5 Magnetic Fusion Reactor Economics

G
5
Title: MAGNETIC FUSION REACTOR ECONOMICS

~ Submitted to:

16th IEEE Symposium on Fusion Engineering Champaign, IL 61820 September 30-October 5, 1995

!

LosAlamos N A T I O N A L L A B O R A T O R Y
Los Alamos NafionalLaboratory, an affirmativeactionlaqua1opportunity empldyer. is operatedby the Universityof Californiafor the US. Departmentof Energy under contractW-74q5-ENG-36.By acceptance of this article, the publisher recognizesthat the US. Government retainsa nonexclusive,royalty-freelicense to publish or reproduce the publishedform of this contribution, or to allow othersto do so,for US. Government purposes. The Los Alamos National Laboratory
requeststhat the publisheridentifythis article as work performedunder the auspices of the U. Form No. 036 R5 ST 2629 10191
DISTRIBUTION^ OF THIS DOCUMENT IS UNLIMEDi,&

Portions of this document may be illegible in electronic image products. h a g = are
produced from the best available original
document.
':
TOSSII TUel anarthe environmental costs imposed on its use __- -

MAGNETIC FUSION REACTOR ECONOMICS
Robert A. Krakowski Los Alamos National Laboratory, Systems Engineering & Integration Group
P.O.Box 1663, MS F607, Los Alamos, NM 87545

ABSTRACT
An almost primordial trend in the conversion and use of energy is an increased complexity and cost of conversion systems designed to utilize cheaper and more-abundant
fuels; this trend is exemplified by the progression fossil +
fission 3 fusion. The present projections of the latter indicate that capital costs of the fusion "burner " far exceed any commensurate savings associated with the cheapest and most-abundant of fuels. These projections suggest competitive fusion power only if internal costs associate with the use of fossil or fission fuels emerge to make them either uneconomic, unacceptable, or both with respect to expensive fusion systems. This "implementationby-default" plan for fusion is re-examined by identifying in general terms fusion power-plant embodiments that might compete favorably under conditions where internal costs (both economic and environmental) of fossil and/or fission are not as great as is needed to justify the contemporary vision for fusion power. Competitive fusion power in this context will require a significant broadening of an overly focused program to explore the physics and simbiotic technologies leading to more compact, simplified, and efficient plasma-confinement configurations that reside a t the heart of an attractive fusion power plant.
INTRODUCTION

will push the associated COE into and eventually beyond the range presently being projected for tokamak-based fusion power plants.
While waiting for the competition to price itself out of the market within 50-100 years represents one marketpenetration plan for fusion, other less-fortuitous scenarios can be envisaged (e.g., fission succeeds, solar photovoltaichydrogen competes with both fusion and unconventional fossil). These scenarios suggest a better understanding is needed of the causes that drive upward the cost of (tokamak) fusion power. Means must be found by which these costs can be reduced while assuring an environmentally and publicly acceptable product on a reasonable time schedule and for a reasonable development cost. These cost drivers and the means by which they can be amelic-rated are addressed in a context where fission does solve its problems, breakthroughs in solarhydrogen occur, and/or the real costs of fossil fuel do not escalate out of the range of future competition. While, fusion market-penetration studies that accommodate both an escalating competition and (tokamak) concept improvements have been reported [2], the present investigation focuses on concept improvements need in the event that the costs of alternative energy sources do not escalate significantly, in which case new fusion approaches may be required.

Because the central role of fusion reactors has been as a purveyor of electrical energy to commercial and public sectors in competition with other means of electrical power generation, the projected cost of that product [Cost of
Electricity, COE( 1.O mill/kWeh = 3.6 $/GJ)] has served as
an important figure-of-merit since the inception of fusion reactor studies. When possible, environmental and' safety advantages suggested for fusion over other long-term energy sources (e.g., nonconventional coal, nuclear fission) have been translated into potential economic advantage with which to counter the added costs associated with generally massive (low-power-density) and/or powerintensive (low engineering energy gain) fusion power cores (FPCs) that characterize many designs based on Magnetic Fusion Energy (MFE) concepts. Detailed reactor studies based largely on a range of scientific and technological extrapolations of the tokamak suggest that environmental and safety advantages are insufficient to project competitiveness with advance fission systems that are safe, licensable, publicly acceptable, and operable within a closed fuel cycle [I]. Consequently, two scenarios are suggested to justify continued investment in the present magnetic-fusion R&D direction: a) advanced fission will not achieve the above-mentioned goals while remaining economic; b) advanced fission is disallowed for environmental or safety reasons, and the escalating cost of fossil fuel and the environmental costs imposed on its use

DESIRABLE ATTR~BUTES
The "optimal " fusion power plant can be described as follows in terms of flexibility, competitiveness, simplicity, and safety attributes:
flexibility in (net-electric) power output at acceptable (competitive) cost: total cost, unit costs, development and implementation cost; flexibility in end-product delivered (electricity, process heat, hydrogen, nuclearwaste transmutation, fissile fuel); competitive energy-generation costs: acceptable (high) power density, high overall efficiency (high thermal-toelectric conversion, low recirculating power), simplicity of operation and maintenance (reduced and/or combined plasma support functions, few- or single-piece FPC maintenance of the fusion power core), high availability; overall design and operational simplicity: steady state, reduced and/or combined plasma support functions, few- or single-piece FPC maintenance, reduced radioactivity (active inventory and waste stream); enhanced safety and environmental attributes: inherently or passively safe, reduced radioactivity (active inventory and waste stream), acceptable resource (raw-material) commitment.

J

, Many of these desirable attributes are counteracting and cannot simultaneously be maximized. Additionally, different attributes share common elements (e.g., the elements of reliability, availability, and maintainability contribute not only to the competitiveness attribute, but are important to most of the others also).

The development of commercial MFE power plants that exhibit these attributes can learn much from the experience
of U S fission power-plant developers. While many of the
problems faced today by fission power were not controllable (e.g., cheap fossil fuel and high discount rates, both of which impact a capital-intensive fission power plant that offers primarily a reduced fuel charge), many of these problems have been driven from within the fissionpower community [3]and in one form or another can be attributed to: appraisal optimism; premature choice (focus); and cost of complexity. The (commercial) development history of fusion is not sufficient to assess the impact of appraisal optimism (Le., projecting a surprisefree future and anticipating large savings compared to past projects), but the development history of nuclear fission presents ample cause for concern. An early indication of “appraisal optimism I‘ creeping into MFE projections, even at the preconceptual design level, is the increase in cost projections away from a competitive position reported preliminarily in [4] versus subsequent more-detailed and realistic studies [1,5],which in themselves have projected increased cost as the designs evolved. Comparing the complexity of an operating fission power plant with that of a conceptual fusion power plant based on a linear extrapolation of the present leading fusion concept [6] increases cost-of-complexity concerns even more. At the present stage of MFE development, the economic concerns driven by appraisal optimism and system complexity would not be as great had not the choice to focus and reduce opportunities for serious corrective action been made.
DESIRABLE DIRECTIONS
A. Status
A change in the direction of fusion development is needed to ameliorate the cost-related concerns discussed in the previous section and summarized quantitatively in [I]. To some extent, these changes are reflect in recent shifts to advanced tokamak physics [7], as embodied primarily in
ultra-low-aspect-ratio geometries [SI and plasmas with reversed-shear magnetic-field profiles [9]. Interim
projections of advanced tokamaks based on reversed-shear
physics indicate [4]COEs close to and possibly below values estimated for the advanced ARIES-IYIV tokamak concepts [I]; Fig. I summarizes these ARIES results and
makes a comparison with a number of fission and fossilfuel power plants [IO].
As elaborated in [I], “economic competitiveness” is measured against an advanced nuclear power plant that is assumed: a) to be accepted by the customer (utilities, power generators, and ultimately the public); b) to be licensed in an acceptable period of time; and c) to have developed and implemented safe and economic means to

-,
Fig. 1. Histogram of Cost-of-Electricity (COE, 1992)
values projected for both ARIES [I] and a range of
fossil- and fissile-fuel power plants [lo] of
comparable net-electrical capacity, PE = 1.0-1.2 GWe (PWR = Pressurized-Water Reactor; I = Improved; A = Advanced; MU = Multiple Units; ARIES-f = first-stability-region tokamak; ARIES-
IYIV = second-stability-region tokamak (different blankets); ARIES = second-stability-region D-3He
tokamak).
close both ends of the nuclear fuel cycle. While a possibility for fusion to exploit opportunities for enhanced public acceptance, reduced licensing burden, generation of a more acceptable (radioactive) waste form, and an economic “closure” of the nuclear fuel cycle, this possibility is a matter of conjecture at this time. Similarly, the competitiveness of fusion with fossil fuel over the next century will depend both on the projected cost of fossil fuels from both conventional and unconventional (e$., synfuels) sources, as well as the implementation and severity of carbon taxes; estimate of the former have been made as an aide to defining better the fusion economic

,

window [2], but scarcity-driven fuel-price increases for all fossil fuels remain to be detected [111.

B. Direction Finders

1) Global Energy Assessments: The COE values used for comparative analyses are derived from technology-based economic assessments, wherein physics and technology constraints are imposed to arrive at a constrained cost optimum for a given set of physics, engineering, materials, and costing assumptions. These analyses yield a discounted COE for comparison with, but in isolation from, other
contributors to a regional energy market (Fig. 1). Within a limited scope, however, the cost-benefit analyses reported in [I21 has been performed in a global context, wherein the economic impact of fusion on the total mix of available energy-producing technologies is estimated for a given
COE assigned to a new technology like fusion. The use of this forced market-equilibrium model [13] to assess the
(global) benefit of fusion introduced at a given COE (obtained from a separate technology-based economic assessment of the tokamak-based reactor [6]) gives a view of the impact of COE on the viability of fusion electric power that is broader than that provided by a “one-on-one” comparison of COE (Fig. 1). Fig. 2 displays the (global) incremental Gross National Product, AGNP(B$), as a function of the COE assumed for fusion for a range of economic (e.g., discount rate, carbon tax) conditions in a situation where fusion as a new technology impacts the global energy mix through an ability to shift the secondaryenergy (e.g., solids, liquids, gases, and electricity [131) supply curve. Instead of comparing COE values amongst competitive fuels and electricity generators, the global
energy/economics/environmental (E3) model compares net present-value GNP to the cost of developing fusion to generate electrical power at a given COE ascribed to fusion by the technology-based economic assessment (e.g.,
- ARTES). As seen from Fig. 2, the sensitivity of net benefit
to fusion COE is strong (AGNP l/COE2.4-3*9,depending on the discount rate and the taxhariff attached to carbon
burning). More recent results from this global E3 assessments have been reported [141.

-,

Fig. 2.

Net (Global) Value of Fusion, AGNP(B$), as a

Function of Projected Cost of Electricity,

COE(mills/kWeh, 1976) for a Business-as-Usual

(BAU) Case and a Carbon Tax (CT) Case under

- two assumptions of discount rate, x; respective

slopes, v , are indicated, where AGNP plotted from values reported in [12].

COEV;

2) Technology-Based Assessments: Use of the above-
described E3 global model is a logical step after the physics and engineering dependence of COE is assessed using technology-based economic studies like ARIES [1,15]. While cost-base systems models like ASC [15] or Supercode [16] incorporate all key physics, engineering, and materials models and constraints to arrive at economic optima and to elucidate relevant trade-offs, a simplified “gauge” model can provide valuable guidance for improved economic prospects for MFE. Specifically, a toplevel costing model [17] is used to project the cost of electricity on the basis of two highly aggregated reactor parameters: the mass power density, MPD(kWe/tonne), and the engineering energy gain, Q. Fig. 3 gives a cost
and functional condensation of a generic MFE power plant into Site (SITE), Fusion Power Core (FPC), HeaTinG (HTG), and Balance of Plant (BOP) power-plant “macrosystems”. Unit costs for each of these macrosystems

are used along with the indicated plant energy balance to give [17] the following relationship between COE, MPD,
and QE:

r

ir

1

where pf is the plant availability factor; ACRj(l/yr) are
annual charges related to capital, indirect, Operations and Maintenance (O&M), and Decommissioning and Decontamination (D&D) costs; UCj are aggregated unit
costs associated with macrosystems, qm is a thermal-to-
electric conversion efficiency, and gj is a “Jacobian” that
assures all powers are ultimately reduced to electric units.
For the typical parameters listed in Table II of [17], Fig. 4

J

FPC ReactorEquipment,(=.I.)

P,, = P#/O, + 0.8 M, +0 2 )
- pn =%tlpm
pE=P ~ iI/aE)

E Pdmav StNCtureand support
L! lmpudty Control Dlrect EnergyConversion

Heat Rejedion

Fig. 3.

Condensed MFE Reactor Power Flows and
Nuclear Costing Structure. Pj are power, (N = neutron, F = fusion, TH = thermal, ET = total electric, E = net electric, HTG = heating, AUX = plant auxiliary, C = recirculated).

illustrates this dependence of MPD on Q E for a range of Q values. This “gauge” contains no physics, which is provided by concept-specific technology-based assessments, the results from which are also indicated [ 1,15,18-221.
C . Directions
This COE comparison using the MPD versus QE cost metric suggests directions for improved commercial prospects. Assignment of concept-specific attributes and limitations expressed in the broader terms listed above remain for a more detailed study. While limited in scope, however, the approach and results presented in Fig. 4 serves as one of a number of lodestones with which to guide MFE research along more optimal path to competitive commercialization: high mass power density (MPD 2 500 kWe/tonne) and high engineering gain (QE 2 6). The increase ‘of the advanced-tokamak (e.g., reversed magnetic shear with high bootstrap current, as
approximated in Fig. 4 by ARIES-II + ARIES-II*) and an
number of poloidal-field-dominated systems (PFDs, high
engineering p for nearly self-confined plasma, along with
efficient current drive vis a‘ vis efficient injection of magnetic helicity, as modeled by the TITAN [16] reversedfield pinch or the CSR [19] spheromak) offer candidate systems.

-I

Fig. 4.

Parametric dependence of Mass Power Density,
MPD(kWe/tonne) on Engineering Gain, QE,
showing comparisons between ARIES steady-state
tokamaks [1,18]; the PULSAR pulsed tokamak [15]; the TITAN reversed-field pinch [19]; the HELIAC [20] and HSR [21] stellarators; a highbeta ARIES-II* [l]; and the CSR [22] spheromak reactors.

SUMMARY

The price that must be charged by any producer of electrical power to pay for all annual capital, O&M, and fuel charges must be reflected in the bus-bar COE. As seen from Fig. 1, a major part of the cost for an MFE power plant is associated with the intense capital investment required to bum a cheap and abundant fusion fuel. This increase in capital charges needed to utilize fuel with ever increasing resource, in fact, follows the progression fossil
+ fissile + fusion. For the MFE system on which the
world R&D program is now focused, the escalating capital cost is projected to outstrip any potential savings in reduced fuel charge. That increased capital charge has two sources: a) increased cost of the Fusion Power Core reflected in the relatively low MPD values; and b) the increased capital charges associated with an oversized balance-of-plant if the engineering gain is insufficient (e.g., the recirculating power fraction, 1 / Q , is too large). Equation (1) and Fig. 4 quantify the related impact on COE.

Wliether accessed through advanced tokamak physics [8,9,15], PFD systems [19,22], or other less-developed but more compact systems [17,231, the associated configuration and confinement physics must function simbiotically with the engineering and materials constraints imposed by high-power-density operation in a
system that is efficient [e.g., both high & and high qm,
leading to high plant efficiency, q, = (1-UQp) qm, as
indicated on Fig. 31, passive with respect of afterheat cooling, and manageable in terms of quality and quantity of radioactive waste. The following generic attributes for such a competitive system are listed as follows:

highly radiating plasmas (bulk plasma, edge plasma,

or both) to alleviate divertor heat-transfer

requirements; the divertor, ideally, should only handle

particles;

high neutron wall loadings (10-20 MW/m2 DT-fusion

neutrons) with the commensurate technology and

configurational capability to operate with:
- high-power-density blankets that more than likely

- preclude solid tritium breeders and gas cooling; annual changeout of entire first wall and innerblanket structure;

material and configurational choices that assure ES&H

attributes are maintained;

all components designed to acceptable engineering

criteria at high (but generally more uniform) power

density to maintain operational reliability;

approach ‘few’-piece (or ideally, single-piece)

maintenance and accrue major benefits related thereto:
- factory fabrication of (nearly) fully operational

- FPCs; fully operational pre-service, non-nuclear FPC testing;
- minimum in situ electrical, fluid, vacuum

connections during FTC replacement;
- shortened scheduled maintenance/replacement

period;
- standard and rapid recovery from unscheduled

events related to major FPC malfunction;
- ability to incorporate major physics and technology

-

advances into FPC during life of plant
neutron-damage life times that exceed

15

MWyr/m2.

While generic in nature and in need of demonstration for specific confinement systems, many of the attributes listed in Sec. 11should come to fruition if these characteristics of high-MPD, high-Q, systems can be achieved. Furthermore,
these characteristics may allow some of the assumptions (e.g., 75% plant availability) necessary to obtain even the COE values listed in Fig. 1 to actually be achieved.

CONCLUSIONS

The focus of these projections and prognoses has been on fusion as a provider of electrical energy to improve the living conditions (e.g., prosperity and security) of a growing

world population (1.8%/yr over the period 1989-90; projected [25] to decrease to 1.4%/yr over the period 19902020) having hopes of achieving a per capita energy utilization comparable to that of North America (NA) plus Western Europe (EUR) [13.8% of the world population and 41.1% of the 8,807 Mtoe (tonne oil equivalent and equal to
42 GJ) energy consumption in 1990; 5.0 toekapita (NA +
EUR) and 7.8 toekapita (NA), compared to a world-wide average of 1.7 toekapita]. Electricity, as one of four secondary-energy sources (Le., gases, liquids, solids, and electricity) accounted for 11.3% (11,607 TWh or 1,325 GWyr) of the secondary energy used in 1990 [23] (ranging from 13.8% in NA to 6.6% in Sub-Sahara Africa), with the linear growth rate of this percentage being -0.1 l%/yr.
The energy demand required for a global increase of living standard, as measured by the ratio GNP/capita (18,559 $/capita for NA, 13,403 $/capita for EUR, and 2,144 $/capita for the Rest of the World) [21]), will depend on the efficiency with which energy is utilized (toe/k$ or MJ/$) to achieve and maintain that standard of living, with environmental and other internal costs of each component to the energy spectrum being accounted; if the energy efficiency needed to build an infrastructure required for improved living conditions is increased (e.g., reduced energy intensity, toe/k$) compared to past experiences [25], global energy demands can be significantly reduced , compared to linear projections. Much of this nevertheless significant energy requirement, however, will be nonelectric; it seems prudent, therefore, for fusion correspondingly to broaden its end-use spectrum to either directly or indirectly (Le., in simbiosis) contribute to these future non-electric needs. An expanded niche for fusion may also deal simbiotically with cost and complexity issues related to fusion as an “on-line “ producer of
electrical power. This broadened role for fusion can
become even more important in any future that, through carbon taxes, carbon sequestering, or complete banishment, limits carbon burning.
Magnetic fusion reactor economics have been addressed primarily at a technical level, wherein the least expensive system that meets safety and environmental goals would ultimately enjoy the largest portion of the market. As pointed out in [26], however, straight economics rarely plays a central role in deciding which energy resource to develop. Important quasi- or non-economic considerations that factor into the choices of which energy paths to develop include [26]: a) the poliiical need to control balance of payments, import vulnerability, and energy dependence; b) internal and international pressures related to the environment; c) merits of international cooperation not having direct economic roots; and d) a range of economic/costing biases and/or distortions related to inconsistenthon-uniform assumptions and hidden subsidies. Nevertheless, a more affordable means of meeting energy needs is an important ingredient in presenting an economically, environmentally, and politically manageable solution to the long-term global energy problem; at this early stage of fusion reactor development, the projection of a versatile and economic commercial

- end-product would go a long way in attracting the attention
fusion deserves as a long-term solution to global energy needs.
REFERENCES
[ l ] R. A. Krakowski, C. G. Bathke, R. L. Miller, and K. A Werley, “Lessons Learned from the Tokamak Advanced Reactor and Evaluation Study (ARIES),” Fus. Technol., 26(3, Pt 2), 1111 (1994). [2] R. L. Miller and the ARIES Team, “Starlite Economics: Requirements and Methods,” Proc. 16th
IEEWPSS Symp. on Fus. Eng., Urbana, IL, (September 30 -
October 5, 1995). [3] G. MacKerron, “Nuclear Costs: Why Do They Keep Rising?,” Energy Policy, 20(7, Pt l), 641 (July 1992). [4] J. P. Holdren, et. al, “Report of the Senior Committee on Environmental, Safety, and Economic Aspects of Magnetic Fusion Energy, ” Lawrence Livennore National Laboratory report UCRL-52766 (September 25, 1989). [5] C. G. Bathke, R. L. Miller, “The Evolution of ARIES-I Design in the ARIES-I and PULSAR Programs,” Proc. 16th
IEEWNPSS Symp. on Fus. Eng., Urbana, E,(September 30 -
October 5, 1995). [6] R. S . DeVoto, et al., “Projections for a Steady-State Tokamak Reactor Based on the International Thermonuclear Experimental Reactor,” Fus. Technol., 19, 251 (1991). [7] C. G. Bathke and the ARIES Team, “A Preliminary Systems Assessment of the Starlite Demo Candidates,” Proc, 16th IEEWPSS Symp. on Fus. Eng., Urbana, IL, (September 30 - October 5, 1995). [8] Y.-K. M. Peng, “Prospects and Status of Low-AspecRatio Tokamaks,” Trans. Fus. Technol. 27, 138 (1995). [9] C. Kessel, J. Manickam, G. Rewoldt, and W. M. Tang, “Improved Plasma Performance in Tokamaks with Negative Magnetic Shear,” Phys. Rev. Lett. 72(8), 1212 (1994). [101 J. D. Delene, “Updated Comparison of Economics of Fusion Reactors with Advanced Fission Reactors,” Fus. Technol., 19(3), 807 (1991). [l 13 N. D. Uri, “Energy Scarcity and Economic Growth Reconsidered,” Intern. J. Energy Research, 19(9), 615 (1995). [I21 J. F. Clarke, “The Cost and Benefit of Energy Technology in the Global Context - The Case for Fusion Power,” Proc. Con8 of Technology Response to Global Environmental Changes: Energy Collaboration of the 21s’ Century, p 521, Kyoto, Japan (November 6-8, 1991). [13] J. A. Edmonds and J. M. Reilly, Global Energy: Assessing the Future, Oxford University Press, New York (1985). [14] J. F. Clark, “Economics, the Environment, and Fusion R&D Strategy,” Proc. 16th IEEEAVPSS Symp. on Fus. Eng.,
Urbana, IL, (September 30 - October 5, 1995).
[15] ARIES Team, “The ARIES-II and ARIES-IV Tokamak Reactor Study Final Report,” University of California at Los Angeles report UCLA-PPG-1461 (to be published). [16] S . W. Haney, et al., “A Supercode for Systems Analysis of Tokamak Experiments and Reactors,” Fus. Technol., 21, 1749 (1992).

[17] R. A. Krakowski, “Simplified Fusion Power Plant Costing: A General Prognosis and Call for ‘New Think‘,’’ Fus. Technol. 27(3) 135 (1995).
[18] C. G. Bathke and the ARIES Team, “A Comparison of Steady-State ARIES and Pulsed PULSAR Tokamak
Power Plants,” Fus. Technol., 26, 1163 (1994). [19] F. Najmabadi and the TITAN Team, “The TITAN Reversed-Field-Pinch Fusion Reactor Study,” University of
California at Los Angeles report UCLA-PPG-1200 (1990). [20] G. Grieger, et al., “Modular Stellarator Reactors and Plans for Wendelstein 7-X,” Firs. Technol. 21, 1767 (1992). [21] R. L. Miller, “Advanced Stellarator Power Plants,”
Fus. Technol., 26, 127 (1994). [22] R. L. Hagenson and R. A. Krakowski, “The Spheromak as a Compact Fusion Reactor,” Los Alamos
National Laboratory report LA-10908 (March 1987).
[23] U.S.Napan Physics of Fusion Concepts for D-3He Burning and Advanced Approaches for Economical Fusion
Power Workshop, Monetrey, CA (September 11-14, 1995).
[24] WEC Commission, Energy for Tomorrow’s World - the
Realities, the Real Options, and the Agendafor Achievement,
St. Martin’s Press Inc, New York, NY (1993).
[25] A. K. N. Reddy and J Goldemberg, “Energy for the Developing Countries,” Sci. Amer., p.111, (September
1990). [26] J. P. Holdren and R. K. Pachauri, An Agenda of Science for Environment and Development into the 219, Century, Chap. 4 (Energy), p. 103, Cambridge University
Press (1991).