- DOE's "science-based" stockpile stewardship and management program
is oversized for its mission of indefinitely maintaining the safety
and reliability of U.S. nuclear weapons. This oversized program
is not just expensive but is leading directly to the introduction
of new, untested nuclear weapons designs into the U.S. arsenal, reducing
confidence in those weapons.
- Many aspects of the planned program are massively redundant, premature,
and tangential to the program's stated mission. The program
includes projects with dangerous ambitions to design new weapons and
contains plans to abrogate U.S. treaty commitments.
- Viable alternatives to DOE's proposed Stockpile Stewardship Program
exist that will cost less, are better suited to maintain the safety
and reliability of the arsenal, support U.S. arms control and nonproliferation
objectives, and have fewer environmental impacts.
The Department of Energy (DOE) has presented Congress with a proposed
FY1999 budget for nuclear weapons activities of $4.5 billion, up $353
million (8.5%) from its $4.147 billion FY1998 appropriation. Both
numbers are substantially more than the average Cold War cost for the
same activities (research, development, testing, weapon production, and
associated waste management), which, in today's dollars, was only $3.7
The unique political opportunity created by possible ratification of the
Comprehensive Test Ban Treaty (CTBT) has allowed DOE to overshoot the
"right size" for this program by a substantial margin.
Not only is DOE's program too costly, but it will also have the inevitable
result, barring direct congressional prohibition, of changing the nuclear
components in the fully-tested weapons which now make up the U.S. arsenal.
Such changes are almost certain to degrade confidence in weapon reliability
and could degrade weapon safety as well, for the simple reason that the
new designs, unlike existing weapons, would not be tested.
DOE's program has other costs as well. Many nations point to the
excessive scale, aggressive course, and concrete results of the program
(such as the introduction of the B61-11 earth-penetrating weapon into
the U.S. arsenal) as proof that the United States is insincere about its
long-term commitment to arms control and nonproliferation. In addition,
the huge new facilities DOE seeks to build and operate, and the large
quantities of nuclear waste the agency will generate, raise serious environmental
Maintaining nuclear weapons does not require substantial new investments
in "science-based stewardship."
Nuclear weapons are made of parts. More than 95% of the parts in
a weapon lie outside the nuclear portion of a weapon, the so-called "physics
package," and hence outside the domain affected by a nuclear test ban.
These non-nuclear parts have been, can be, and are fully tested
-- individually, as systems, and in integrated flight tests, independent
of full system nuclear explosive testing. The physics package itself
is sophisticated, but not very complicated.
Likewise, many components within the physics package, such as neutron
generators, tritium reservoirs, detonators, and high explosive assemblies,
can be fully tested with or without a test ban.
The only parts that cannot be tested for full function under a CTBT are
the so-called fission "primary" and thermonuclear "secondary." The
problem of maintaining weapons under a test ban boils down to maintaining
or replacing components of the primary and secondary as needed.
Test-ban-related safety issues arise only in relation to the primary,
as the secondary contains no high explosive. The safety of primaries
is an intrinsic property of the design, and is not increased by aging.
All U.S. weapons have been certified for their safety prior to entry
into the stockpile, and there is no way a CTBT could degrade the safety
of existing nuclear designs. Reliability problems have been exceedingly
rare in secondaries. The two or three that have occurred since surveillance
began decades ago have been readily solved by remanufacture or routine
maintenance. Test-ban-related reliability concerns are thus confined
to the primary stage only. All primary components, except the plutonium-containing
"pit," can be readily replaced with new, fully-tested components as needed,
using existing facilities.
According to Los Alamos managers, there are no short-term aging problems
that could affect the performance of pits for the first several decades
after manufacture. In addition, the nation has a reserve of more
than 10,000 extra pits, some of which can be used in stockpile weapons,
an approach already confirmed by nuclear tests.
In sum, the problem of maintaining the present large U.S. nuclear arsenal
indefinitely under a CTBT is not nearly as difficult a problem as DOE
is representing to Congress. Existing DOE facilities and programs
have the proven capacity to maintain the safety and reliability of the
U.S. stockpile. The expansive and expensive Stockpile Stewardship
program envisioned by DOE contains much that is both wasteful and dangerous.
Steps to a Better Program
Step one: eliminate redundant projects and facilities.
- DOE is maintaining, building, or planning: eight pulsed
- six dynamic radiography facilities;
- four inertial confinement fusion (ICF) facilities (DoD has
another);two nearly-duplicative billion-dollar physics laboratory
(Los Alamos and Livermore);
- two plutonium pit prototyping sites (Los Alamos and Livermore);
- three uranium component prototyping sites (the two physics
labs and a huge manufacturing facility, the Y-12 Plant in Tennessee);
- two separate development programs for tritium production
(hardly a scientific challenge); and
- three separate R&D programs in supercomputing (Los Alamos,
Livermore, and Sandia, not counting research at several universities).
But all too often, this "peer review" simply means that "I won't question
your unwarranted program if you don't question mine." Far less
expensive and more objective means of providing peer review include
the use of retired experts, internal review teams, and independent scientists,
not to mention providing the greater public transparency upon which
all forms of independent oversight are ultimately based. Some
of this redundancy has been sold as an expansive version of "peer review."
Some redundancy reflects territorial agency attitudes. For example,
DOE maintains or is planning a number of facilities for the long-term
storage of nuclear materials. The Nuclear Materials Storage Facility
(NMSF), built at Los Alamos in 1987, is one such facility. It
was so badly designed and constructed that it has never been used. The
program to rebuild it has muddled on for years and wasted millions without
even producing a workable design. The FY1997 line item which would
rebuild this facility is now sufficiently troubled to have become a
$32M "design only" request -- this for a facility whose total cost was
only $19.5M in 1987. The total cost of the facility is unknown.
Meanwhile DoD maintains, at the Manzano Base 90 miles away, a
bigger, more secure, better-designed storage facility -- one already
built -- than the DOE is planning to build at Los Alamos.
Step two: delay premature projects.
Example: tritium production. The DOE is requesting at least
$157M this year for the production of tritium, a radioactive gas used
in all U.S. weapons. Tritium decays with a half-life of about
12.4 years. Because current U.S. nuclear strategy requires retention
of a "hedge" arsenal of several thousand weapons beyond what can be
legally deployed under START II, U.S. tritium supplies will begin to
be "inadequate" in about ten years.
The existing Clinton-Yeltsin agreement to reduce total stockpiles below
the START II level would, if formalized, postpone the "need" for this
tritium and reduce the rate of any production required. The option
of purchasing tritium from Russia -- already a worthwhile option --
might then appear more attractive. Further arms control measures
(see step six) would allow DOE to further postpone, downscale, or eliminate
this program entirely.
Example: pit production. Another program which should be
delayed is the acquisition of greater capacity to produce plutonium
"pits" at Los Alamos National Laboratory. In 1997, DOE informed
Congress that, if funded, this project would cost about $1.1 billion
through FY2004, when a capacity of 50 pits/year was expected to be in
place, up from the 20 pits/year capacity DOE then said it already had.
This year's budget request asks for a $67 million down payment on this
project, plus an even greater amount in "hidden" pit production costs
contained in several line items for "maintaining infrastructure and
plant," several facility upgrades, and a new $71M security upgrade.
DOE now says these pit production costs are necessary to establish a
pit production capacity of just 20 pits/year by FY2007, the capacity
DOE once said LANL already had.
There is no need to produce pits. Los Alamos says no aging mechanism
will affect pit performance for at least two decades, if not longer,
and even if it did do so in some weapons, DOE says it can refurbish
and "requalify" some of its extra pits to give new working lifetimes
equal to the original ones. (4)
Step three: halt runaway projects.
Example: strategic computing. The Accelerated Strategic
Computing Initiative (ASCI) is an example of a runaway program.
Although the program is only a few years old, the FY1999 request for
"ASCI and stockpile computing" is already $518M, and DOE expects to
be spending a staggering $754M per year on this program by FY2003.
In the present "anything-goes" climate, yesterday's computing fantasy
has become today's "need." For example, Los Alamos now says it
"needs" a computer so huge that it will require 20-30 megawatts of power
and hence its own cooling tower(s). The building that will house
this machine is expected to cost $100M, not including the computer itself.
For comparison, the most modern weapons in the arsenal today were
designed with computing power roughly comparable to today's engineering
Example: DARHT. Another runaway project is the Dual-Axis Radiographic
Hydrotest (DARHT) facility at Los Alamos. Originally budgeted
at $53M for a complete facility with two axes, this year's budget request
gives a total project cost (TPC) of $270M, not including related road
upgrades, plutonium recovery and processing requirements, and other
This year's cost escalation alone is $73M. DARHT's TPC is almost certain
to rise further, as the project now provides for a second axis whose
capabilities far exceed those of the first. These costs do not include
operational costs of over $18 million per year.
Step four: reconsider tangential projects.
DOE is planning to spend billions of dollars on a variety of machines
to obtain more detailed data on the operation of thermonuclear "secondaries."
This is the chief, although not the only, purpose of the inertial
confinement fusion (ICF) program and of the $1.2 billion National Ignition
Facility (NIF), as well as of the numerous pulsed power facilities at
all three DOE weapons laboratories. NIF's total lifecycle cost
has soared to over $5 billion.
Even if NIF and the other machines actually achieve their goals -- an
open question -- the data they will supply is not needed. DOE
maintains a huge manufacturing facility for secondaries at the Y-12
plant in Tennessee, and a much more cost-effective and reliable
approach to maintaining the reliability of these parts is simply to
make them whenever they are needed, or to requalify older secondaries
(see note 4, below).
During the past year, numerous weapons scientists have stepped forward
to say that the safety and reliability of the arsenal is in no way contingent
upon the completion or operation of NIF. When pressed, DOE argues
that NIF and other tangential ICF and pulsed power projects are needed
to test the quality of the scientists, not the weapons.
There are far less costly and dangerous ways to do this.
Step five: scale back unrealistic and dangerous ambitions
Congress needs to tell the DOE that its job is to maintain and if
necessary reproduce the current arsenal for as long as is needed --
and that its job is not now, and will never be, to design, certify,
and produce a brand-new and inherently untestable arsenal.
The U.S. possesses about 12,070 intact nuclear weapons today, of which
8,420 are operational, 2,300 are in long-term storage as part of a
reserve or "hedge" arsenal, and 1,350 are awaiting dismantlement.
Operational and reserve weapons are of nine major types, in 13 different
configurations, many of which have a selectable explosive yield, ranging
from the very high to the very low.
(5) All these weapons have been fully tested, and all have
been repeatedly certified as safe and reliable. Our existing
nuclear weapons are deterrence enough. No new or newly modified
designs will change this.
New and modified weapons come in three types:
1. Those with new or modified "physics packages" (e.g.
a replacement submarine-launched weapon, now under design at all three
2. Those with new military characteristics created by innovation
external to the physics package (e.g. the B61-11 earth-penetrator,
deployed in early 1997, and the proposed nuclear glide bomb); and
3. Additional weapons of an existing type, as for example if
existing W76 Trident warheads were gradually replaced with higher-yield,
hard-target-capable W88 Trident II warheads.
All these attempts to "re-optimize" the arsenal and "exercise" the weapons
complex carry a heavy price, either in confidence (the first), in arms
control, nonproliferation, and related U.S. diplomatic objectives (all
three), or in the required manufacturing infrastructure, with its associated
costs and impacts, (the first, if pits are not reused, and the third).
Step six: reflect on U.S. treaty commitments
and their security value.
Article VI of the Nuclear Nonproliferation Treaty (NPT) commits the
U.S. to good-faith negotiations leading to complete nuclear disarmament.
This promise was the price we paid for a ban on nuclear weapon
development in non-nuclear-weapon states prior to the NPT. In
1996, the International Court of Justice unanimously agreed that there
was a binding legal requirement for the nuclear weapon states to successfully
conclude these negotiations, leading to total nuclear disarmament.
Despite this, the DOE's stockpile stewardship program devotes very
large resources to obtaining background data on "problems" that will
not affect the stockpile for decades to come, if ever. One example
is the program of underground "subcritical" nuclear tests, whose stated
purposes are 1) to collect data on the long-term performance of plutonium
components, and 2) to prepare for CTBT breakout.
The political costs of such a breakout would be devastating. The NPT
was indefinitely renewed in 1995 only after the nuclear weapon states
agreed to complete a CTBT by the end of 1996, and any resumption of
nuclear testing would cast both the CTBT and NPT into the dustbin.
U.S. testing would also provide China with a "green light" for the
testing it needs to miniaturize its warheads, encourage threshold
nuclear states to weaponize their nuclear devices, and downgrade other
security-related treaties as well.
The DOE's budget request includes a "rapid reconstitution" manufacturing
plan, one that would allow for the large-scale manufacturing of nuclear
weapons should tensions between the U.S. and Russia rise again. This
treaty breakout plan ignores the reality that Russia is now so economically
crippled that the number of nuclear weapons it can maintain and actually
deploy falls further each year, a process largely independent of treaty
negotiation and ratification. These processes are expected to
result in a Russian arsenal considerably smaller than that allowed
under START II by 2008. (6)
1. Stephen Schwartz, "Expenditures for U.S.
Nuclear Weapons Research, Development, Testing and Production--1948--1998,
2. For one example, see the DOE Congressional
Budget Request, Vol. I (Weapons Activities), p. 78, under "Future
3. A better policy would be to make any minor
changes that may be needed because certain materials or manufacturing
processes are no longer available now, before the original
designers, who could certify those changes, are no longer around,
as senior Livermore physicist and weapons designer Ray Kidder has
argued. See Ray Kidder, "Problems with Stockpile Stewardship,"
Nature, April 17, 1997, pp. 645-647.
4. "Reuse of pits in new primaries has been
demonstrated with successful nuclear tests of several designs that
also incorporated fire resistant shells and insensitive high explosive.
Pit reuse offers the possibility of manufacturing new primaries at
a reduced cost without requiring an expansion of currently planned
Pu fabrication facilities....It is expected that in the next few decades,
most plutonium pit requirements can be satisfied through the requalification
and reuse of existing, older pits....To meet refurbishment needs for
the foreseeable future, methods are being developed to allow pits
and secondaries to be requalified and placed back into the stockpile.
These requalified components are expected to have a design life comparable
to the original weapon life. Facilities will be established
at the weapon assembly site [Pantex] by the end of the decade to accomplish
this work." DOE, Stockpile Stewardship and Management Plan,
February 29, 1996, pp. V-10, 16, and 18.
5. William Arkin, Robert Norris, and Joshua
Handler, Taking Stock: Worldwide Nuclear Deployments 1998,
Natural Resources Defense Council, March, 1998.
6. Ibid., pp. 11-13.