COMPARISON OF

STEADY-STATE AND PULSED-PLASMA

TOKAMAK POWER PLANTS

F. Najmabadi,

University of California, San Diego

and The ARIES Team

IEA Workshop onTechnological Aspects of Steady State Devices

Max-Planck-Institut Fur PlasmaPhysikGarching, Germany, February 20–23, 1995

Conceptual Tokamak Power Plants Studies

Steady State:

• ARIES-I: First Stability, high bootstrap.

• ARIES-IV: Second Stability, high bootstrap.

� Discharge duration ∼ months

Pulsed-Plasma:

• Pulsar

� Discharge duration ∼ hours

• Analysis showed that using current drive assist to extend thepulse length of a pulsed-plasma power plant does not lead to anyimprovements because pulsed-plasma and steady-state powerplants operate in different physics regime.

There Are Four ARIES Tokamak Plant Designs

• ARIES-I is based on modest extrapolations in physics and ontechnology which has a 5 to 20 year development horizon (oftenby programs outside fusion).

• ARIES-III is an advanced fuel (D–3He) tokamak plant.

• ARIES-II/IV is based on greater extrapolations in physics(e.g., 2nd stability).

Objectives of the PULSAR Study

• Study the feasibility and potential features of a tokamak with apulsed mode of plasma operation as a fusion power plant.

• Identify trade-offs which lead to the optimal regime of operation.

• Identify critical and high-leverage issues unique to a pulsed-plasmatokamak power plant.

• Compare steady-state and pulsed tokamak power plants.

Approach: Build upon the ARIES designs and focus on issuesunique to pulsed-plasma tokamak power plants:PULSAR-I: SiC/He blanket;PULSAR-II: V/Li blanket.

Pulsed vs Steady-State Plasma Operation

• Steady-state tokamaks (i.e., ARIES-I) reduce the current-drivepower by maximizing the bootstrap current fraction. The plasmashould have a high βp:

� β would be low;

� Substantial (∼ 100 MW) recirculating power is still needed.

• Pulsed-plasma tokamaks use “efficient” inductive current drive:

� Constraint on βp is removed and β could be higher;

� Recirculating power for current drive is eliminated.

• Also, pulsed-plasma tokamak operation is “perceived” to be closerto the present plasma physics data base and understanding.

Trade-off Among MHD, Bootstrap, and CurrentDrive Determines Optimum, Steady-State Plasma,Tokamak Power Plant

• For 1st stability, it is not possible to exceed IBS/Ip ∼ 0.7 forconventional profiles.

Operate at high aspect ratio (low current) and raise qo. β is low,requiring high B to achieve reasonable fusion power density whichscales as (βB2)2.

• Alternatively, use the current-drive system to produce favorableplasma current profiles which allow increasing βN at εβp ∼ 1. (2ndStability, reversed-shear mode).

Values of IBS/Ip>∼ 0.9 are possible for several configurations:

� Centrally peaked current with q0 ∼ 2 and βN ∼ 5;� Off-axis peaked current with q′ < 0 and βN ∼ 5;� No second stability found for kink n = 1 mode with no

conducting wall.

Tokamak Physics Regimes

.01

.02

.03

.04

.05

.06

.07

.08

.09

.000.0 0.1 0.2 0.3 0.4 0.70.60.5 1.00.90.8 1.31.21.1 1.4

β A

S

β p A

q* = 2 q* = 3

q* = 4

q* = 5

q* = 6FS

PU

RS

SS

The Operating Space of Tokamak Power Plants

1st Stability 2nd StabilityRegime Regime

Modest β Modest to high βModest βp High βp

Steady StateARIES-I ARIES-II to -IV

A ∼ 4.5, R ∼ 7 m A ∼ 4, R ∼ 6 m

High β, low βpor

Modest β, Modest βp Not PossiblePulsed

PULSARA ∼ 4, R ∼ 9 m

Internal of ARIES-IV Tokamak Fusion Core

Engineering Features of ARIES-I and ARIES-IV

• ARIES-I and -IV blankets are He-cooled design with SiC/SiCcomposite structural material, Li2O solid tritium breeder, and Beneutron multiplier.

• SiC bulk material and B4C are used as the shield material.

• An advanced Rankine power conversion cycle as proposed forfuture coal-burning plants (49% gross efficiency).

• A high level of safety assurance is projected for both ARIES-I andARIES-IV.

Internals of an FPC module

Engineering Features of ARIES-II

• ARIES-II blanket is made of vanadium alloy (V-5Cr-5Ti)structural material with liquid lithium as the coolant and tritiumbreeder.

• ARIES-II blanket utilizes an insulating layer (TiN) which willreduce the MHD pressure drop by a factor of 20 (< 1 MPa).

• Because of the low pressure drop, the blanket design has beenoptimized toward heat transfer and simplicity.

• Because of high coolant outlet temperature, a gross thermalefficiency of ∼45% is estimated.

Potential Attractive Safety and EnvironmentalFeatures of Fusion

Waste Disposal:

Fusion power plant waste can qualify for Class-C shallow-land burialby using low-activation structural materials (e.g. ferritic steels,vanadium alloys, SiC-composites).

Accidental Release of Inventory:

By proper choice of material, the radioactivity inventory of a fusionplant can be reduced substantially. Through care in design, theimpact of accidents is small.

Radioactivity Levels in Fusion Power Plants

• Low-activation materials reduce the activity level by6 orders of magnitude.

Principal Conclusions of ARIES Study –Blanket Engineering

• SiC composite is an attractive structural material for fusionapplication, but many developmental issues are to be resolved.

• The safety characteristics of fusion power plants can becompromised by small components (i.e., coating of the divertors)

• Even with Li2O as the breeding material, tritium self-sufficiencycannot be assured without Be in the blanket.

• The tritium inventory in high temperature Li2O as well as themaximum operating temperature of Li2O are uncertain.

Principal Conclusions of ARIES Study –Blanket Engineering

• An insulating coating is required for self-cooled liquid lithiumdesigns. The development and demonstration of the reliability ofsuch an insulating coating in fusion environment remains.

• The demonstration of safe operation of a large liquid metal systemin fusion environment is necessary.

• The recovery of bred tritium from lithium to a concentration of∼1 ppm is a critical issue.

The PULSAR-I Fusion Power Core

The PULSAR Operation Cycle

Major Parameters of PULSAR Power Plant

Current ramp time, τr 54 s

Plasma ignition time, τi 53 s

Plasma de-ignition time, τo 38 s

Plasma shutdown time, τq 54 s

Total: Dwell time, τd 200 s

Burn time, τb 9,000 s

Number of cycles 2,700 /year

Pulsed vs Steady-State Plasma Operation

• Pulsed-plasma mode of operation, however, requires many criticalissues to be resolved such as:

� Large and expensive power supplies for PF system;

� Thermal energy storage;

� Magnet design (cyclic fatigue, larger PF coils, rapid PF ramprates);

� Fatigue in the first wall, blanket, shield, and divertor;

� Reliability of complex components under cyclic operation.

Power Flow in a Pulsed Tokamak

• Utilities require a minimum electric output for the plant to stay onthe grid;

• Grid requires a slow rate of change in introducing electric powerinto the grid.

• Large thermal power equipment such as pumps and heatexchangers cannot operate in a pulsed mode. In particular, therate of change of temperature in the steam generator is ∼ 2◦C/minin order to avoid boiling instability and induced stress.

=⇒ Therefore, Steady Electric Output Is Required AndAn Energy Storage System Is Needed.

PULSAR Energy Storage System

• An external energy storage system which uses the thermal inertiais inherently very large:

� During the burn, Tcoolant > Tstorage

� During the dwell, Tcoolant < Tstorage

� But coolant temperature should not vary much. Therefore,thermal storage system should be very large.

• PULSAR uses the outboard shield as the energy storage systemand uses direct nuclear heating during the burn to store energy inthe shield;

� This leads to a low cost energy storage system but the dwelltime would be limited to a few 100’s of seconds.

The PULSAR Thermal EnergyStorage System

Energy Accumulated in Outer ShieldDuring Burn Phase

SHIELD BLANKET PLASMAPRIMARY

HX

Thermal Power Is Regulated by Mass FlowControl During Dwell Phase

SHIELD BLANKET PLASMAPRIMARY

HX

The PULSAR Operation Cycle

• Plasma physics sets a “lower” limit for the dwell time.

• Because of the large step change in the cost of thermal storagesystem, the dwell time is basically set by the thermal storagesystem.Pulsar Dwell Time = 200 s

• The burn time is determined through trade-offs in the magnetsystem:� Shorter burn time increase the number of cycles and lowers the

allowable stresses;� Longer burn time requires larger PF coils (higher volt-seconds)

and larger out-of-plane loads on the TF coils.

• The cost of electricity is insensitive to burn time between 1 to 4hours.Pulsar Burn Time = 9,000 s

Fatigue in First Wall and Blanket

V alloys:

• Fatigue (at 40,000 cycles) reduces the design stress by ∼20%.

• No thermal fatigue data is available.

• No irradiated fatigue data is available.

SiC Composites:

• No fatigue data.

• Data for SiC fiber/Si3N4 matrix indicate that fatigue reduced theallowable stress in the material by 65%.

Fatigue in First Wall and Blanket

• The limited fatigue data for V-alloy structural material suggeststhat ARIES-II type designs can be utilized for PULSAR-IIdivertor, first wall, blanket, and shield.

• Assuming that SiC fiber/Si3N4 matrix data applies to SiCcomposite, ARIES-IV type designs can be utilized PULSAR-I firstwall, blanket, and shield. The ARIES-IV type divertor design maynot be feasible for PULSAR-I.

• Much More Engineering Data Is Needed In Order ToMake A Sound Assessment Of The Impact Of Fatigue OnThe Design Of The Pulsed Tokamak Power Plants.

Maintenace of The PULSAR-I Fusion Power CoreRemove Vaccum Vessel Access Port

Maintenace of The PULSAR-I Fusion Power CoreRemove Fusion Core Sector

Principal Conclusions of ARIES Study –Blanket Engineering

• The advanced physics of ARIES-II/-IV did not lead to a majorreduction in extrapolation from present technology data base.

• A large gap exists between present data base and needs of a fusionpower plant. The technology program is in an early scientific stageand requires considerable effort even to enter an engineering phase.

• An extensive fusion technology and materials R&D effort mustbegin NOW in order for these technologies to be tested at nearlyfull scale in ITER and be ready for the fusion DEMO. Intenseneutron source is not sufficient to provide this data base.

• The material development program should emphasize materialsthat also have other applications (e.g., aerospace and automotiveindustries). This will ensure that these materials are developed atreasonable costs.

Principal Conclusions of the PULSAR Study

• Both steady-state and pulsed power plants tend to optimize atlarger aspect ratio and low current.

• Even though the plasma β is larger in a pulsed tokamak, the fusionpower density (wall loading, etc) would be lower because for thesame magnet, the achievable maximum field at coil would be lower(due to lower allowable stresses n the coils).

• A major innovation of the PULSAR study is the low-cost thermalstorage system using the outboard shield.

• Much more engineering data base is needed to assess the impact offatigue on the design of the blanket and shield of a pulsed-plasmapower plant.

• The magnet system and the fusion power core are much morecomplex in a pulsed-plasma tokamak.

• Assuming the same availability and unit cost for components,PULSAR is ∼25% more expensive than a comparable ARIES-I-class steady-state-plasma power plant.