe” radiation shield, and possibly the fuel capsuleas well. In the event of a breach of the radiationshield, but not the fuel capsule, the major riskwould be to cleanup workers who would recoverthe capsule for removal and disposal, In the eventthat the fuel capsule itself is breached, the ra-dioisotope material would be exposed to the en-vironment. It is unlikely that there would besubstantial airborne dispersal, as the Sr-90 wouldremain bound in the form of solid strontium ti-tanate. The titanate might break apart, but mostof the particulate would be large, and thus wouldbe likely to fall no farther away than the otherdebris from the explosion. The range of dispersalwould depend on the power of the explosion. Ifthe strontium titanate is released, some soilwould be contaminated. However, strontium ti-tanatc is highly insoluble, so it would not tend tomigrate deep into soils, nor contaminate water-ways in dissolved form. Water runoff in theevent of a rainstorm following an act of vandal-ism could carry and disperse some of the materialaway from the site. Due to the biological unavail-

ability of the titanate fuel form, plant uptake ofSr-90 and entry into the food chain would beminimal.

Aircraft and Transportiation AccidentsIf a decision is made to phase out the RTGs

soon, or in any case at the end of the useful lifeof the Burnt Mountain Seismic Observatory, itwill be necessary to remove them from the site.The Air Force expects that the units will bemoved by helicopter out of Burnt Mountain, witheventual storage at the Hanford site in the state ofWashington. Handling accidents with the RTGsat the Burnt Mountain site should not be a prob-lem. The major sources of risk considered arehelicopter flight accidents and crashes. Bothtypes of accidents present risks of breaching theradiation shield and the fuel capsule containingthe Sr-90 fuel. The possible sequence of eventsfollowing a breach of the radiation shield and/orfuel capsule are the same as those discussedabove.

The two major risks to the integrity of theRTG in the event of a helicopter accident are:

. the explosive force of a major fuel explosion,either in mid-air or on impact with the ground;and

. the impact force of the entire RTG unit or thefuel capsule falling a long distance from anairborne helicopter,

The force necessary to breach the RTG hous-ing and radiation shield would be considerablyless than that needed to breach the fuel capsuleitself. This is so for two reasons: first, the metalcasing and tungsten radiation shield would ab-

sorb much of the energy, shielding the fuel cap-sule; and second, the fuel itself is encased i nstainless steel and a very ductile and rupture-re-sistant nickel alloy, Hastelloy C. If the fuel cap-sule remains intact, it presents a radiation riskmainly to cleanup workers, who would recover itfor proper disposal. If the capsule is breached,


the radioisotope material would be exposed tothe environment. In no event would there be anymelting of the ceramic fuel material-strontiumtitanate-although some of the fuel could crum-ble, especially in the case of the explosion sce-narios. It is unlikely that a substantial fraction ofthe strontium titanate would be converted intofine particulate, so in the event of a ground-levelbreach of the capsule, the great bulk of the mate-rial would fall out within a short distance of theaccident. An explosion in the sky could cause amuch wider dispersal of the material.

All of the dispersed radioisotope materialwill remain in the form of the ceramic materialstrontium titanate, which is highly resistant todissolving into water, and relatively inert withrespect to uptake by plants. Most of the contami-nation near a ground-based accident should beable to be recovered by a cleanup crew. The re-mainder of the material would present a riskmainly to animals that come into close contactwith it, as the beta radiation travels no more thanabout 10 inches through the air. In addition to therisk of helicopter accidents during an RTG re-moval trip, there is also the risk of a helicoptercrashing into an RTG at Burnt Mountain duringa routine maintenance and inspection trip. Thesequence of events following the impact of acrash would be the same as that following anRTG removal accident.

PROPANE-FUELED THERMOELEC-TRIC GENERATORS

Delivery, handling, and storage operationsfor the propane fuel for TEGs are the main sourceof environmental risk associated with the use ofTEGs at the Burnt Mountain Observatory.Propane fuel would be stored in onsite tanks thatcould be either buried or installed above ground.Fuel would have to be brought to the site fromFort Yukon, some 50 miles away, by helicopter.Access by land is difficult, even via all-terrainvehicle, because there are no roads and the tundrais fragile. The tanks must be able to store more

than a year’s supply of fuel in order to allow refu-eling to be accomplished on an annual basis.Semiannual refueling operations may also beconsidered. A system of piping and valves deliv-ers the propane from the fuel tank to the combus-tor, which is part of the generator unit. TEGs andpropane storage tanks are designed to be safefrom damage by dropping and fire.

Propane is a high-volatility hydrocarbon fuelthat can be liquefied at ambient temperature un-der moderate pressure conditions. It is stored andhandled in liquid form. The major environmentalrisks associated with propane fuel are fire andexplosions. Propane is not considered to be atoxic substance, and any material spilled that isnot burned will evaporate into the atmosphere.The boiling point of propane at atmosphericpressure is -44oF.

Four accident scenarios are examined forTEG power systems. The scenarios are identifiedby their initiating events: offsite tires, vandalism,fuel transportation and handling, and equipmentand material transportation and construction. Thefirst three of the scenarios cover accidents thatcould occur to the TEG energy systems duringtheir operation and maintenance. The fourth sce-nario involves the environmental risks encoun-tered during the project implementation phase ofinstalling the TEG energy systems. Transporta-tion and construction risks are similar to thosethat would be encountered with the implementa-tion of any new energy system at Burnt Moun-tain.

The major source of environmental risk forthe TEG energy systems is the heat source.Propane, a hydrocarbon fuel, is highly volatileand flammable, and in some conditions, explo-sive. It is stored under pressure in order to main-tain it as a liquid, so it tends to escape quickly inthe event of a leak. The only real concerns aboutthe use of propane fuel at Burnt Mountain are inconnection with fires and explosions. Fuel thatescapes but does not burn will simply evaporateand dissipate in the atmosphere. The small

381 Power Sources for Remote Arctic Applications

amounts of fuel that will be stored onsite at anyone time should not be a concern with regard toregional atmospheric hydrocarbon pollution,

I Accident ScenariosFires

In the event of a major offsite fire, it is possi-ble that both the fuel lines and above-ground fueltanks would fail, even though they would be de-signed to withstand such conditions. At the least,the onsite fuel would be consumed in the fire; atworst, there could be an explosion. Burial of thefuel tanks would help to lessen the probability ofa major explosion, although it would entail in-creased environmental disruption at the site dueto the need for digging into and disrupting thepermafrost.

The presence of propane fuel at the BurntMountain site would increase the risk of damageto the seismic sensing facility in the event of amarginal fire encroaching on the site, includingpossibly a fire similar to the one that passedthrough the area in the summer of 1992. Shouldthere be a breach of either a fuel line or storagetank, the presence of propane fuel at the sitecould spread the fire further than it would other-wise go, and/or cause explosive damage thatwould not otherwise occur.

VandalismCasual vandalism at the Burnt Mountain Ob-

servatory could cause a rupture of either the fuellines or fuel tanks at the site, resulting in the re-lease of propane fuel and system shut down. Theleaking fuel would then present a substantial riskof tire and explosion, depending in large part onthe presence of a spark or other ignition source.The ignition source could be supplied by eitherthe vandal, the site’s existing power system, oranother natural source. Propane is very easily ig-nited, and explosive at low concentrations in theair. If the propane leaks away without ignition, it

will simply evaporate and disappear, causing notoxic effects at the site, and leaving no residualsonsite,

Delivery and Handling of PropaneThe propane fuel must be flown to the Burnt

Mountain site via helicopter from the Air Force’sstaging area in Fort Yukon, more than 50 milesaway from the seismic observatory. Each step inthe fuel delivery process-handling at FortYukon, inflight transportation, and handling atBurnt Mountain-entails some risk of fire andexplosion. Standard safety practices have beenestablished for handling and using propanefuel-indeed propane is safely used and trans-ported in a wide variety of everyday situations,but residual risks would remain at each step.

Two different configurations are under con-sideration for TEG deployment at the BurntMountain Observatory: 1) a centralized configu-ration in which the TEGs are installed near theU3 site, where the data multiplexer and transmit-ter are currently located, and a power distributionnetwork must be installed in order to deliverpower to each of the remote terminal (RT) sites,and 2) a distributed configuration equivalent tothe current RTG power system, in which TEGsare deployed at each of the RT sites, and no elec-tricity distribution system is required. The cen-tral-generator configuration would consumeapproximately 25 percent more fuel than the dis-tributed configuration because of the need tocover electricity distribution losses, and bringingin this much additional fuel would increase therisks of fuel supply accidents by some increment.However, deployment in a distributed configura-tion would require fuel to be distributed on theground to each of the RT sites from the centralheliport staging area, which would present an en-tire set of risks that do not pertain to the centralpower-generating configuration.


Deployment of a TEG Power SystemThe deployment of a TEG energy system at

Burnt Mountain entails environmental risks, Allof the material needed for the deployment of thesystem must be flown in by helicopter, includingthe TEG generators and the tankage, as well asall pertinent parts and equipment. Each heli-copter flight entails a level of risk, both in flyingto the Burnt Mountain site and in landing at thesite. The distribution of equipment and materialat the site would entail risk of environmental dis-ruption, as would installation operations, particu-larly the installation of the fuel tanks. Theconfiguration used for the TEG systems wouldhave a major influence on the types of onsite en-vironmental impacts that are of concern. With adistributed configuration, there would be instal-lations of generators and fuel tanks at each of thefive RT sites. With a centralized configuration,there would be only one power system installa-tion, but this configuration would also requirethe installation of the power distribution system.

PHOTOVOLTAICSPV energy systems present minimal risks for

safety and the environment under routine operat-ing and maintenance conditions. During thetransportation and construction phases of deploy-ing a PV system at Burnt Mountain, the risks aresimilar to those of TEGs. There are also safetyand environmental risks ‘associated with damageto the batteries in PV systems caused by trans-portation accidents, fires, and bullets—shoteither accidentally or with malicious intent.Breach of the batteries could release toxic fumesand lead, nickel, cadmium, or other heavy metalsinto the environment. These heavy metals pose avariety of environmental health hazards. If theycontaminate the air, water, or soil, human expo-sure might occur through breathing dust, throughskin contact with the soil, or through ingestion ofwater or contaminated plants or animals. For ex-ample, breathing cadmium can cause damage tothe lungs and kidneys, while long-term exposureto cadmium through ingestion can result in harmto the kidneys and bones. Given the relativelysmall scale of use of batteries, the low likelihood

of a significant breach and subsequent transportto the environment, and the distance from anylocal population, these risks are very small. Spe-cial attention to the structural integrity of the bat-tery containment vessel could reduce, but notentirely eliminate, the risk of contaminating thesurrounding soil and water sources. Additionalrisk arises if booster charging of the batteries isever required. This is the risk of a helicopter acci-dent while transporting the fuel for the chargingequipment.

CONCLUSIONSThe three power systems (RTGs, TEGs, and

PVs) examined in this background paper all in-corporate human safety and environmental qual-ity as important design criteria. During routineoperation and maintenance, the three systemspresent little risk to the Burnt Mountain environ-ment and to the safety of maintenance personneland nearby populations. The safety and environ-mental risks are also very low in most accidentand vandalism situations.

The risk associated with RTGs is the possi-ble exposure of radioactive material (Sr-90) tohumans, plants, animals, and the environment.The radiation exposure received from physicalproximity to an operating RTG is very low. TheRTGs are designed such that radiation levels areless than 10 millirem per hour at a distance of 1meter from the RTG surface. At the maximumrate of 10 millirem per hour, exposure for 41/2hours would yield a dose equivalent to a typicalchest x-ray. Of greater concern is the possibleexposure caused by breach of the inner shields ofthe RTG units and release of Sr-90 into the envi-ronment. Natural disasters and most accidents as-sociated with human activities present little risk

of such release of radioisotope material to theenvironment. Dedicated vandalism presentsgreater risk in this regard, but measures can betaken to lower the risk somewhat. In the eventthat radioisotope material is released, there will

probably be minimal long-range dispersal, socleanup activities should be able to recover mostif not all of the radioisotope. Residual Sr-90 ma-terial in the environment will remain in a fairly


inert form (strontium titanate ceramic), with min-imal uptake by plants and incorporation into thefood chain.

The risks associated with TEGs are the possi-ble fires and/or explosions connected withpropane, Propane is highly volatile and flammable,and is explosive under some conditions. Propaneaccidents can happen in delivering the fuel to theBurnt Mountain site, and while distributing fuelon the ground at the site. Propane accidents dur-ing unattended operation of the observatory canbe caused either by natural events, like fires andearthquakes, or by vandalism. The construction

phase of installing a TEG system could alsocause environmental impacts at Burnt Mountain,but these could be minimized with proper design.

PV systems, while benign in most respects,are not without safety and environmental risks.With them, the risk is the possible release oftoxic fumes and heavy metals from the batteries.Such releases could result from damage to thebatteries caused by bullets, fires, or transporta-tion accidents. There are also risks of helicopteraccidents in deploying and maintaining a PV sys-tem, as there are with other systems.

A p p e n d i x A :

Acronyms

ACAFBAFTAC

CFRCiCMOSCOC

DCDet-460DOT

IAEA

NiCdNRC

QAP

PV

ROFRTRTG

Sr-90

TEGTEMTTL

USAF

Appendix-A

Acronyms

alternating currentAir Force BaseAir Force Technical Applications Center

Code of Federal Regulationscuriescomplementary metal-oxide siliconCertificate of Compliance

direct currentAir Force Detachment 460Department of Transportation

International Atomic Energy Agency

nickel-cadmiumNuclear Regulatory Commission

Quality Assurance Program

photovoltaic

remote operating facilityremote terminalradioisotope thermoelectric generators

strontium 90

thermoelectric generatorthermoelectric moduleTransistor-Transistor Logic

U.S. Air Force

power sources for remote arctic applications

Documents