# Consolidated Record of Decision for Tritium Supply and Recycling

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A99-12019

## Record

- **Collection:** Federal Register
- **Document type:** Notice
- **Published:** May 14, 1999
- **Citation:** 64 FR 26369

## Text

DEPARTMENT OF ENERGY

Consolidated Record of Decision for Tritium Supply and Recycling

AGENCY: U.S. Department of Energy.

ACTION: Consolidated Record of decision for tritium supply and
recycling.

-----------------------------------------------------------------------

SUMMARY: The U.S. Department of Energy (DOE) completed the Tritium
Supply and Recycling Final Programmatic Environmental Impact Statement
(PEIS) (DOE/EIS-0161) in October 1995. The Tritium Supply and Recycling
PEIS assessed the potential environmental impacts of technology and
siting alternatives for the production of tritium for national security
purposes as well as the impacts of constructing a new Tritium
Extraction Facility (TEF) at the Department's Savannah River Site near
Aiken, SC.
On December 5, 1995, DOE issued a Tritium Supply and Recycling
Record of Decision (ROD) [60 FR 63878] that selected the two most
promising alternative technologies for tritium production and
established a dual-track strategy that would, within 3 years, select
one of those technologies to become the primary tritium supply
technology. The other technology, if feasible, would be developed as a
backup tritium source. Under the dual-track strategy, DOE would: (1)
Initiate the purchase of an existing commercial reactor (operating or
partially complete) or irradiation services with an option to purchase
the reactor for conversion to a defense facility; and (2) design,
build, and test critical components of an accelerator system for
tritium production. Any new facilities that might be required, the
production-scale accelerator and a Tritium Extraction Facility to
support the commercial reactor alternative, would be constructed at
DOE's Savannah River Site. Subsequent to the PEIS and the December 5,
1995 ROD, DOE prepared three site-specific EISs: the Accelerator
Production of Tritium at the Savannah River Site (APT) (DOE/EIS-0270),
the Production of Tritium in a Commercial Light Water Reactor (CLWR)
(DOE/EIS-0288), and the Tritium Extraction Facility at Savannah River
Site (TEF) (DOE/EIS-0271). The December 1995 ROD also stated that,
although it was rejected as a reasonable long-term supply alternative
in the PEIS, DOE's Fast Flux Test Facility (FFTF) at the Hanford
Reservation in Washington would be re-examined to determine whether it
should play any tritium production role.
On December 22, 1998, the Secretary of Energy announced his
selection of the commercial light water reactor alternative as the
primary tritium supply. This consolidated Record of Decision documents
that decision and announces a series of three tiered decisions which,
taken together, comprise the Department's plans for establishing a new
domestic source of tritium to support the nuclear weapons stockpile.
Each decision results from the preparation of a related environmental
impact statement. In the order presented, this consolidated record of
decision makes the following decisions based on their associated
environmental impact statements (EIS):
1. Supplemental Programmatic Decision for Tritium Supply and
Recycling: Documents the Secretary of Energy announcement of December
22, 1998; selects the purchase of irradiation services using commercial
light water reactors as the primary tritium supply technology; and
designates the accelerator system at the Savannah River Site as the
backup technology. This ROD supplements the December 1995 ROD described
above. Environmental analysis is contained in the Tritium Supply and
Recycling PEIS (DOE/EIS-01621, October 1995).
2. Site-specific Decision for the Production of Tritium in a
Commercial Light Water Reactor. Selects the Tennessee Valley
Authority's (TVA) Watts Bar Unit 1, Sequoyah Unit 1, and Sequoyah Unit
2 reactors for use in irradiating tritium-producing burnable absorber
rods (TPBARs). This decision is tiered from and implements the
supplemental programmatic decision described above. Environmental
analysis is contained in the Final EIS for the Production of Tritium in
a Commercial Light Water Reactor (DOE/EIS-0288, March 1999). This EIS
is tiered from the Tritium Supply and Recycling PEIS.
3. Site-specific Decision for Construction and Operation of a
Tritium Extraction Facility at the Savannah River Site. Selects the
alternative that would design, construct, test, and operate a new TEF
in the H-Area immediately adjacent to and west of Building 233-H at the
Savannah River Site. This facility is an essential element of the
system for producing tritium using commercial reactors. This decision
is tiered from and implements the supplemental programmatic decision
described above. Environmental analysis is contained in the Final EIS
for Construction and Operation of a TEF at the Savannah River Site
(DOE/EIS-0271, March 1999) which is tiered from the Tritium Supply and
Recycling PEIS.
4. Site-specific Decision for the Accelerator Production of Tritium
(APT). Selects the specific location at the Savannah River Site and the
technologies to be used for the backup tritium supply technology,
should its construction be required. This decision is tiered from and
implements the supplemental programmatic decision described above.
Environmental analysis is contained in the Final EIS for Accelerator
Production of Tritium (DOE/EIS-0270, March 1999) which is tiered from
the PEIS.

FOR FURTHER INFORMATION CONTACT: For further information on the
commercial reactor program and the Tritium Extraction Facility, contact
Stephen M. Sohinki, DP-62, 1000 Independence Avenue SW, Washington, DC
20585, by phone (202-586-0838), or electronically (Tritium web site:
www.dp.doe.gov and click on ``Tritium Project Office Home Page'') For
further information on accelerator production of tritium, contact
William P. Bishop, DP-61, 1000 Independence Avenue SW, Washington, DC
20585, by phone (202-586-0046).
For general information on the DOE National Environmental Policy
Act process, please contact: Carol M. Borgstrom, Director, Office of
NEPA Policy and Assistance, EH-42, U.S. Department of Energy, 1000
Independence Avenue, SW, Washington, DC 20585, (202) 586-4600 or leave
a message at (800) 472-2756.

SUPPLEMENTARY INFORMATION:

I. Background

DOE has prepared this consolidated ROD pursuant to the Council on
Environmental Quality (CEQ) regulations for implementing the procedural
provisions of the National Environmental Policy Act (NEPA)(40

[[Page 26370]]

CFR 1500-1508) and the DOE NEPA regulations (10 CFR part 1021). This
ROD is based on the Tritium Supply and Recycling Programmatic
Environmental Impact Statement (PEIS), and the three site-specific EISs
identified above. Non-environmental considerations such as cost,
technical maturity, and policy issues are also discussed in this ROD.
The Department of Energy is responsible for supplying nuclear
materials for national security needs and for ensuring that the nuclear
weapons stockpile remains safe and reliable. Tritium, a radioactive
isotope of hydrogen, is an essential component of every nuclear weapon
in the current and projected U.S. stockpile. Unlike other materials
used in nuclear weapons, tritium decays at a rate of 5.5 percent per
year. Accordingly, as long as the Nation relies on nuclear weapons,
tritium in each weapon must be replenished periodically. Currently, the
U.S. nuclear weapons complex does not have the capability to produce
tritium to support the Nation's stockpile.
The President's Nuclear Weapons Stockpile Plan sets forth national
security requirements for the current and projected nuclear weapons
stockpile. At present, this plan is based on the Strategic Arms
Reduction Treaty (START I) between the U.S. and former Soviet
Republics. START I, which was signed in July 1991 and became effective
in December 1994, reduces the number of strategic nuclear weapons in
each side's stockpile. Under the Presidential guidance, new tritium
would be needed by about fiscal year (FY) 2005 to offset the decay of
tritium in the stockpile, in the required 5-year reserve, and in
various operating inventories. Although the actual requirement is
classified, the unclassified representation of the steady-state
production rate to offset decay would be about 2.5 kilograms per year.
If needed to replenish the tritium inventory, the new tritium source
should be able to achieve a maximum production rate of around 3
kilograms per year. The START II agreement, which further reduces
nuclear stockpiles, was signed in July 1991, but has not been ratified
by Russia and is, therefore, not in force. If Russia ratifies START II,
the date when new tritium is needed may be as late as 2011 and the
steady-state production rate may be as low as about 1.5 kilograms per
year.
The Department has not produced any new tritium since the shutdown
of the last of its nuclear materials production reactors in 1988. Since
that time the Department has examined various methods of producing new
tritium. The Department announced on November 11, 1991, that analyses
of tritium production alternatives would be incorporated into a
programmatic environmental impact statement for the Reconfiguration of
the Nuclear Weapons Complex. On October 28, 1994, the Department
announced that a separate PEIS for Tritium Supply and Recycling would
be prepared (59 FR 54175). On October 27, 1995, the Notice of
Availability of the Final PEIS was published (60 FR 55020). Following
publication of the Final PEIS, a Record of Decision was issued on
December 5, 1995, which stated that the Department would pursue a dual
track on the two most promising tritium supply alternatives: (1) to
initiate the purchase of an existing commercial reactor (operating or
partially complete) or irradiation services with an option to purchase
the reactor for conversion to a defense facility; and (2) to design,
build, and test critical components of an accelerator system. Within a
three-year period, the Department would select one of the tracks to
serve as the primary source of tritium. The other alternative, if
feasible, would be developed as a backup tritium source. The ROD
further stated that the Savannah River Site is selected as the location
for an accelerator, should one be built. The ROD also stated that a
tritium extraction facility will be constructed at the Savannah River
Site if a commercial reactor alternative becomes the primary tritium
source. Finally, the ROD stated that the existing tritium recycling
facility at the Savannah River Site would be consolidated and upgraded.
In the December 1995 ROD, the Department indicated that the FFTF,
which had been rejected as a reasonable long-term production
alternative, would be re-evaluated to determine whether it could
reasonably play any role in meeting future tritium requirements. In
January 1997, the reactor was placed in a stand-by status while
additional evaluations were conducted. At the time, placing the reactor
in a stand-by condition was thought to provide near-term insurance
while the study of the two dual-track options continued.
On December 22, 1998, the Department announced that commercial
light water reactors would be used for the production of new tritium
and the accelerator would be developed, but not constructed, as the
backup technology. Selection of the commercial light water reactor
confirms the prior plan to construct a new TEF, an element of the
system to produce tritium using reactors. The use of existing
commercial reactors was chosen as the preferred alternative. In
addition, the Department decided that the FFTF would have no role in
tritium supply plans because the Department has high confidence that
the primary and back-up roles assigned to the commercial light water
reactor and accelerator technologies, respectively, would assure that
future tritium requirements are met.
During the 30-day waiting period following publication of the three
project-specific EISs in March 1999, DOE received four letters. One
from the Department of Human Health and Services regarding the Final
EIS for the Tritium Extraction Facility. That letter stated that the
potential concerns of the Department of Human Health and Services were
addressed in the Final EIS, and that there were no additional comments.
The second letter was received from the Department of the Interior
regarding the Final EIS for Accelerator Production of Tritium at the
Savannah River Site and expressed a number of concerns relating to the
biota. Since the APT has been designated as the backup, none of these
impacts to biota are expected. However, if a decision is made to pursue
the APT at a later date, these concerns would be addressed. The third
and four letters, which were from the Environmental Protection Agency's
(EPA) Region 4 Office in Atlanta, Georgia, concerned the APT and TEF
EISs. The letters stated that DOE adequately responded to all EPA
comments, but that EPA continues to have environmental concerns related
to the wetlands, surface water, and groundwater impacts for the APT
project, and the response to, and potential environmental impacts,
associated with accidental releases for the TEF project. If a decision
is made to pursue the APT, these concerns would be addressed. The
concerns regarding the TEF project will be addressed in further detail
during the design and permitting process. No other comments or letters
were received.

II. Supplemental Programmatic Decision for Tritium Supply and
Recycling

A. Tritium Supply and Recycling Alternatives

The dual-track strategy established in the December 1995
Programmatic Record of Decision defined the alternatives that would
remain under consideration: (1) the purchase of an existing commercial
reactor (operating or partially complete) or irradiation services with
an option to purchase the reactor for conversion to a defense facility;
and (2) design and construction of an accelerator system for tritium
production. New construction of an

[[Page 26371]]

accelerator and/or a new tritium extraction facility would be located
at DOE's Savannah River Site near Aiken, SC. No new tritium recycling
capabilities or facilities are required or contemplated. This decision
was based on the Final Tritium Supply and Recycling Programmatic
Environmental Impact Statement (PEIS) (DOE/EIS--0161, October 1995).
This supplemental programmatic ROD makes a choice between the two
programmatic alternatives. It compares the alternatives with regard to
their ability to meet military requirements in terms of technical
maturity, capacity, and schedule risk; regulatory and licensing issues;
cost; nonproliferation policy issues; flexibility to meet changing
requirements, and environmental impacts.
The commercial reactor alternative has narrowed somewhat since
1995. DOE sought proposals from electrical utilities that operate
commercial light water reactors (CLWR). No proposals were submitted to
sell a reactor (operating or partially complete) to DOE. The Tennessee
Valley Authority (TVA) offered to provide irradiation services using an
incomplete reactor, for which DOE would provide funds to finish, plus
use of its currently operating reactors as needed. TVA also offered the
use of its currently operating reactors alone.
1. Description of Tritium Production Using Commercial Reactors
This section describes the process of producing tritium in a CLWR.
Current tritium requirements dictate that two CLWRs would be utilized
at any given time. DOE-designed Tritium Producing Burnable Absorber
Rods (TPBARs) would be placed in the reactors. DOE would have TPBARs
manufactured commercially under contract. A maximum of approximately
3400 TPBARs would be inserted in any one reactor for one fuel cycle.
TPBARs perform the same functions as burnable absorber rods, which are
used or have been used in commercial reactors to absorb excess neutrons
to control local power levels and fuel burnup rates. Commercial
burnable absorber rods absorb excess neutrons using the isotope Boron-
10 in ceramic form. TPBARs would also use a ceramic but substitute the
isotope Lithium-6 for Boron-10. Lithium-6 changes to tritium when
neutrons are absorbed. TPBARs would be placed in the reactors during
normal refueling outages. The TPBARs would remain in the reactors
throughout their normal operating cycle, usually a 15-18 month period.
The irradiated TPBARs would be replaced in the reactors with new ones
during refueling operations. Reactors potentially engaged in tritium
production must have their operating licenses amended by the Nuclear
Regulatory Commission (NRC). To meet current requirements, DOE plans
for the first irradiation cycle to begin in early FY 2004.
After irradiation, TPBARs would be transported in approved shipping
casks to a new TEF which would be constructed at DOE's Savannah River
Site and ready for operation no later than February 2006. The tritium
in each TPBAR is not gaseous, but is held in a solid matrix by several
internal structures. These structures are so effective in retaining the
tritium that a high-temperature furnace must be used to remove the
tritium as a gas. The TEF would use remotely operated handling
equipment and the furnaces that would heat the irradiated TPBARs to
around 1,000 degrees Celsius. The gases removed from the TPBARs would
be partially purified and pumped to the existing Tritium Recycle
Facility at the Savannah River Site for further processing and delivery
to the nuclear weapons stockpile. Following extraction, TPBARs,
classified as low-level radioactive waste, would then be sent to a low-
level radioactive waste disposal facility at the Savannah River Site.
2. Description of Accelerator Production of Tritium
The production of tritium in the proposed Accelerator Production of
Tritium facility may be viewed as a four-step process. First, protons
are accelerated to high energies. Second the protons strike tungsten to
produce neutrons through a nuclear process called spallation. Tritium
is produced in the third step, when the neutrons are captured by a
helium-3 feedstock (He-3) causing a nuclear reaction which produces
tritium and other isotopes of hydrogen. The final step is to separate
the tritium from the feedstock and purify it for use in the stockpile.
The APT would use radiofrequency waves to accelerate protons
(positively charged atomic particles). Electrical power would be
converted to radiofrequency waves outside the accelerator beam, and
waveguides (hollow metal conduits) would transmit the waves to cells
along the beam path. The accelerator design would enable the proton
beam to intersect with the radiofrequency waves in the proper
orientation to cause the protons to accelerate; in other words, the
radiofrequecy waves would push the protons down the beam tube faster
and faster.
Once the protons reached the desired energy, they would be directed
toward a target/blanket assembly of tungsten surrounded by lead. The
high energy of the protons striking the tungsten target would cause the
nuclei of the tungsten atoms to break into fragments, ejecting neutrons
and secondary particles in all directions (spallation). These neutrons
and some protons would be scattered to surrounding lead blanket modules
where more neutrons would be produced through additional nuclear
reactions. The neutrons freed during spallation would strike and be
absorbed by the feedstock material (i.e., He-30) in the target/blanket.
This absorption of neutrons would result in the production of tritium
and byproduct atoms. The tritium would then be separated from the
feedstock and purified. The purified tritium would be transported to
the Tritium Loading Facility at the Savannah River Site where it would
be used to refill tritium reservoirs in nuclear weapons.

B. Comparison of Non-Environmental Impacts of Tritium Supply
Alternatives

DOE is responsible to the President and its primary customer, the
Department of Defense, for establishing an assured source of tritium on
a schedule that meets the requirements discussed in the background
section above. Several factors, not directly related to environmental
impacts, are important in assessing the probability that each tritium
supply alternative will meet that responsibility. The factors discussed
below are: ability to meet military requirements; regulatory and
licensing issues; cost; nonproliferation issues; and flexibility to
meet changing requirements.
1. Ability To Meet Military Requirements
To meet military requirements, a tritium source must have low
technical risk, must have the capacity to produce tritium at required
rates, and must meet schedule deadlines. The tritium supply options are
assessed in these terms below:

Technology Maturity/Risk

Since its inception, the APT Project has sought to develop and
demonstrate critical components of a tritium production system and to
reach a level of maturity in the design of a full-scale production
system so that its technical risks, costs, and schedule can be fully
understood. At this point a majority of the accelerator system's
preliminary design has been completed and a low-energy demonstration
accelerator at the Los Alamos National Laboratory in New

[[Page 26372]]

Mexico has undergone construction and successful operational testing.
Several external reviews have revealed no technical ``showstoppers.''
However, accelerators have never made tritium on a continuous
production scale, and the APT would be a first-of-a-kind facility.
Tritium production in reactors has been demonstrated to be safe and
technically straightforward. Although there are variations in the
technical details, in the past the only method used to produce tritium
has been with reactors and tritium-producing ``targets'' containing
lithium. DOE began considering commercial reactor target designs for
tritium production in the 1960s. The TPBAR to be used in commercial
reactors was designed and extensive development and testing done during
DOE's previous New Production Reactor Program (1988-1992). Commercial
nuclear power is supported by a well developed, mature industrial
infrastructure. During that program, rods of essentially the same
design as those to be used in commercial reactors were irradiated in
DOE's Advanced Test Reactor at the Idaho National Environmental and
Engineering Laboratory. Post-irradiation non-destructive and
destructive examinations have shown that the rods performed even better
than predicted. Various laboratory tests have consistently shown TPBAR
component performance to be as good or better than expectations.
Following two extensive technical reviews by the Nuclear Regulatory
Commission (NRC) and the approval of an amendment for its operating
license issued in September 1997, the Tennessee Valley Authority's
(TVA) Watts Bar reactor irradiated 32 TPBARs over a normal operating
cycle for a confirmatory demonstration. Frequent monitoring of the
reactor coolant and neutron flux indicated no problems with the rods.
Following irradiation, the rods were removed from the reactor's spent
fuel on March 19, 1999, and visually inspected. The inspection of the
32 TPBARs showed no indications of any kind of problem. In February
1999, DOE submitted the Tritium Production Core Topical Report to the
NRC. NRC's review of the report has raised no significant concerns and
a Safety Evaluation Report to this effect is now being finalized by the
NRC.
Conclusion: While much progress has been made in addressing the
technical issues that existed regarding the APT at the time of the 1995
Record of Decision, tritium production technology for light water
reactors is more technically mature, and carries with it less technical
risk than the APT.

Capacity

The commercial reactor alternative and the APT alternative would
both have a maximum production capacity of about 3 kilograms of tritium
per year. Commercial reactors routinely operate at full power for
extended periods of time. The national average capacity factor for
commercial reactors is in excess of 75 percent, including all refueling
shutdown periods. The Watts Bar reactor, while irradiating 32 of DOE's
TPBARs, recently shut down for refueling, having been in continuous
high-power operation for 353 consecutive days. The availability of
multiple candidate reactors for irradiating TPBARs also provides high
confidence that tritium production capacity requirements can be met.
Although much progress has been made, the APT project has not yet
demonstrated its tritium production capacity.
Conclusion: Although either alternative should be able to meet
capacity requirements, the availability of multiple commercial reactors
and their demonstrated capacity factors provides a greater degree of
confidence that production goals can be met consistently.

Schedule

The commercial reactor alternative could begin producing its first
batch of tritium in October 2003 when one of the candidate reactors is
scheduled to complete a refueling outage. Because many technical and
regulatory issues have been addressed already, there is a high degree
of confidence that this initial irradiation schedule can be met. The
first batch could be delivered to the stockpile as tritium gas as soon
as the TEF is operational. Selection of the incomplete reactor approach
would not impact the schedule because an existing reactor would be used
to irradiate the initial batch of TPBARs. Under both reactor
alternatives, current START I requirements would be met without the use
of the 5-year tritium reserve. The APT alternative would be operational
around 2008 and would begin continuous tritium production at that time.
This would require that 3 years of the 5-year reserve be utilized for
stockpile support. The APT would need to operate at its maximum
capacity for a number of years to replace the depleted reserve.
Conclusion: There is a high likelihood that, with adequate funding,
the reactor alternatives can meet the schedule and the tritium reserve
would not be impacted. The APT would require that at least 3 years of
the reserve be consumed and that the machine operate at maximum
capacity until the reserve has been restored. Any schedule delay beyond
2008 would potentially utilize the balance of the reserve and thus
potentially impact the stockpile. If START II is ratified and
implemented, any schedule risk would be eliminated. However, for
current stockpile requirements, the commercial reactor alternative has
the best chance for meeting schedule requirements.
2. Regulatory and Licensing Issues
Both the reactor and accelerator alternatives would be overseen by
bodies external to DOE. The potential for oversight/regulatory issues
to impact the tritium alternatives is discussed below.
The NRC would have to amend the operating licenses of existing
commercial reactors to permit production-scale irradiation of tritium-
producing rods. Requests for license amendments would be submitted in
the middle of calendar year 2000. It is expected that the NRC would be
in a position to act upon the amendment requests well in advance of the
planned October 2003 start of irradiation. Some experience has already
been gained in this area because the Watts Bar reactor's operating
license was amended in September 1997 to permit the confirmatory test
irradiation of 32 TPBARs. That licensing process was completed in a few
months. The NRC has completed two reviews of technical reports on the
TPBAR submitted by DOE and a third review of a reactor-specific request
to amend the Watts Bar reactor's operating license for the confirmatory
irradiation demonstration. No significant safety issues were
identified.
If a partially complete reactor were finished and brought on line,
the facility would have to be licensed as a new nuclear power plant.
The licensing process is likely to take up to 5 years. As discussed
above, this would not impact national security because initial tritium
production would begin with an existing reactor. However, delays in
getting the incomplete plant into operation could delay and possibly
reduce DOE's receipt of revenues from the plant's power sales. Thus,
the only potential regulatory impact would be financial in nature.
The APT design, construction, and operation would be overseen by
the Defense Nuclear Facilities Safety Board (DNFSB). To date, the DNFSB
has not identified any issues that would affect the availability of
this facility. The APT would not require a license for its construction
or operation.

[[Page 26373]]

Conclusion: The APT option appears to have no regulatory and
licensing issues. The existing-reactor sub-option is not likely to be
impacted by regulatory and licensing issues. The incomplete reactor
sub-option has potential for these issues to impact its schedule, but
is not likely to affect tritium production because initial irradiation
would be with an existing reactor.
3. Cost
Cost is determined in terms of investment cost and life-cycle cost.
Investment cost is defined as the total of all remaining up-front costs
necessary to design, develop, construct, startup, or otherwise
establish tritium production capacity. Investment costs are generally
the same as project costs. Life-cycle cost is defined as the total
amount of money spent to produce 100 kilograms of tritium over the life
of the alternative to meet current START I requirements. Life-cycle
cost includes investment cost, all operating costs, and decontamination
and decommissioning (D&D) costs. All cost discussions refer to constant
FY 1999 dollars.
The investment cost remaining (FY 1999-2008) to develop, design,
construct, and startup the APT facility, sized to meet START I tritium
requirements, would be $3.4 billion. The investment cost remaining to
establish capabilities to produce tritium through irradiation services
with existing commercial reactors and to design, construct, and startup
the TEF would be $580 million. This investment cost would increase by
$1.2-1.8 billion if finishing an incomplete reactor is included.
The annual operating cost of the APT would be $135 million when
meeting START I tritium requirements. The annual operating cost to
produce START I quantities of tritium using existing reactors would be
$20-60 million. At the high end of this range DOE would pay for the
incremental increase in the enrichment of the host reactors' fuel as
needed to accommodate TPBARs for tritium production. At the low end of
the range DOE would provide blended-down highly enriched uranium from
its national security stocks, and the host utility would reimburse DOE
for that portion not directly attributable to tritium production. If
DOE provides funds to finish an incomplete reactor, under some
scenarios, the Government would share in the power sales revenue of
that reactor. These revenues would depend on the amount of investment
money provided and whether the funds were provided over a short period
or an extended period. Large ``block'' investment payments would result
in the highest revenue share. Reduced, extended payments would provide
no revenue share. Depending on the investment, the annual operating
cost to DOE would range from around $30 million of net income to around
$25 million of net outlay.
D&D costs for the APT would be $260 million. For the reactor
alternative, DOE would be liable only for D&D of the TEF at $8 million.
DOE would have no liability for reactor D&D costs.
The APT and TEF would be designed for a 40-year life. Although the
NRC licenses of currently operating reactors would expire before then,
extension of the reactors' operating licenses is possible, either to
meet power demand or tritium requirements or both. For purposes of this
cost analysis, it is assumed that suitable reactors will be available
throughout the 40-year period.. Thus, all alternatives were compared on
the same life-span basis. Life-cycle cost for the APT is estimated to
be $9.2 billion. Life-cycle cost for the use of commercial reactors is
estimated to be $1.2 billion to $2.9 billion, depending on the
investment-revenue combination discussed above.
The present discount value of the APT alternative, using a 3.6
percent discount rate, would be $5.2 billion. The present discount
value of the commercial reactor alternative would range from $880
million to $2.0 billion, depending on the investment and fuel
enrichment strategies, as discussed above.
Conclusion: Under current requirements, the commercial reactor
alternative would cost significantly less than the APT alternative in
terms of investment costs, operating costs, D&D costs, life-cycle
costs, and present discount value.

Cost To Meet Reduced START II Requirements

If START II comes into force, the tritium need date could be around
2011 and the maximum tritium production rate may be reduced to about
1.5 kilograms per year. If so, a smaller accelerator could be
constructed, reducing its investment cost to $2.8 billion. The existing
commercial reactor alternative's investment cost remains about the same
as the START I case. The accelerator alternative's life-cycle cost
under this reduced-requirement scenario would be $7.5 billion. Life-
cycle cost for the commercial reactor alternative, using existing
reactors would be $2.2 billion or less, depending on the fuel
enrichment strategy. Adding completion of an unfinished reactor could
drive the life-cycle costs up or down, depending on the investment
strategy.
Conclusion: Under START II requirements, the commercial reactor
alternative would cost significantly less than the APT alternative in
terms of investment cost, operating costs, D&D costs, and life-cycle
costs.
4. Nonproliferation Issues
Concerns have been expressed by members of Congress and other
individuals and groups regarding the use of a civilian reactor to
assist a defense mission. As a result of these concerns, the Congress
requested the Department to facilitate a high-level interagency review
of the nonproliferation implications of the various tritium production
technologies. Participants in the review included the National Security
Council, the Department of Defense, the Department of State, the Arms
Control and Disarmament Agency, the White House Office of Science and
Technology Policy, the Office of the Vice President, and the NRC. The
report, Interagency Review of the Nonproliferation Implications of
Alternative Tritium Production Technologies Under Consideration by the
Department of Energy, was provided to the Congress in July 1998. A
summary of conclusions of the report follows:
The interagency report noted that tritium is not a fissionable
material, and thus there is no legal prohibition on the production of
tritium in a commercial reactor to support the stockpile. The report
concluded that ``the nonproliferation policy issues associated with the
use of a commercial light water reactor are manageable, and that the
Department should continue to pursue the reactor option as a viable
source for future tritium production.'' This conclusion was based on a
number of factors, including the following:
Use of commercial reactors for tritium production is not
prohibited by statute or international treaty;
There have been several exceptions over the past several
decades to the practice of distinguishing between the civilian and
military uses of nuclear power.
Commercial reactors engaged in tritium production would
remain eligible for the application of International Atomic Energy
Agency safeguards.
The commercial reactor option would be operated in
compliance with international agreements imposing restrictions on use
of transferred materials for peaceful purposes only, e.g., no reactor
fuel or component

[[Page 26374]]

transferred under these agreements would be used by any reactor making
tritium; and
Further mitigation is offered if the existing reactors are
operated by TVA. TVA's statutory charter assigns it a national security
mission. TVA's reactors are already government facilities. TVA has made
contributions to national security in the past including production of
munitions and providing power for the enrichment of uranium for
civilian and military purposes. It would, therefore, be entirely
appropriate for TVA to be assigned the tritium production mission.
The interagency review concluded that the accelerator option would
raise no significant nonproliferation policy issues, assuming that
export control measures are maintained. Subsequent to the issuance of
the report, concerns have been expressed, applicable to both the APT
and to the completion of an unfinished reactor, that the commitment to
a major new weapons facility would be inconsistent, either in fact or
in appearance, with our commitment to further stockpile reductions and
thus to our obligations under the Nuclear Nonproliferation Treaty.
These concerns were considered in the tritium technology decision
process.
Conclusion: Although concerns have been expressed about each of the
tritium production alternatives, nonproliferation policy issues would
not preclude the selection of any alternative.
5. Flexibility To Meet Changing Requirements
Since tritium production stopped in 1988, the U.S. tritium
requirements have been reduced by almost 75 percent, primarily because
of the stockpile reductions resulting from bilateral arms control
agreements. The current tritium production requirement is based on
supporting a stockpile sized for START I. If START II is ratified by
the Russian Duma (legislature), the U.S. may decide to reduce its
tritium production requirements, thus moving the need date to 2011 and
reducing tritium production requirements. Stockpile reductions beyond
START II are possible and would hopefully occur, potentially resulting
in further extension of the tritium need date and reductions in tritium
production requirements.
The APT has significant flexibility to change its rate of tritium
production and therefore its operating costs. It is less flexible in
its avoidance of capital investment costs. The APT project plan calls
for construction of a ``modular'' accelerator sized to produce about
1.5 kilograms per year, the capacity sufficient for a START II
stockpile. According to the plan, if current tritium requirements are
not reduced by early FY 2000, accelerator construction would proceed
with a full-size machine having a capacity of 3 kilograms per year with
a $500 million increase in investment cost. If tritium requirements are
reduced after early FY 2000 much of the investment cost of the APT
would be ``sunk.''
The use of the existing, operating reactors is the most flexible
option with respect to changing stockpile levels. If the tritium need
date is extended during FY 1999-2000, most investment for this
alternative could be suspended indefinitely and then restarted later. A
substantial portion of DOE's operating costs would be based on tritium
demand on a pay-as-you-go basis. Except for minimal standby costs, DOE
would pay for irradiation services, TPBAR manufacturing, and
transportation operations only during those years when tritium is
actually required. The amount spent for irradiation services would, to
a great degree, depend on the amount of tritium produced. If the
tritium need date is extended before the TEF handles its first
increment of radioactive material, that facility could remain in
standby indefinitely for less than $1 million per year.
If completion of an unfinished reactor is considered, the reactor
alternative's flexibility characteristics become much like those of the
APT. While there is great flexibility in amounts of tritium that can be
produced, the large up-front investment cost would have no relation to
tritium requirements. Once DOE committed itself to completion of the
reactor, there would be no opportunity to reduce investment costs if
stockpile tritium requirements were reduced. Revenues would be returned
to DOE whether tritium is needed or not, but the cost per kilogram
would obviously be higher if tritium requirements were substantially
reduced as a result of further arms reduction agreements. The annual
net operating cost (positive or negative) of this alternative would
vary somewhat with tritium demand because of reductions in the cost for
TPBAR manufacturing and transportation, thus reducing the total-life
cycle cost.
Conclusion: The use of existing reactors potentially results in the
greatest degree of flexibility to meet changing requirements,
especially in view of the potential for future reductions in the
nuclear weapons stockpile.

C. Comparison of Environmental Impacts of Tritium Supply Alternatives

Since the December 1995 Tritium Supply and Recycling PEIS ROD, a
substantial amount of work has been accomplished on both the CLWR
tritium production alternative and the APT alternative, including the
issuance of project-specific Environmental Impact Statements. In the
course of preparing this supplement to the December 1995 ROD for the
tritium supply technology decision, in order to select between the two
technologies, DOE reviewed the Tritium Supply and Recycling Final PEIS
to ensure that the information contained there is still valid. The
conclusion of that review is that the Tritium Supply and Recycling PEIS
remains a valid basis for the programmatic portion of this consolidated
ROD.
In the December 5, 1995 ROD for the Tritium Supply and Recycling
PEIS, environmental impacts of the various tritium supply technologies
were compared and a general conclusion was reached that ``[for all of
the reasonable tritium supply technology alternatives] the
environmental impacts are generally small and, except for the
commercial reactor options to purchase an existing reactor or
irradiation services, the impacts are within the same range. The
Department considers the commercial reactor options of purchasing an
existing reactor or irradiation services to be the environmentally
preferred alternative.'' [60 FR 63889] As discussed below, these
conclusions remain true.
Described below are the relative differences in environmental
impacts between tritium production in operating CLWRs (TVA's Watts Bar
Unit 1 and Sequoyah Units 1 and 2 are used in the analysis) and an
incomplete CLWR (TVA's Bellefonte Unit 1 is used in the analysis), and
construction and operation of the APT at the Savannah River Site. For
an incomplete CLWR, the environmental analysis attributes all of the
impacts from completing construction and operating the plant to the
tritium production mission. Additionally, because any tritium produced
by a CLWR would need to be extracted from TPBARs prior to delivery to
the nuclear weapons stockpile, the impacts associated with operation of
a TEF are included in the discussion below, as appropriate. DOE has
decided previously that a TEF capability would be constructed
regardless of whether the CLWR option is selected as the primary or the
backup tritium supply [60 FR 63890]. In the latter case the TEF would
be needed as part of a viable backup system and could have been

[[Page 26375]]

incorporated as part of the APT facility. Therefore, construction
impacts of TEF apply if either the CLWR or APT option is chosen, but
TEF operating impacts apply only to the CLWR. Because of the
availability of data in the tiered, final EISs for use of commercial
reactors for tritium production, the TEF, and the APT, the discussion
below is based upon the best available information and analyses that
have been developed to date.
1. Construction Impacts
For tritium production in a CLWR, construction impacts would range
from none (for operating CLWRs) to minor (for a CLWR which is currently
approximately 90 percent complete, and would only require internal
modifications). The predominant construction impact associated with an
incomplete CLWR would be on socioeconomics, as approximately 4,500
direct jobs and 4,500 indirect jobs could be created during the peak
year of construction. The creation of approximately 9,000 total jobs
would have a significant positive impact on the economic area
surrounding the incomplete reactor. For the APT at the Savannah River
Site, construction impacts would consist of: land disturbance of
approximately 250 acres; water use of less than 1 percent of current
use; and socioeconomic impacts associated with a peak-year construction
workforce of approximately 1,400 direct jobs and approximately 900
indirect jobs. The creation of approximately 2,300 total jobs would
have a significant positive impact on the economic area surrounding the
Savannah River Site. Construction impacts associated with a TEF at SRS
would be minimal. Land disturbance would occur in a densely developed
industrial area. Water use would be less than 1 percent of current site
use. Socioeconomic impacts associated with a peak-year workforce would
be about 740 direct jobs which would have a positive stabilizing
influence on SRS employment but an insignificant impact on regional
employment.
Conclusion: With respect to construction impacts associated with
tritium production, use of an existing CLWR would have the least impact
on the natural environment. Completion of an unfinished reactor would
have positive socioeconomic impacts, as would the APT at SRS. Using an
existing CLWR would have no socioeconomic impacts. For all
alternatives, the environmental impacts associated with construction
are considered small.
2. Operating Impacts
For an operating CLWR, there would either be no impacts, or
negligible impacts, to resources such as: land, infrastructure, noise,
visual, air quality, water resources (use and quality), geology and
soils, archeological and historic, and socioeconomics. Tritium
production and extraction could cause additional impacts in the
following resources: spent fuel generation; human health (normal
operations and accidents); low-level radioactive waste (LLW)
generation; and transportation.
For the alternative that would complete, start up, and operate an
incomplete reactor, the operating impacts include those impacts
associated with a new commercial nuclear power plant. The following
resources would be affected: infrastructure (including visual
resources); water resources; spent fuel generation; human health
(normal operations and accidents); LLW generation; transportation; and
socioeconomics.
Operation of a TEF at the Savannah River Site would affect the
following resources: infrastructure; water resources; human health
(normal operations and accidents); LLW generation; and socioeconomics.
For the APT, tritium production could cause impacts in the
following resources: infrastructure; surface water; human health
(normal operations and accidents); LLW generation; and socioeconomics.
For the resources potentially affected during operation of any tritium
supply technology, the most significant discriminators between
alternatives are: infrastructure, spent fuel, human health (including
impacts from accidents), low-level waste generation, and
socioeconomics. These resources are discussed below for the tritium
production alternatives, as appropriate.

Infrastructure

The production of tritium in an operating CLWR would have no impact
on the local infrastructure. The impacts of operating a newly completed
reactor would produce more than 1,200 megawatts of usable electric
power. In an area such as the Tennessee Valley, this beneficial impact
would tend to reduce the need for operation of coal-fired or gas-fired
power plants, or could offset the need for additional power plants in
the future, potentially reducing future air emissions. Although visual
resources surrounding the incomplete reactor site would be negatively
impacted by a cooling tower plume, this would not be significant enough
to change the plant's existing visual resource classification. For the
operation of the TEF, estimates for base load electricity use are
approximately 2.4 megawatts of electric power, which would be provided
through the existing infrastructure at SRS.
For the APT, estimates for base load electricity use are up to 350
megawatts of electric power. Environmental impacts associated with
production of electricity by a coal-fired or gas-fired power plant
would consist mainly of increased air emissions; however, no air
quality standards are expected to be exceeded. The visual impacts of
the APT are not deemed significant because the facility would not be
visible from the Savannah River Site boundaries to ground-level
observers.
Conclusion: Operation of a newly completed reactor would produce a
positive impact on the local infrastructure by producing more than
1,200 megawatts of electric power. An operating CLWR used for tritium
production would have no additional impact on the local infrastructure.
The TEF would have a negligible impact on the local infrastructure at
the Savannah River Site. The APT would have a minor negative
environmental impact on the local infrastructure by requiring
approximately 350 megawatts of electric power.

Spent Fuel

The reactors considered here each use 193 fuel assemblies when
operating. At each refueling a percentage of these assemblies are
removed from the reactor and placed in the reactor's spent fuel storage
pool. The number of assemblies of spent fuel generated by an existing
reactor could increase as a result of tritium production. Increases
could range from approximately 60 spent fuel assemblies per cycle if a
CLWR is loaded with a maximum of 3,400 TPBARs, to no increase in spent
fuel if a CLWR is loaded with less than approximately 2,000 TPBARs. The
environmental impacts associated with long-term, on-site, dry-cask
storage of spent fuel are not significant. For a newly completed CLWR,
approximately 72 spent fuel assemblies would be generated during
reactor operations without tritium production. For nominal tritium
production, the amount of spent fuel generated would not increase as
long as less than approximately 2,000 TPBARs are loaded into the
reactor. If maximum tritium production is needed, up to 3,400 TPBARs
would be used and approximately 69 additional spent fuel assemblies
would be generated per cycle. In this regard, it is DOE's

[[Page 26376]]

intention to minimize, if not eliminate, the generation of additional
spent fuel by limiting the number of TPBARs inserted in a single
reactor. Neither the TEF, nor the APT, would generate spent fuel.
Conclusion: Operation of a newly completed reactor would generate
the most additional spent fuel. Use of currently operating reactors
could lead to a limited incremental increase in spent fuel. The APT
would generate no spent fuel.

Human Health (Normal Operations)

By adding tritium production to the currently operating reactors,
there would be additional radiation doses to workers and the public
from tritium production. The incremental increase in annual average
worker dose is estimated at approximately 1.1 millirem, while the total
population dose within 50 miles is estimated to increase by
approximately 2.0 person-rem per year during normal operations. In
terms of potential impacts, these values are not significant. For
example, a 2.0 person-rem dose translates into a latent cancer fatality
risk of 1 in 1,000 years. For the average worker, a 1.1 millirem annual
dose translates to a risk to that worker of a latent cancer fatality
every 2.3 million years.
By finishing the incomplete reactor and operating it to produce
electricity and tritium, there would be radiation doses to workers and
the public that do not currently occur. The average annual worker dose
is estimated at a maximum of approximately 105 millirem, of which 104
millirem would result from operation of the reactor to produce
electricity, and 1.1 millirem would be from tritium operations. The
annual total population dose within 50 miles is estimated to be a
maximum of approximately 2.3 person-rem. In terms of potential impacts,
these values are not significant. For example, a 2.3 person-rem dose
translates into a latent cancer fatality risk of 1 in 870 years. A 105
millirem annual dose translates to a risk to an average worker of a
latent cancer fatality every 23,000 years.
Operation of the TEF at the Savannah River Site would result in
small radiological impacts to workers and the public from tritium
production. The average annual worker dose is estimated at
approximately 40 millirem, while the total population dose within 50
miles is estimated to increase by approximately 0.77 person-rem per
year. In terms of potential impacts, these values are not significant.
For example, a 0.77 person-rem dose translates into a latent cancer
fatality risk of 1 in 2600 years. For the average exposed worker, a 40
millirem annual dose translates to a risk to that worker of a latent
cancer fatality every 62,500 years.
Operation of the APT would result in small radiological impacts to
workers and the public from tritium production. The average annual
worker dose is estimated at a maximum of approximately 144 millirem,
while the total population dose within 50 miles is estimated to be
approximately 2.0 person-rem. In terms of potential impacts, these
values are not significant. For example, a 2.0 person-rem dose
translates into a latent cancer fatality of 1 in approximately 1,000
years. A 144 millirem annual dose translates to a risk to an average
worker of a latent cancer fatality approximately every 17,600 years.
Conclusion: Radiological impacts for normal operations are
considered small for all alternatives. The APT and commercial reactor
options would have comparable impacts to the population. Use of an
operating CLWR would have the smallest impact to workers.

Human Health (Accidents)

Based upon tests and analyses that had been performed previously as
part of the DOE's New Production Reactor program, the Tritium Supply
and Recycling PEIS concluded that ``it appears that no new significant
safety hazard is introduced as a result of a decision to produce
tritium in an existing CLWR.'' [PEIS, page 4-524] Nonetheless, the PEIS
also acknowledged that a complete reactor-specific evaluation remained
to be completed. The CLWR EIS provides a detailed evaluation of impacts
from accidents on a site-specific basis for the CLWR reactor
alternatives. Based upon the CLWR EIS evaluation, the conclusion in the
PEIS is further supported. The CLWR EIS documents that the potential
impacts from tritium production on accident impacts is small. For
design-basis accidents at operating reactors, the risk of a latent
cancer fatality to an average individual from tritium production in the
50-mile population surrounding a CLWR would be approximately 1 in 490
million years. At the incomplete reactor site, this risk would be
approximately 1 in 1.3 billion years. For beyond design-basis
accidents, tritium production would result in very small changes in the
consequences of an accident. This is due to the fact that the potential
consequences of such an accident would be dominated by radionuclides
other than tritium. At the operating reactors, the additional risks to
the 50-mile population from adding tritium production would be less
than one additional cancer per every 100,000 years from a beyond
design-basis accident. At the incomplete reactor site, the total risk
of the new reactor and the added tritium mission to the 50-mile
population would be approximately 11 latent cancer fatalities per
100,000 years from a beyond design-basis accident.
The potential impacts to the public from accidents associated with
operation of the TEF at the Savannah River Site are extremely small.
For the design-basis accident, the risks to the 50-mile population
would be approximately 7 latent cancer fatalities per 100,000 years.
The potential impacts to the public from either a design-basis or
beyond design-basis accident from the APT are small. For a design-basis
accident, the risk of a latent cancer fatality to an average individual
in the 50-mile population would be approximately 1 in 470 million
years. For beyond design-basis accidents, the risks to the 50-mile
population would be approximately 3 latent cancer fatalities per
100,000 years.
Conclusion: The risks associated with accidents are small for all
the tritium production alternatives. Differences between the CLWR and
APT are not deemed to be significant.

Low-Level Radioactive Wastes

LLW generation at the operating reactors could increase by 0.43
cubic meters annually as a result of tritium production. The impact of
disposing of the additional LLW at the Barnwell commercial disposal
facility at Barnwell, South Carolina would represent much less than 1
percent of the total LLW that is currently disposed of at that
facility. The newly completed reactor would generate approximately 40
cubic meters of LLW annually, which would also be less than 1 percent
of the total LLW that is disposed of annually at the Barnwell LLW
commercial disposal facility. Operation of the TEF would generate
approximately 232 cubic meters of LLW annually. These wastes would be
manageable using existing waste management treatment, storage, and
disposal facilities at the Savannah River Site.
The APT would generate approximately 1,400 cubic meters of LLW
annually. These wastes would be manageable using existing waste
management treatment, storage, and disposal facilities at the Savannah
River Site. The environmental impacts of all waste types for all
alternatives, including LLW, would be small and manageable with
existing facilities.
Conclusion: Although all of the waste generation impacts are
acceptable, the

[[Page 26377]]

use of currently operating reactors would generate the smallest amount
of low-level wastes from tritium production. For all alternatives, the
environmental impacts of all waste types, including low-level waste
would be small and manageable with existing facilities.

Socioeconomics

Little or no socioeconomic impact is expected by adding the tritium
production mission at an operating CLWR. Operation of a newly completed
CLWR would add approximately 800 direct and 800 indirect jobs. The
socioeconomic impacts of the 1,600 total jobs would have a positive
impact on the economic area surrounding the reactor site. Operation of
the TEF would add approximately 108 direct jobs. This would not have
any significant impact on the local socioeconomic area. Operation of
the APT would add approximately 500 direct jobs and 335 indirect jobs.
The socioeconomic impacts of the 885 total jobs would have a positive
impact on the economic area surrounding SRS.
Conclusion: Operation of a newly completed reactor and the APT
would have the greatest positive socioeconomic impacts, while use of
currently operating CLWRs to produce tritium would involve
insignificant socioeconomic impacts.

Transportation

There will be impacts associated with transporting irradiated
TPBARs from the reactor sites to the TEF at the Savannah River Site.
There would be approximately 13 shipments of TPBARs annually to SRS
which would result in an annual human health risk, over the entire
route of the shipments, of less than 1 latent cancer fatality every
100,000 years. The impact on any one individual would be less than
that. Because the Tritium Loading Facility and the APT would be located
at SRS, there are no impacts directly associated with transportation.
Conclusion: Although all the transportation impacts are negligible,
the APT has the least impact.
3. Overall Environmental Conclusion
As described above, and as documented in the environmental analyses
that have been developed, it is expected that the overall environmental
impacts associated with tritium production in either a CLWR or the APT
would be small. Consequently, the environmental impacts associated with
the two alternatives are not considered a major discriminating factor
in this tritium technology decision. The December 1995 Programmatic ROD
stated that the use of existing CLWRs for tritium production would be
the environmentally preferred alternative. Subsequent analyses,
discussed here, confirm this still to be true.

D. Programmatic Decision

Both technology alternatives are feasible. Consistent with the
Department's December 22, 1998, announcement, and based on the above
analysis, DOE selects the use of existing commercial light water
reactors as the primary technology to produce tritium for national
security purposes. In implementing this decision, DOE will construct a
new Tritium Extraction Facility on the Savannah River Site.
The use of commercial light water reactors is selected to be the
primary tritium supply technology because analysis leads to the
conclusion that this technology:
Would have the best chance of meeting all military
requirements due to:
Lowest technical risk.
Lowest schedule risk.
Highest confidence for meeting capacity requirements.
Would have the lowest investment and life-cycle costs.
Offers potential to be the most flexible in meeting
changing requirement.
Offers potential to have the least environmental impact.
The Accelerator Production of Tritium (APT) is designated as the
backup tritium production technology. The APT Project will complete
Engineering Development and Demonstration (ED&D) activities and final
design for a few key elements of the accelerator system. Completion of
these activities would permit expedient initiation of facility
construction if the accelerator is called upon.
In January 1997, the Fast Flux Test Facility (FFTF) was placed in a
safe standby condition as near-term ``insurance'' given the
uncertainties at that time with the dual-track technologies for tritium
production. Because it could not produce enough tritium to meet
production requirements, it could not serve as a potential primary
long-term tritium supply source. The Department's evaluation of FFTF
has focused on whether it can or should play any role as an interim
source of tritium until one of the other technologies is implemented.
The Department is fully confident that the tritium supply strategy
embodied in this decision can meet any current or future tritium
requirements. Consequently, the Department's FFTF will have no tritium
production role. A separate study is being conducted to determine if
that reactor should be restarted and operated for other purposes.

III. Site-specific Decision for the Production of Tritium Using
Commercial Light Water Reactors (CLWR)

A. CLWR EIS Alternatives

In conformance with the Department's December 22 announcement, the
preferred alternative identified in the CLWR Final EIS is to produce
tritium in the Watts Bar and Sequoyah reactors. As a result of the
programmatic decision in this ROD (see section II), DOE will produce
tritium in a CLWR, and the APT is designated as the back-up technology.
Consequently, the comparisons described in this section are focused
solely on the TVA reactor alternatives, and not the APT.
The CLWR EIS evaluates the following alternatives: (1) No Action
Alternative (which would result in the production of tritium in an
accelerator at the SRS); and (2) Tritium production at one or more of
the following Tennessee Valley Authority (TVA) CLWRs: Watts Bar Nuclear
Plant Unit 1 (Spring City, TN); Sequoyah Nuclear Plants Units 1 and 2
(Soddy Daisy, TN); and Bellefonte Nuclear Plants Units 1 and 2
(Hollywood, AL). The Watts Bar and Sequoyah reactors are existing,
operating CLWRs that produce electricity. Tritium production could be
performed in these reactors without any significant modifications to
these facilities and would not affect electricity production. The
Bellefonte units are unfinished nuclear reactors. Bellefonte Unit 1 is
approximately 90% complete, and Unit 2 is approximately 58% complete.
In order to produce tritium in a Bellefonte reactor, construction would
have to be completed and an operating license would have to be received
from the NRC.

B. Non-Environmental Comparison of CLWR Reactor Alternatives

1. Cost and Flexibility Factors
Investment cost is defined as the total of all remaining up-front
costs necessary to design, develop, construct, startup, or otherwise
establish tritium production capacity at each of the CLWRs. Investment
costs are generally the same as project costs. Life-cycle cost is
defined as the total amount of money spent to produce about 100
kilograms of tritium over the life of the alternative. Life-cycle cost
includes investment cost, all operating costs, and decontamination and
decommissioning (D&D) costs.

[[Page 26378]]

Cost to Meet Current Requirements (cost comparisons are expressed
in constant FY 1999 dollars).
The investment cost for the tritium-supply system that would use
the Watts Bar and Sequoyah reactors is estimated to be about $580
million, of which approximately $350 million are associated with
designing, constructing, and starting up the new TEF. Total investment
costs for the tritium-supply system that includes the Bellefonte
alternative are estimated to be $1.8 billion to $2.4 billion, depending
on the plan selected for payments to TVA to complete the reactor. The
Watts Bar/Sequoyah alternative could be accommodated within the DOE
Defense Programs budget but the Bellefonte alternative cannot.
The life-cycle cost for the Watts Bar and Sequoyah reactors ranges
from $1.4 billion to $2.9 billion, based on the letter agreement
between DOE and TVA signed on February 25, 1999. This includes $8
million for D&D of the TEF. The upper end of the life-cycle cost range
assumes that DOE would pay cash for the incremental increase in reactor
fuel enrichment needed for a reactor to accommodate TPBARs. The lower
end of the range assumes that highly enriched uranium, drawn from DOE's
defense stocks, would be blended down to provide all the fuel for the
host reactors. TVA would reimburse DOE at a market-based rate for that
portion of the fuel cost not directly attributable to tritium
production. Present discount value for the Watts Bar/Sequoyah option
would be in the range of $880 million to $1.6 billion.
Life-cycle cost of the Bellefonte alternative would be $1.2 billion
to $2.8 billion, depending on the plan for payments to TVA and DOE's
share of Bellefonte's power sales revenues. Because annual budget
limitations would likely prevent DOE from making large up-front
payments to TVA to complete Bellefonte, the lower-revenue-share/higher-
life-cycle-cost scenario is far more likely than the high revenue/low
life-cycle cost scenario. For the Bellefonte alternative, no fuel
transactions are assumed. Present discount value would be in the range
of $1.6-2.0 billion. D&D of the TEF, but no other facility, is
included.
Conclusion: The Watts Bar/Sequoyah alternative has the lowest
investment cost which can be accommodated within the DOE national
security programs budget. There is also strong potential for the Watts
Bar/Sequoyah option to have the lowest life-cycle cost because of the
likelihood that Bellefonte life-cycle costs would be near the high end
of the range. In addition, the Watts Bar/Sequoyah alternative has a
significantly lower financial risk because DOE would not pay until
tritium is produced. With the Bellefonte alternative there is a degree
of risk that, having paid for the plant, DOE would not receive any
return from net power revenues because of changes in the power market
or failure of the reactor to go into operation.

Cost To Meet Reduced START II Requirements

If START II comes into force, the tritium need date could be around
2011 and the maximum tritium production rate may be reduced to about
1.5 kilograms per year. If so, the existing commercial reactor
alternative's investment cost would remain about the same as the
current case. Life-cycle cost for the commercial reactor alternative,
using the existing TVA reactors would be in the range of $2.2-2.5
billion, based on the DOE-TVA letter of agreement of February 25, 1999.
The upper end of this range assumes DOE pays cash for incremental
increases in reactor fuel enrichment. The low end of this range assumes
DOE fuel stocks are blended to provide for the incremental increase in
fuel enrichment. The range could be lower still if TVA purchased all
its fuel from DOE. The Bellefonte alternative's relatively high
investment costs would not change under a START II scenario and the
life-cycle cost would be reduced by $100 million or less.
Conclusion: Under a START II scenario, investment and life-cycle
costs would be lowest for the Watts Bar/Sequoyah alternative.

Flexibility To Meet Changing Requirements

If START II is ratified, the U.S. may decide to reduce its tritium
production requirements, thus moving the need date to around FY 2011
and reducing tritium production requirements. Stockpile reductions
beyond START II are also possible and would result in further extension
of the tritium need date and reductions in tritium production
requirements.
The Bellefonte reactor alternative's flexibility characteristics
are limited. While there is great flexibility in amounts of tritium
that can be produced, the large up-front investment cost would have no
relation to tritium requirements. Once DOE committed itself to
completion of the reactor, there would be no opportunity to reduce
investment costs if stockpile tritium requirements were reduced. The
annual net operating costs of this alternative would vary slightly with
tritium demand only because of reductions in the cost for TPBAR
manufacturing and transportation, thus reducing the total life-cycle
cost.
The use of the existing Watts Bar and Sequoyah reactors is the most
flexible with respect to changing stockpile levels. If the tritium need
date is extended, most investment for this alternative could be
suspended indefinitely and then restarted later. A substantial portion
of DOE's operating costs would be based on tritium demand on a pay-as-
you-go basis. Except for minimal standby costs, DOE would pay for
irradiation services, TPBAR manufacturing, and transportation
operations only during those years when tritium is actually produced.
The amount spent for irradiation services would be dependent on the
amount of tritium produced. If the tritium need date is extended before
the TEF handles its first increment of radioactive material, that
facility could remain in standby indefinitely for less than $1 million
per year.
Conclusion: The use of the existing Watts Bar and Sequoyah reactors
results in the greatest degree of flexibility to meet changing
requirements, especially in view of the potential for future reductions
in the nuclear weapons stockpile.

Arms Control/Nonproliferation

The use of the currently operating Watts Bar and Sequoyah reactors
has unique advantages not available with any other alternative,
including the Bellefonte option, which serve to offset the
nonproliferation implications of using these reactors. It is the only
option that does not require a very large up-front capital expenditure.
It is the only option that allows the nation to pursue the goal of
further arms reductions without commitment to a major new weapons
facility. By selecting Watts Bar and Sequoyah, the nation is assured of
a long-term option to make tritium, which may not have to be exercised
for many years if arms reduction efforts are successful, as DOE hopes
they would be.
By not committing itself to the construction of a major new weapons
facility, the U.S. can underscore to other nations, especially would-be
proliferant nations, its continuing pursuit of smaller nuclear weapons
stockpiles. This would be consistent with recent U.S. actions,
including cessation of underground nuclear testing, the stoppage of
plutonium production, and closure or withdrawal of defense missions
from several sites in the nuclear weapons complex. Commitment to a
major new weapons facility could be seen as building up U.S. nuclear

[[Page 26379]]

weapons production capabilities at a time when the U.S. is seeking to
reassure other nations of its commitment to nuclear arms reductions.
These factors offset the fact that the use of the three reactors
for tritium production would depart from the general practice of
maintaining a distinction between U.S. defense and civilian nuclear
activities. Moreover, the Department has determined that the impact of
this issue on U.S. nonproliferation policy is manageable, given the
surrounding circumstances enumerated above.
Conclusion: The use of the currently operating Watts Bar and
Sequoyah reactors is most consistent with stated U.S. nuclear weapons
stockpile reduction and nonproliferation goals.
2. Technical Factors

Capacity and Schedule

The Bellefonte alternative and the Watts Bar/Sequoyah alternative
could both achieve a production capacity of about 3 kilograms of
tritium per year. No matter which alternative is selected, the first
batch of tritium could begin production in early FY 2004 when the Watts
Bar reactor is scheduled to complete a refueling outage. Because many
technical and regulatory issues have been addressed already, there is a
high degree of confidence that this initial irradiation schedule can be
met. The first batch of tritium gas could be delivered to the stockpile
as soon as the TEF is operational. Because the Watts Bar and Sequoyah
reactors would be used to irradiate the initial batches of TPBARs,
delays in completing the Bellefonte reactor would not be expected to
impact the tritium production schedule. Under current START I
requirements, neither reactor alternative would require the use of the
tritium reserve.
Conclusion: Each reactor alternative can achieve capacity
requirements. There is a high likelihood that, with adequate funding,
each of the reactor alternatives can meet the schedule and the tritium
reserve would not be affected.

Regulatory and Licensing Issues

The Bellefonte alternative would have to be licensed as a new
nuclear power plant. The plant's initial NRC operating license would
also permit tritium production. This process is likely to take up to 5
years. This would not affect national security because initial tritium
production would begin with the Watts Bar reactor. Delays in getting
Bellefonte in operation would, however, delay and possibly reduce DOE's
receipt of revenues from Bellefonte power sales, if any.
The NRC would have to amend the operating licenses of the Watts Bar
and Sequoyah reactors to permit production-scale irradiation of
tritium-producing rods. DOE expects that NRC would be in a position to
act upon the amendment requests well in advance of the planned October
2003 start of irradiation. Some experience has already been gained in
this area because the Watts Bar reactor's operating license was amended
to permit the confirmatory test irradiation of 32 TPBARs.
Conclusion: The Bellefonte alternative has potential for these
issues to impact its schedule, but is not likely to affect tritium
production. However, delays in getting Bellefonte on line would reduce
the Government's receipts from its share of Bellefonte revenues, if
any. The Watts Bar/Sequoyah option is not likely to be affected by
regulatory issues. Watts Bar and Sequoyah are preferred over Bellefonte
because the completion and initial licensing of a new nuclear facility
entails greater technical and financial risk than obtaining a license
amendment for existing facilities.

C. Comparison of Environmental Impacts of CLWR Alternatives

The relative differences in environmental impacts between tritium
production in operating CLWRs (Watts Bar and Sequoyah) and the
completion and operation of the incomplete Bellefonte Unit 1 reactor
are described in the Supplemental Programmatic Record of Decision,
Section II.C, above. As described in that section and as documented in
the CLWR EIS, DOE expects that the overall environmental impacts
associated with tritium production in a CLWR would be small.
Consequently, the environmental impacts associated with the CLWR
alternatives are not considered a major discriminating factor in this
decision. Based on all of the environmental factors considered, the use
of the Watts Bar and Sequoyah reactors is the environmentally preferred
alternative.

D. CLWR Decision

DOE selects the Watts Bar Unit 1 and the Sequoyah Unit 1 and 2
reactors as the specific CLWRs to produce tritium for national security
purposes. Compared to completing the Bellefonte reactor, the use of the
currently operating Watts Bar and Sequoyah reactors for tritium
production would have the:
Lowest investment cost and lowest life-cycle cost under
most-likely scenarios.
Lowest financial risk.
Greatest flexibility to meet changing requirements.
Most consistency with stated arms reduction goals.
Lowest overall incremental environmental impact.
By selecting the Watts Bar and Sequoyah reactors, highly enriched
uranium, drawn from DOE's defense stocks, would be blended down to
provide for the enrichment increase.

IV. Site-Specific Decision for Construction and Operation of a
Tritium Extraction Facility (TEF) at the Savannah River Site

A. TEF Alternatives

The proposed action addressed in the Final EIS for the Construction
and Operation of a TEF at the Savannah River Site (SRS) is to design,
construct, test, and operate TEF at SRS to provide tritium extraction
capability to support tritium production technology. The purpose of TEF
is to extract tritium-containing gases from TPBARs irradiated in a CLWR
or from targets of similar design, and deliver the tritium-containing
gases to Building 233-H, the existing Tritium Loading Facility, for
final purification. As described below, DOE evaluated two reasonable
alternatives and a no-action alternative in the TEF Final EIS.
1. Construct a New Facility in the H-Area (Preferred Alternative)
As identified in the TEF Final EIS, the preferred alternative is to
locate TEF in H-Area, immediately adjacent to and west of Building 233-
H within the boundaries of SRS. The reasons for co-locating TEF close
to Building 233-H are: (1) To share common support facilities,
services, and some personnel; (2) to facilitate the transfer of tritium
between the two facilities; and (3) to use certain gas-handling
processes located in H-Area. TEF would consist of a concrete industrial
facility constructed partly below grade. The facility would be divided
into two major areas: (1) A remote handling area (RHA) and (2) a
tritium processing building. The tritium processing building would be
entirely aboveground; the floor of the RHA would be below grade.
Construction of the proposed facility would require approximately 4 to
5 years. Major process and operation systems included within the
proposed TEF would be: (1) The Receiving, Handling, and Storage System
that would support all functions related to the receipt, handling,
preparation, and storage of incoming radioactive sources and outgoing
radioactive waste materials; (2) the Tritium Extraction System that
would

[[Page 26380]]

get tritium and other gases from irradiated TPBARs, remove contaminants
from the gas stream, and store the hydrogen isotope/helium mixture; (3)
the Tritium/Product Processing Systems that would separate and purify
process gases from the irradiated TPBAR materials; (4) the Tritium
Analysis and Accountability Systems that would support monitoring and
tritium accountability; (5) the Solid Waste Management System that
would receive solid waste generated by TEF for management and storage
prior to disposal in the SRS E-Area vaults; and (6) the Heating,
Ventilation, and Air Conditioning System that would provide and
distribute conditioned supply air to the underground RHA and the
aboveground tritium processing area and also discharge exhaust air to
the environment via a 100-foot stack.
2. Upgrading the Existing Allied General Nuclear Services (AGNS)
Facility
An alternative to constructing a new TEF within H-Area is to
refurbish and use the existing Allied General Nuclear Services (AGNS)
facility located in Barnwell County, adjacent to the eastern boundary
of SRS. AGNS was completed in 1976, and portions of the facility were
tested with natural uranium in anticipation of obtaining an operating
license to process commercial spent nuclear fuel. However, due to a
change in government policy on reprocessing commercial spent nuclear
fuel, the facility never opened. It was cleaned up and placed in
standby in 1977 and shut down in 1983. The AGNS facility was designed
and built to NRC standards. It would not meet all applicable DOE Orders
without major modifications as discussed below. Utilization of AGNS
would necessitate some new construction and some modifications.
Extraction furnaces would have to be designed, built, and installed. A
drying oven to remove pool water from CLWR TPBAR bundles or bundles of
targets of similar design unloaded in the wet basin would be required
(at AGNS, TPBARs would be stored in existing fuel storage basins). A
process gas stripper would have to be added to reduce stack tritium
releases. Although rail lines to the existing facility have been
removed, the tracks within the facility staging area and into the cask
unloading bays are still in place. Roads on the AGNS property need
moderate repair; and a short connecting road tying AGNS into the SRS
road system would have to be constructed. Other requirements include
refurbishing the heating, ventilation, air conditioning (HVAC) fans,
motors, high-efficiency particulate air (HEPA) filters and dampers; and
replacing the chiller water, fire protection, electrical, security, and
personnel protection systems.
3. No Action Alternative
Under the no-action alternative, DOE would not construct and
operate a TEF either at the preferred location in H-Area or at the
alternate location at AGNS. Under the no-action alternative, DOE could
incorporate tritium extraction capability into the APT facility at SRS.
However, because the use of existing commercial light water reactors
has been chosen as the primary tritium supply, selection of no action
for the TEF would result in the inability to extract tritium from the
irradiated TPBARs because an APT (with extraction capabilities) would
not be built. In that case, DOE would not be able to fulfill the
purpose and need for the proposed action. Such a decision would be
inconsistent with the December 5, 1995 ROD for the Tritium Supply
Programmatic EIS, as well as the programmatic decision documented in
this Consolidated ROD. Based on the supplemental Tritium Supply and
Recycling ROD, the no-action alternative for tritium extraction is
unreasonable and is not further discussed in this portion of the
Consolidated ROD.

B. Non-Environmental Comparison of Alternatives

1. Cost and Technical Factors

Cost

The life cycle cost estimate for the TEF at the preferred
alternative (H-Area) is $920 million compared to the AGNS facility
upgrades which is $1085 million. Both estimates are in constant FY 1999
dollars. Because of its close proximity to other tritium facilities in
H-Area, the H-Area alternative for TEF enables the sharing of common
support facilities, services, and some personnel; to facilitate the
transfer of tritium between the two facilities; and to use certain gas-
handling processes located in H-Area. Consequently the life-cycle cost
of operating the TEF at this location is less than AGNS. The AGNS
estimate exceeds the TEF estimate due to the added cost of logistics in
moving the tritium containing gases from the AGNS location to the H-
Area location for final processing and loading and the additional gas
processing equipment needed at the AGNS location.
Conclusion: Locating the TEF in the H-Area would have a lower life-
cycle cost than locating it at AGNS.

Technical

Several technical aspects were considered in evaluating the
alternatives. For the AGNS facility, these technical aspects included:
construction of several new buildings to house the gas processing
equipment needed (existing facilities were not large enough to house
the needed gloveboxes), installation of a drying oven to remove
moisture from TPBARs wetted during underwater cask unloading, the
addition of a waste processing facility, and an overhaul of the AGNS
ventilation system to facilitate the tritium gas processing
requirements. Technical factors involving the location of the preferred
alternative are: (1) To share common support facilities, services, and
some personnel; (2) to facilitate the transfer of tritium between the
two facilities; and (3) to use certain gas-handling processes located
in H Area.
The design basis of the Tritium Extraction Facility (TEF) requires
that tritium-containing gasses be supplied to the existing Tritium
Loading Facility (Building 233-H). Extracted gasses would not be
isotopically separated at TEF but would utilize existing equipment in
Building 233-H for separation of the hydrogen isotopes. In addition,
the TEF would not be designed to separate hydrogen and non-hydrogen
isotopes. The cost savings to the TEF project by not including this
separation equipment is approximately $50 million. If the TEF were
built at the AGNS facility, the TEF would have to include all of the
necessary separation equipment as well as the infrastructure required
for the facility (electrical, waste water, fire protection, staffing,
etc.). The hydrogen isotopic separation equipment would need to
``purify'' the extracted tritium-containing gasses prior to loading on
a hydride bed for transporting to the 233-H facility. Additionally,
utilization of AGNS would require the unloading of shipping casks
underwater which in turn would require the addition of a drying area
for the TPBARs prior to extraction. The introduction of water in or
around a tritium source greatly increases the hazard to operations
personnel in the form of tritium oxide, which is 10,000 times more
hazardous to humans than elemental tritium. However, collective doses
to the population are expected to equal those of the H-Area
alternative.
Conclusion: The ability of the preferred alternative to deliver gas
directly to the 233-H facility offers several technical advantages over
the AGNS alternative.

[[Page 26381]]

C. Comparison of Environmental Impacts of TEF Alternatives

In general DOE considers the expected impacts on the physical,
biological, and human environment for both reasonable alternatives to
be minor and consistent with what might be expected for an industrial
facility. In the comparison of impacts, DOE determined that changes
from current site environmental conditions of less than 5 percent are
within the margin of error and the conservatism inherent in the
analyses. Therefore, DOE finds that in those instances there would be
no measurable change from current environmental conditions. As
documented in the TEF Final EIS, overall, there are not expected to be
any significant differences in environmental impacts between the two
reasonable alternatives. Except for the no-action alternative, the
construction and operating impacts of the TEF would be added to the
impacts of the CLWR alternatives discussed in Section III above.
1. Construction Impacts
Minor differences between the alternatives are expected due to
construction. Because much of the AGNS alternative involves internal
modifications to an existing facility, less land would be disturbed and
less construction waste generated. However, because the land at H-Area
is already a densely developed, industrial area, impacts associated
with land disturbance are not a significant factor. With respect to
construction waste volumes, potential impacts to SRS waste treatment,
storage, and disposal facilities would be small for both alternatives
because of the low volumes of waste to be generated. At the AGNS site,
construction noise and activity could have localized adverse effects on
wildlife; however, this is not expected to be significant. Impacts
associated with socioeconomics would be similar as each alternative
would have a 5-year construction duration and a similar peak workforce
(740 for H-Area, 685 for AGNS). While the creation of these jobs would
have a positive stabilizing effect on the SRS employment, the overall
impact would be minor since either alternative would change the
regional employment by less than one-half of one percent.
Conclusion: Although the environmental impacts associated with
construction are considered small for both alternatives, the AGNS
alternative would have a smaller construction impact.
2. Operating Impacts
Operation of the TEF at H-Area or at AGNS could cause impacts in
the following areas: human health (normal operations and accidents);
waste generation; and socioeconomics. These areas are discussed below:

Human Health

A primary difference between the preferred alternative at H-Area
and the alternative at AGNS is AGNS's proximity to non-government land,
and therefore, its greater potential for impacting offsite individuals
due to releases near the site boundary. Additional differences include
stack height and radionuclides released to the environment. The
quantities released at AGNS would differ from those emitted at H-Area
because each rod would have to be cut three times in order to fit in
the AGNS furnace, while full-height TPBARs would be punctured at H-
Area. While processing CLWR TPBARs, the contributions of
nonradiological air constituents at AGNS would be 0.13 percent of the
applicable standard, and still lower for the onsite H-Area alternative.
The radiological dose for the offsite maximally exposed individual
would be 0.15 millirem per year for AGNS and 0.02 millirem per year for
H-Area. Both of these would be well below the regulatory annual limit
of 10 millirem from airborne releases. Because of the location of AGNS,
some minority or low-income communities could be disproportionately
affected by radiological and nonradiological air emissions; however,
such impacts are expected to be minor and within all regulatory
standards. Compared to the proposed action, for the maximally exposed
individual the AGNS alternative is projected to have a 0.13 millirem
per year higher radiation (due to its closer proximity to the boundary)
but nearly equal collective population doses.
With respect to impacts from potential accidents, the lower
population density in the communities near AGNS would result in a
slightly smaller collective doses from potential accidents. For each of
the alternatives, the design-basis accident would yield risks to the
50-mile population of approximately 7 latent cancer fatalities every
100,000 years.
Conclusion: Although the differences between the two alternatives
are not significant, the preferred alternative (H-area) would have a
lower impact on human health because of its greater distance from the
site boundary.

Waste Generation

Both alternatives would generate 232 cubic yards of waste annually.
The potential impacts to SRS waste treatment, storage, and disposal
facilities would be small because the volumes would be small relative
to existing waste management capabilities.
Conclusion: There is no apparent difference between the two
alternatives' generation of waste.

Socioeconomics

Because of its proximity to other tritium facilities in H-Area, the
H-Area alternative for TEF facilitates the use of common support
facilities, services, and some personnel. Consequently, the operations
workforce for the H-Area alternative is approximately 60 percent as
much as the AGNS alternative (108 versus 175). While the socioeconomic
impact for each alternative is considered minor, the reduced staffing
requirement for the H-Area alternative is a major factor in its reduced
life-cycle cost compared to the AGNS alternative.
Conclusion: Although the AGNS alternative would provide 67 more
jobs for facility operators, the difference is not significant.
3. Environmentally Preferred Alternative
As described in the TEF Final EIS, the potential impacts from the
preferred alternative or the AGNS alternative on the physical,
biological, and human environment would be minor and consistent with
what might be expected for an industrial facility. The preferred site
for TEF is within H-Area, a densely developed, industrialized area near
the center of SRS, approximately 6.8 miles from the nearest (western)
SRS boundary. There are four existing tritium-related facilities in the
immediate vicinity of the proposed TEF site. Advantages to locating TEF
within H-Area include minimal environmental impacts associated with
construction and operation of the proposed TEF due to the developed
nature of H-Area; availability of site infrastructure (i.e., power,
steam, potable water, sewerage); and proximity to existing tritium-
related facilities and processes to support TEF operations. Both the
nonradiological air constituents and annual radiological dose are lower
for the preferred alternative compared to the AGNS alternative.
Consequently, the H-Area alternative is the environmentally preferred
alternative.

D. TEF Decision

The preferred alternative, to design, construct, test, and operate
a new TEF in H-Area immediately adjacent to and

[[Page 26382]]

west of Building 233-H, at the SRS, is selected for implementation.
This alternative has the lowest life-cycle cost, has technical
advantages, and is environmentally preferred.

V. Site-Specific Decision for Accelerator Production of Tritium
(APT)

DOE has prepared this part of the Consolidated Record of Decision
to implement that portion of the December 22, 1998 announcement
designating the APT as the backup technology. It is based on the
analysis from the Accelerator Production of Tritium at the Savannah
River Site Final Environmental Impact Statement (DOE/EIS-0270) issued
in March 1999, along with other factors such as DOE statutory mission
requirements, national security policy, cost, schedule and technical
risks.

A. APT Design Features and System Alternatives Considered

The EIS evaluated the no action alternative, and technology and
siting alternatives relating to radiofrequency power, accelerator
operating temperature, feedstock material, cooling water system, APT
site, electric power supply, and APT design variations. The following
section summarizes these alternatives.
1. No Action Alternative
No action for the APT is to produce tritium in a commercial light
water reactor and to construct and operate a tritium extraction
facility. Under the no action alternative the APT is designated the
back-up technology for tritium production. As back-up, DOE would
complete key research and development, and preliminary design
activities for the APT at SRS (but would not construct the facility).
Selection of APT technology and siting alternatives would support the
research and development and preliminary design activities and
facilitate implementation should construction and operation of the APT
be called for in the future.
2. Radiofrequency Power Alternatives
APT would use radiofrequency waves to accelerate protons in the
accelerator. Specially designed vacuum electron tubes would convert
electric power to radiofrequency waves outside of the accelerator. The
waves are then transported into the accelerator and used to accelerate
the protons. The APT EIS evaluated two alternatives to supply the
radiofrequency power for the accelerator, (1) klystron radiofrequency
power tubes (DOE's preferred alternative), and (2) high order mode
inductive output radiofrequency power tubes.
3. Operating Temperature Alternatives
The operating temperature affects the electrical components in the
accelerator. The greater the power converted to heat the greater the
amount of electricity used. If the temperature of some materials (e.g.,
niobium) falls to values near absolute zero (-459 deg.F), the
electrical resistance becomes essentially zero, and the component uses
much less electricity. This is called superconductivity. The APT EIS
evaluated two operating temperature alternatives for the accelerator:
(1) operating electrical components at essentially room temperature,
and (2) operating high energy accelerating structures at
superconducting temperatures and the rest at room temperature (DOE's
preferred alternative).
4. Feedstock Material Alternatives
The feedstock material absorbs the neutrons freed during spallation
resulting in the production of a tritium atom and a byproduct atom. DOE
would use the same target/blanket as the neutron source regardless of
the feedstock material. The APT EIS evaluated two feedstock materials,
(1) Helium-3 (DOE's preferred alternative) and (2) Lithium-6.
5. Cooling Water System Alternatives
The APT requires cooling water to keep target/blanket components,
radiation shielding, beamstops and other components from overheating.
DOE proposes to use a similar method for cooling each component. This
is a primary coolant loop isolated from the environment through heat
exchangers. Components with the potential for radioactive contamination
would require a secondary loop to cool the primary loop and isolate
potential contamination from the environment. The final cooling system,
regardless of the number of loops, would use a cooling water system to
discharge heat to the environment. The APT EIS evaluated four designs
to provide the necessary cooling capacity for the APT: (1) Mechanical-
draft cooling towers with makeup water from the Savannah River and
discharge into pre-cooler Ponds 2 and 5 of Par Pond (DOE's preferred
alternative); (2) mechanical-draft cooling towers with makeup water
from groundwater wells and discharge into pre-cooler Ponds 2 and 5 of
Par Pond; (3) once through cooling using Savannah River water and
discharge into pre-cooler Ponds 2 and 5 of Par Pond; and (4) use the
existing K-Area cooling tower with Savannah River water makeup and
discharge to Pen Branch via Indian Grave Branch. A design variation for
the first three alternatives would be to discharge the heated water to
the head of Pond C of Par Pond but downstream from pre-cooler Ponds 2
and 5.
6. Siting Alternatives
DOE conducted a screening process to select potentially suitable
sites within the SRS for the APT. Based on a weighing and balancing of
the criteria, DOE selected two sites for further analysis. The APT EIS
evaluated (1) a site 3 miles northeast of the Tritium Loading Facility,
and approximately 6.5 miles from the SRS boundary (DOE's preferred
alternative); and (2) a site 2 miles northwest of the Tritium Loading
Facility, and approximately 4 miles from the SRS boundary.
7. Electric Power Supply Alternatives
APT requires large amounts of electricity to operate. Therefore,
DOE evaluated two alternatives for the source of electricity for the
APT: (1) Obtain electricity from existing commercial capacity and
through market transactions (DOE's preferred alternative); (2) obtain
electricity from the construction and operation of a new coal-fired or
a natural-gas-fired generating plant.
8. APT Design Variations
In addition to the cooling water discharge design variation
described above, the APT EIS evaluated two other variations. The first
is a modular, or staged, accelerator configuration. It would use the
same accelerator architecture as the baseline but could be constructed
in stages. An initial stage would produce less tritium than the
baseline APT but would be capable of producing as much tritium as the
baseline APT with the addition of a second stage.
The second variation would combine tritium separation and tritium
extraction facilities to take advantage of common process systems and
would be capable of handling both Helium-3 and Lithium-6 feedstock
material.

B. Non-Environmental Comparison of APT Design Features and System
Alternatives

Technical comparisons are presented for each set of alternatives
described above. These are based on various studies completed for each
alternative.

[[Page 26383]]

1. Technology Factors

Radiofrequency Power

The klystron is an established technology that has been used for
years. Thus, this technology has proven reliability and presents no
technical challenges to its use in the APT. The inductive output tube
has several commercial applications, but additional design and
prototyping is needed to demonstrate the applicability to APT. These
demonstration tests are scheduled for completion this spring. The
inductive output tubes have a greater efficiency in converting DC power
to RF power which would reduce power requirements by 15 percent. The
inductive output tube also uses one half of the voltage resulting in
reduced shielding requirements.
Conclusion: The preferred alternative of klystron power tubes would
be used as the basis for the preliminary design. The inductive output
tube offers technical advantages and reduces operating costs (less
electricity used) and capital costs (less shielding needed). The
continued development is justified to achieve these benefits.

Operating Temperature

The room temperature accelerator technology is based on technology
demonstrated at the Los Alamos National Laboratory. The accelerator
cavities are cooled by the primary water cooling system. As part of the
accelerating structures the cavity lengths would increase in size in
proportion to the increasing proton velocity. This results in greater
complexity of maintenance because each cavity is unique.
The superconducting technology uses two sizes of cavities which are
cooled with liquid helium to almost absolute zero. This cooling method
eliminates the need for water cooling in the superconducting cavities.
The two different sizes of cavities allows for simplified maintenance.
The engineering development and demonstration program has completed the
design and prototyping of these cavities.
Conclusion: The superconducting cavities allow for easier
accelerator maintenance. Experience has shown that liquid helium
distribution systems are less prone to leakage than water systems.

Feedstock Materials

Helium-3 is a nonradioactive gas that exists naturally in small
quantities and is produced through the radioactive decay of tritium.
The helium-3 is contained in tubes within the target/blanket. The
helium-3 would absorb neutrons which converts it to tritium and
hydrogen The helium-3 and tritium mixture would be continuously or
semi-continuously transported via piping to the Tritium Separation
Facility. The helium-3 purified in the separations process is returned
to the target/blanket to produce additional tritium. This results in
reduced inventories of tritium in the target/blanket and prevention of
pollution since the helium-3 is recycled. The production of tritium can
also be varied through controlling the number of neutrons but without
sacrificing continuous separation.
Lithium-6 would be in the form of rods that would be placed in the
blanket area. These rods would be similar to the rods DOE used when it
operated the SRS tritium production reactors. Because the lithium-6 is
incorporated into solid rods, batch production of tritium is required
resulting in a higher inventory of tritium in the target/blanket than
the helium-3 alternative. Also the rods could not be recycled.
Conclusion: The improved safety factors from reduced inventory of
tritium in the helium-3 alternative along with the ability to recycle
the helium-3 provides advantages for the helium-3 alternative. The
added flexibility of varying production rates also makes the helium-3
alternative attractive.

Cooling Water System

The cooling water system alternatives were evaluated using three
evaluation criteria, capital cost, life cycle cost, and permitting
risk. The mechanical draft cooling tower with river water makeup was
rated the lowest capital cost, the lowest life cycle cost, and the
least risk associated with obtaining permits. The evaluation of risks
associated with permits is based on the scope of changes to existing
systems that would require regulatory reviews as well as the
temperature of the blowdown water compared to the threshold limit. This
evaluation placed the mechanical draft cooling tower with river water
makeup as the best alternative. The mechanical draft cooling tower with
ground water makeup was ranked second, once through cooling was third
and the use of K-Area cooling tower was fourth. A separate evaluation
for the design variation of discharge to Pond C of Par Pond was also
completed. This evaluation showed a reduction in costs due to avoidance
of costs associated with upgrades to the pre-cooler ponds.
Conclusion: The mechanical draft cooling tower with river water
makeup was evaluated as the best alternative based on capital cost,
life-cycle cost, and permitting risk criteria. The design variation of
discharging to Pond C of Par Pond added the benefit of reducing costs.

Siting

The two sites evaluated in the EIS, a site 3 miles northeast of the
Tritium Loading Facility (northeast site), and a site 2 miles northwest
of the Tritium Loading Facility (northwest site), were similar in most
characteristics. No differences in engineering factors were identified
in the Site Selection Study (WSRC-TR-96-0279). The ranking factors
where there is a difference between the two sites were in ecology,
where the northeast site was better; depth to groundwater, where the
northwest site was better; and buffer distance to the public off-site,
where the northeast site was better.
Conclusion: Due to increased buffer distance which would reduce
public radiological exposure in the case of an incident, the northeast
site is a better location.

Electrical Supply

The two alternatives evaluated present different technical and
financial challenges. The alternative to construct a new dedicated coal
or gas fired plant would probably require both contractual and
financial guarantees by DOE to the utility providing the electricity.
Prior to a utility constructing a plant, the DOE would need to enter a
long-term power purchase agreement to provide assurance to the utility
that it would have a market for the output of the plant. The
contractual arrangement would therefore entail take-or-pay obligations
on the part of DOE for an amount of time necessary for growth in system
demand to absorb the generating capacity constructed.
In the alternative of relying on existing capacity and contracting
for power purchased on the market, the take-or-pay and/or notice-of-
termination provisions associated with a dedicated plant can be
minimized or entirely avoided. Shorter term retail sales contracts (2
to 5 years) can be accommodated which would permit the DOE to
periodically recompete the APT purchase arrangements. This would also
allow DOE to take advantage of renewable energy opportunities that
could become available in the future.
The electric power industry is presently subject to significant and
widespread changes, with approximately 40 states presently addressing
the issue of restructuring the retail power market to permit
competition among suppliers. A long-term power supply contract tied to
the construction of a dedicated generating

[[Page 26384]]

facility would eliminate DOE's flexibility in taking advantage of
changes in the power supply market over the life of the plant.
Conclusion: The alternative of purchasing power from the electric
grid through market transaction provides DOE with greater long-term
flexibility and avoids the need to commit to a long-term contract for
power.

Modular Design

The modular design was developed to provide tritium production
flexibility in the face of changing stockpile requirements, and to
optimize the project costs and funding profile. Several different
modular designs were evaluated using cost and schedule, technical and
programmatic risk, and the potential for future upgrades as general
criteria. The preferred design meets current requirements, but allows
for a delay in the decision to construct an APT that meets Start I
requirements for several years, while avoiding the commitment to the
cost of a START I sized facility.
Conclusion: The modular design provides the DOE with enhanced
flexibility to only commit to an APT sized to meet requirements in
several years.

C. Comparison of Environmental Impacts of APT Alternatives

The APT EIS presents an evaluation of environmental impacts for the
combination of the preferred alternatives identified above, and the
differences found for each of the alternatives. This summary presents
the same format for comparison of the environmental impacts.
1. Construction Impacts for the Preferred Technology and Site
Alternatives
APT would require conversion of approximately 250 acres of land
from forest to industrial uses. This land would be graded or leveled
during construction. Additional roads, bridge upgrades, rail lines, and
utility upgrades would be required. No geologically significant
formations or surface faults occur on the site. Soils on the site are
not classified as significant. The change in land use would have no
marked reduction in plant and animal abundance or diversity. There are
no impacts to wetlands or threatened or endangered species.
Impact to surface waters are negligible, however, dewatering of the
construction site could result in short-term increases in solids to
receiving water bodies. Impacts to aquatic organisms in Upper Three
Runs and tributaries would be minor due to the use of soil and erosion
control measures.
Air emissions would be negligible at the site, and purchases of
electricity would be dispersed. There are no radiological emissions
during construction. Visual impacts would be negligible. Noise,
primarily from construction equipment is not audible at the SRS
boundary, however, construction workers could encounter noise levels
that would require administrative controls or protective equipment.
APT would generate hazardous solid waste and sanitary solid and
liquid waste. These would be deposited at SRS, and would require some
landfill construction. Estimated annual volumes of waste are 560 cubic
meters of sanitary solid waste, 30,000 cubic meters of construction
debris, and 3.6 million gallons of industrial wastewater.
Impacts to public health during construction would be negligible
because concentrations of non-radiological constituents are below
applicable limits. Increased traffic would result in a small increase
in traffic fatalities. Occupational injuries are not expected to be
different than those occurring on any large construction site.
The work force required for construction is estimated to peak at
1,400 jobs. This would not result in large regional impacts.
2. Operational Impacts for the Preferred Technology and Site
Alternatives
No impacts would occur to landforms, soils, hydrology or geology
during operations. No dewatering is required for operations. Electrical
use is estimated at 3.1 terawatt-hours per year. Negligible impacts to
terrestrial ecology and threatened and endangered species are expected.
Mechanical draft cooling towers would result in salt deposition on
vegetation, however, maximum levels are below threshold levels.
Operations would result in minor impacts to wetlands due to marginally
higher temperature of blowdown water.
Blowdown rates of approximately 2,000 gallons per minute would
cause negligible impact on surface water levels. Using Par Pond and
pre-cooler ponds as discharge points for cooling water, temperatures
would not exceed 90 degrees F. Contaminated sediments could be
resuspended, resulting in negligible additional fatal cancers from
exposure to the public. Impingement and entrainment from intake of
river water would not substantially affect Savannah River fisheries.
Solids in blowdown water would have no impact on aquatic ecology.
Discharge temperatures would only have small localized effects on
aquatic communities.
Non-radiological air emissions would be well within the applicable
regulatory standards. Radioactive airborne emissions would result in
expected latent cancer fatalities of 0.0008 annually. There would be
negligible impacts to visual resources, with plumes visible under
certain meteorological conditions. Noise generated by equipment and
traffic would not be audible at the SRS boundary.
APT operations would generate solid and liquid wastes but no high-
level or transuranic waste; waste volumes would have a negligible
impact on the capacities of waste facilities. The generation of
electricity would produce various types of waste including fly ash,
bottom ash, and scrubber sludge. Estimated annual amounts of waste
generated are 1,800 metric tons of sanitary solid waste, 3,800 metric
tons of industrial waste, 140,000 gallons of radioactive wastewater,
3.3 million gallons of sanitary wastewater, 920 million gallons of non-
radioactive process wastewater, 1,400 cubic meters of low-level
radioactive waste, 3 cubic meters of high concentration radioactivity
low-level radioactive waste, and 12 cubic meters of high concentration
radioactivity mixed waste.
The public would receive source radiation exposure from APT
emissions and transportation of radioactive material. Workers would
receive radiation exposure from facility operations and transportation
of radioactive material and from electromagnetic fields. These would
result in an annual risk of 0.0016 latent cancer fatalities. There
would be negligible consequences from accidents with a frequency of
less than once in the operating lifetime of the facility.
The operational work force would be approximately 500. This would
not result in large regional impacts. No adverse impacts on minority or
low-income populations are expected.
3. Environmental Impacts of Alternatives

Radiofrequency Power Alternative--Inductive Output Tubes

This alternative would have no change in estimated impacts from the
preferred alternative for construction impacts. The only change in
operational impacts from the preferred alternative is in impacts to
surface waters. The

[[Page 26385]]

inductive output tube would require 7 percent less cooling water.

Operating Temperature Alternative--Operating Electrical Components at
Room Temperature

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative, except
that 100 fewer construction jobs are estimated, resulting in lower
regional community impact; there would be a 9 percent reduction in
sanitary waste generated; and there would be a 6 percent reduction in
occupational injuries. During operations electricity usage is estimated
to be 23 percent higher, and 37 percent more non-radioactive waste
water would be generated.

Feedstock Material Alternative--Lithium-6

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative. For
operations, the impacts would be similar to the preferred alternative
except for slightly increased doses from airborne radiological
emissions which would slightly increase the latent cancer fatalities.
Also, eight percent more low-level radioactive waste, and 25 percent
more high concentration mixed waste would be generated. A minor
decrease in radiological doses from accidents with low probability of
occurrence would also occur.

Cooling Water System Alternative--Once-Through Using River Water as
Makeup

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative.
Impacts from operations would also be similar, except blowdown rates of
125,000 gallons per minute (a 2,000 percent increase) would result in
higher temperatures to receiving bodies of water and would adversely
affect aquatic communities. Also an increase of 1.5 feet in the water
levels of the pre-cooler ponds would possibly affect wetland
communities. Impingement of 2,600 fish, and entrainment of 3.4 million
fish eggs and 6.4 million larvae annually would occur. Resuspension
caused by the increased flows would result in slightly increased doses.
Latent cancer fatalities would increase from 0.0016 to 0.0017 annually.
No mechanical-draft cooling tower noise would be heard at the APT site,
but pump noise would be occasionally audible to river traffic. No salt
deposition would occur.

Cooling Water System Alternatives--Mechanical Draft Cooling Towers
Using Groundwater Makeup

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative.
Impacts from operations would also be similar except the removal of
6,000 gallons per minute on a sustained basis could impact groundwater
flow to streams and compact clay layers. No impingement and entrainment
would occur.

Cooling Water System Design Variation--Discharge to Pond C Avoiding the
Pre-cooler Ponds

This design variation applies to the preferred alternative and the
two cooling water system alternatives above. This variation would have
no changes in the estimated construction impacts as described for the
preferred alternative. The operational impacts would be similar to the
preferred alternative, except that impacts to the pre-cooler ponds are
eliminated, and there would be a minor increase in heated water impacts
to Pond C.

Cooling Water System Alternatives--K-Area Cooling Tower Using River
Water as Makeup

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative except
the wastewater discharges would go to Pen Branch via Indian Grave
Branch. The water levels in the upper reaches of the stream system
would be raised. Additional cooling water piping to the K-Area would
also be needed. The plume from K-Area cooling tower would likely be
more visible. There would be no mechanical-draft cooling tower noise at
the APT site, but pump and cooling tower noise in the K-Area would
increase.

Site Location Alternative--2 Miles Northwest of Tritium Loading
Facility

This alternative would have no changes in the estimated
construction impacts as described for the preferred alternative except
the water table is deeper and would require less dewatering. Also
traffic fatalities during construction would be twenty percent less.
Changes in operational impacts from the preferred alternative are
higher doses due to closer distance to the SRS boundary. The dose from
all sources would increase latent cancer fatalities from 0.0016 to
0.0017 annually.

Electric Power Supply Alternative--Construct New Plant

The impacts of a new plant would be dependent on the specific
location. A new coal facility would require 290 acres and a natural gas
facility 110 acres. The types of impacts presented for the preferred
alternative would also occur at the specific site for a new plant.
Increased amounts of construction waste would be generated.
Construction would require a peak work force of 1,100. Plant operations
would require an additional 200 jobs.

Design Variations--Modular Design

This variation would have no changes in the estimated construction
impacts as described for the preferred alternative except construction
wastes, health impacts, and peak employment all would be 10 percent
lower. Operational impacts would also be similar with the following
exceptions. Both blowdown water rates and non-radiological air
emissions would be 10 percent lower. Electricity usage would be 2.0
terawatt-hours per year, a 32 percent decrease. Wastes from operations
would be 10 percent lower.

Design Variation--Combining Tritium Separation and Extraction
Facilities

This variation would have no changes in the estimated construction
impacts as described for the preferred alternative. Operational impact
differences would result in an increase in doses from airborne
emissions from 0.0008 latent cancer fatalities to 0.0009.

No Action Alternative

For the APT, no action is to not build the APT, but use the CLWR as
a source of tritium. Since the APT would not be built or operated there
would be no change in the existing environment at SRS.
4. Overall Environmental Conclusion
As described above, and as documented in the environmental analyses
that have been developed, it is expected that the overall environmental
impacts associated with tritium production in an APT would be small.
Consequently, the environmental impacts associated with the APT
alternatives and design variations are not considered major
discriminating factors in the decision. Based on all of the
environmental factors considered, the no action alternative is the
environmentally preferred alternative.

D. APT Decision

DOE selects the APT as the

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A99-12019. Public record. Not legal advice.