Startup and load-following transients of a thermoelectric space reactor system with no single point...

American Institute of Aeronautics and Astronautics

1

Startup and Load-Following Transients of a Thermoelectric Space Reactor System with No Single Point Failure

Mohamed S. El-Genk1 and Jean-Michel P. Tournier2 Institute for Space and Nuclear Power Studies and Chemical and Nuclear Engineering Department

The University of New Mexico, Albuquerque, NM 87131, USA

A space Nuclear Reactor Power System (NRPS) has been developed for avoidance of a single point failure in reactor cooling and energy conversion. The Sectored Compact Reactor (SCoRe) in this system is lithium-cooled and the reactor core is divided into six equal sectors with liquid metal heat pipes dividers. These sectors are neutronically, but not thermal-hydraulically, coupled. Each sector has its own primary and secondary circulating lithium loops, which are thermally coupled both in a SiGe Power Conversion Assembly (PCA) and in a Thermoelectric Conversion Assembly (TAC) that powers the electromagnetic pumps in the primary and secondary loops. Each secondary loop also has a separate, segmented radiator panel that is optimized for low specific mass and low liquid lithium inventory. The primary loops transport the thermal power generated in the reactor by fission in 1026 UN fuel pins (or 171 fuel pins per sector) to 6 PCAs that nominally supply a total of 112 kWe to the load at 450 VDC. A Dynamic simulation Model (DynMo-TE) has been developed and used to investigate the transient operation of the SCoRe-TE NRPS during startup from a fully-thawed condition initially at 600 K, to nominal steady-state operation at which the lithium coolant exits the reactor at only 1179 K. Also investigated is the response of the SCoRe-TE NRPS, following a 50% change in the electrical load.

I. Introduction space Nuclear Reactor Power Systems (NRPSs) are enabling to future NASA space exploration missions with power requirements of 10’s to 100’s kWe for up to 15 year, or even longer. These systems could be used for

surface power on the moon and Mars and to provide electrical power for ion propulsion to farthest planets such as Jupiter and its icy moons, Saturn, and Pluto. High power ion propulsion cuts the travel time by more than 50%, and the higher power of NRPSs could also support a host of science experiments and very high rate of data storage and transmission not possible with Radioisotope Power Systems (RPSs). RPSs have operated with remarkable success on many spacecrafts to various planets, but can only provide 10’s to 100’s We. A number of considerations in the design and development of NRPSs include: (a) Nuclear reactor type: Fast spectrum reactors are compact but require significantly more fissile fuel loading than

thermal spectrum reactors. Furthermore, designing the former to remain sub-critical when submersed in wet sand and flooded with seawater, following a launch abort accident is more challenging. However, a proper selection and placement of Spectrum Shift Absorber (SSA) materials (e.g. europium and gadolinium) and using them in the form of a thin coating on the reactor vessel and as small additives to the fuel could alleviate re-criticality concerns for both reactor types (Hawley 1967, Poston 2002, King and El-Genk 2005).

(b) Reactor operation temperature (900 K to 1500 K): Low reactor temperatures (< 1200 K) are favored for using readily available structural materials and significantly lower fuel swelling and fission gas release. Fuel swelling typically increases with fuel temperature raised to the third power but with fuel burnup raised to 0.82 (Ross and El-Genk 1990). On the other hand, low reactor temperature also means low hot side temperature for energy conversion and hence, low conversion efficiency. Consequently, Closed Brayton Cycle (CBC) engines are a likely choice for low temperature gas-cooled reactor because of their relative high conversion efficiency compared to other dynamic and static options, with the exception of AMTEC. At the same hot side temperature of ~ 1150 K, sodium-AMTEC could have a conversion efficiency > 25%, while rejecting waste heat at a

1 Regents’ Professor and ISNPS Director, AIAA Senior Member, 1991-1999, and Associate Fellow since 1999. 2 Research Assistant Professor, AIAA Member since 1998.

S


2

radiator temperature several hundred degrees higher (~ 700 K) (Tournier and El-Genk 2005) than with CBC engines (~350-450 K). As a result, the radiator size of a NRPS with AMTEC would be significantly lower than one with CBC engines.

(c) Reactor cooling: high power space reactors could be cooled using liquid metal heat pipes, circulating liquid metal, or circulating gas. The coolant choice influences the selections of structural, fuel, and fuel cladding materials, and the energy conversion technology. Liquid metal heat pipes provide inherent redundancy with no external pumping but complicate the coupling to the energy conversion subsystem (Tournier and El-Genk 2004). Liquid metal cooling requires ElectroMagnetic (EM) pumps for circulation, but no pressurization of the reactor because of the high boiling points of liquid metals such as NaK, sodium and lithium. Owing to the poor heat transfer characteristics and high pumping requirements of inert gases for cooling space reactors, the system is pressurized to ~ 0.3 - 1.0 MPa, which increases the mass and requires careful consideration of redundancy and avoidance of a single point failure in reactor cooling. On the other hand, corrosion issues with inert gases are benign compared to those with liquid metals and they avoid issues of freeze and thaw of liquid metals.

(d) Energy Conversion: Reactors cooled with either liquid metal heat pipes or circulating liquid metals are most appropriately use static energy conversion such as Thermoelectric (TE), Thermionic (TI), Thermo-photovoltaic (TPV), or Alkali Metals Thermal-to-Electric Conversion (AMTEC) (Refs). These options typically operate are high hot side temperatures ranging from 1100 – 1900 K, and some has a low cold side temperature, increasing the radiator area. Dynamic options of FPSEs and Potassium Rankine Cycle (PRC) are also suitable to use with liquid metal heat pipes or circulating liquid metal cooled reactors. CBC engines, however, are more suitable to use with gas-cooled reactors. Dynamic conversion options typically have high conversion efficiencies than static options, except AMTECs with > 20-25% efficiency, but some like CBC and FPSEs have low cold side temperature, increasing the radiator size. For the same radiator effective surface area, a recent analysis has shown that NRPSs with either CBC or SiGe-TE conversion generate almost the same, but lowest electrical powers. The best technologies in terms of generating the highest electrical power are PRC, followed closely by AMTEC, and then Advanced TE of skutterudite-based materials a distance third (Tournier and El-Genk, 2005).

(e) Launch vehicle: The size, stowed volume, and total mass of a NRPS could also affect by the launch vehicle. The radiator is typically the heaviest and largest component of a NRPS. Therefore, early considerations of the design and optimization of the radiator and its integration with NRPS and packaging into the bay of the launch vehicle, such as the DELTA–IV Heavy or other similar rockets, are necessary.

(f) Safety during launch, ease of startup and ability to load follow. With static energy conversion and nuclear reactors with a negative temperature reactivity feedback, NRPS are inherently load following. Such operation advantage is NOT possible with dynamic conversion. For a NRPS with a gas-cooled reactor that is directly coupled to CBC engines, startup from a frozen state is not an issue. Startup of liquid metal heat pipes from a frozen condition had been shown not to be a concern as with circulating liquid metal cooled NRPS. For the latter, however, heating up the system on the launch pad then covering it with a thermal blanket so the liquid metal coolant remains above its freezing point when starting SNRPS in orbit has been used successfully with SNAP-10A in 1965 and the Russian TOPAZ-I in 1989.

(g) Operation reliability: It is best achieved through a combination of redundancy in energy conversion and avoidance of a single point failure. Space reactors cooled with circulating liquid metals and or gasses such as He-Xe, or He, could be designed for avoidance of a single point failure (Liscum-Powell and El-Genk, 1994a,b; El-Genk, Liscum-Powell and Pelaccio, 1994; El-Genk et al., 2005). An important issue is understating the performance and transient response of NRPSs during startup and following a large change in electrical load.

Transient operation of NRPS with lithium cooled, Sectored, Compact Reactor (SCoRe) designed for avoidance of single point failure (Refs) and SiGe, Thermoelectric (TE) Power Conversion Assemblies (PCAs) between the lithium primacy and secondary loops is investigated. The simulated transients include the startup of the SCoRe-TE NRPS from a fully thawed condition initially at 600 K to steady state nominal operation at 115 kWe and 450 VDC, with the least amount storage batteries power to operate the EM pumps in the secondary loops early in the startup transient. Also investigated is the transient response of this system following a 50% decrease in the load electrical power followed by an equal increase to the nominal steady state power level. At the nominal steady state reactor exit temperature < 1180 K, it possible to use Oxides Dispersed Steels (MA-ODS) or super steel alloys

II. Dynamic Simulation of SRPSs Investigating transient operation of NRPSs requires comprehensive simulation capabilities to develop actual

startup procedures for NRPS without spiking the reactor fission power and temperature and for adaptive control (Metzger and El-Genk paper) and autonomous operation (Metzger and El-Genk, 1992; Metzger, El-Genk, and


3

Parlos, 1991). Dynamic simulation capabilities of different types of NRPSs have been developed at the University of New Mexico’s Institute for Space and Nuclear Power Studies (ISNPS). These include Space Nuclear Power Systems Analysis Model (SNPSAM) of the SP-100 NRPS (Truscello and Rutger, 1992) with SiGe TE conversion and a fast spectrum reactor cooled with circulating liquid lithium (El-Genk and Seo, 1987a, 1987b, 1988; El-Genk, Buksa and Seo, 1987; El-Genk, Seo and Buksa, 1987; Rider, 1989; El-Genk and Rider, 1989; El-Genk and Xue, 1992).

A dynamic simulation model has also been developed at ISNPS of the Russian TOPAZ-II that was successfully benchmarked using actual system performance data (El-Genk, Xue, and Paramonov, 1994; El-Genk and Xue, 1994; El-Genk and Paramonov, 1994; Paramonov and El-Genk, 1994, 1996 and 1997). TOPAZ-II employed a thermal spectrum reactor cooled with circulating liquid NaK, and in-core Thermionic Fuel Elements (TFEs) with UO2 fuel for energy conversion and nominally generated ~ 5.5 kWe at a load voltage of ~ 28 VDC. Recently, a Dynamic Model (DynMo-TE) for NRPSs with TE conversion has been developed (El-Genk and Tournier 2005) using the SIMULINK® platform (Simulink, 2004); an interactive graphical environment that allows rapid development of a library of blocks of the system components that can easily be replaced or exchanged. Through MATLAB® (Matlab, 2004), SIMULINK® has immediate access to an extensive range of tools for numerical computation, time integration, algorithm development, and data visualization and analysis. SIMULINK platform has also been used to perform dynamic simulation of a NRPS with CBC engines (Wright and Sanchez 2005).

DynMo-TE has been used to investigate the startup transient and the effect of using different combinations of alkali metal coolants (Li, Na, and NAK) in the primary and secondary loops of SCoRe-TE NRPS in Fig. 1. Results showed that SCoRe-TE could be started up in orbit from a fully-thawed condition 10’5 using auxiliary batteries to power the EM conduction pumps in the primary and secondary loops, until the voltage developed across the pumps’ TCAs is enough to power the pumps, and disconnect the batteries (El-Genk and Tournier 2005). Results also showed with using lithium coolant in the primary and secondary loops results in the highest system efficiency and electrical power output at the lowest reactor exit temperature,

6.43 m1.5

1.08 m

4.52 m

6 fixed,forwardpanels

5.77 m

12 rear,deployablepanels

7.61 m

SCoRenuclearreactor

NuclearRadiationshield

Transmission line,power conditioningand distribution

Electric propulsionand science payload

Mechanicalboom

Heat rejection radiator

6.43 m1.5

1.08 m

4.52 m

6 fixed,forwardpanels

5.77 m

12 rear,deployablepanels

7.61 m

SCoRenuclearreactor

NuclearRadiationshield

Transmission line,power conditioningand distribution

Electric propulsionand science payload

Mechanicalboom

Heat rejection radiator

Figure 1. A Layout of SCoRe-TE NRPS.

(b) Aft (fixed) radiator segment (1 of 6)

(c) Rear (deployable) radiator segment (1 of 12)

5.952 m

0.59m

1.00m

Radiatorheat pipes

C-C fin

Inlet duct

Outlet duct

Perforated separation wall

6.605 m

0.268m

1.13m

Radiatorheat pipes

C-C fin

Inlet duct


Outlet duct

(a) Layout of segmented radiator panel (design – B) (1 of 6)

Inlet ductto panel

Outlet ductfrom panel

One fixed (aft) segment

Header

Collector

Two deployable(rear) segments



5.952 m

0.59m

1.00m

Radiatorheat pipes

C-C fin

Inlet duct

Outlet duct


6.605 m

0.268m

1.13m

Radiatorheat pipes

C-C fin

Inlet duct


Outlet duct



5.952 m

0.59m

1.00m

Radiatorheat pipes

C-C fin

Inlet duct

Outlet duct


5.952 m

0.59m

1.00m

Radiatorheat pipes

C-C fin

Inlet duct

Outlet duct


6.605 m

0.268m

1.13m

Radiatorheat pipes

C-C fin

Inlet duct


Outlet duct

6.605 m

0.268m

1.13m

Radiatorheat pipes

C-C fin

Inlet duct


Outlet duct

(a) Layout of segmented radiator panel (design – B) (1 of 6)

Inlet ductto panel



Header

Collector


Inlet ductto panel



Header

Collector


Figure 2. A segmented radiator panel.


4

which are attributed to the high heat capacity and thermal conductivity of lithium. This system simulation, however, did not account for the presence of liquid lithium accumulators in the primary and secondary loops, thus the lithium coolant pressure in these loop was not regulated during the startup transient.

In the present investigation, accumulator models are developed and incorporated into DynMO-TE. The accumulators are optimized for structural strength and minimum pressurization of SNPS at nominal steady state operation. Thus, it is currently possible, to account for the effect of changing the reactor power on the pressure of liquid lithium in the primary and secondary coolant loops and for the changes in the volume of lithium in the loops. The total volume of lithium and of the accumulators in the secondary loops were reduced based on a detailed design optimization and analysis of the radiator panels for limiting the pressure losses in each panel to 12 KPa and decreasing the volume of the coolant channels in the panels (Tournier and El-Genk, 2005). SCoRe-TE NRPS described in the next section (Fig. 1) employs 6 segmented radiator panels and the radiator’s effective surface area is 203 m2 .

III. SCoRe-TE NRPS The nuclear reactor, SCoRe, is divided into six equal sectors and each has its own primary coolant and secondary coolant loops and each secondary loop a separate radiator panel. This arrangement provides redundancy in reactor cooling for avoidance of a single point failure. Each radiator panel consists of a forward, fixed part and 2 rear deployable parts (Figs. 2 and3) The three parts are coupled hydraulically in parallel, which together with using perforated dividers between the inlet and exit lithium coolant channels in the panel decreased the lithium inventories in the secondary loops and the radiator of SCoRe-TE NRPS. This radiator panel design also increased the specific power of the radiator (Figs. 3). The segmented radiator (Fig. 3) has effective and geometrical surface areas of 203 m2 and 168.9 m2, respectively, holds only 177 liters (86 kg) of lithium and had a specific mass of 6.46 kg/m2

(Tournier and El-Genk, 2005). It weights a total of 1092 kg, including 86 kg of lithium in the panels.

A. Deployment Sequence Line diagrams of the different parts of the

segmented radiator panel in SCoRe-TE NRPS, with some dimensions are shown in Fig. 2. The stowed configuration in the launch pay of the DELTA IV Heavy and the deployments sequence of SCoRe-TE NRPS is shown Fig. 3. In the stowed configuration, six of the rear deployable parts of the radiator panels are folded over the six forward fixed parts. The other 6 rear deployable parts when folded overlap the other 6 folded parts with a 30o angle (Fig. 3 sequence 1). The length of the forward fixed parts of the radiator is limited by the diameter of the payload fairing in thelaunch vehicle. When deployed, the effective radiation view factor for the inside surface of the fixed parts of the panels is only 1.06 because they are partially obstructed by the rear deployable parts. The extent (5.77 m) of the deployed rear parts of the radiator is limited by the overlap in the stowed configuration, hence dictating their major diameter of 7.61 m. When fully deployed the radiator panels remain within the cone protected by the radiation shadow shield (Fig. 1), which has a cone angle of 30o. The effective view factor for rear, deployed parts is 1.273.

IV. Sectored Compact Reactor (SCoRe) SCoRe has a hexagonal core surrounded by a relatively thick (> 10 cm) BeO radial and BeO axial (~ 4 cm thick)

reflectors. A total of 6 or 12 BeO rotating control drums with thin (> 5 mm) B4C segments are inserted in the radial reflector. In the shutdown mode, the B4C segments face the reactor core and face away at the end of operation life of the reactor. SCoRe core is divided into six equal sectors loaded with UN fuel pins clad in Mo-14Re. The

(1) Stowed configuration

(2) Deployment of payload and 6 outer panels

(3) Deployment of6 inner panels

(4)

(5) Fully-deployedspace power system




(4)





(4)





(4)


Figure 3. Deployment sequence of SCoRe-TE NRPS.


5

cladding has a wire wrap of the same material on the outside surface to maintain uniform flow channels and provide structural integrity of the assembled reactor core (Fig. 4a). Each sector has separate inlet and exit and a separate lithium primary loop. This loop is thermally coupled to a secondary lithium loop and has a separate rubidium heat pipes radiator panel for waste heat rejection. The returning lithium to a core sector flows upward in an annulus on the inside of the reactor vessel wall, then reverses direction at the opposite end to flow through the sector and remove heat generated by fission in UN fuel pins. From the lower plenum the liquid lithium exits the reactor core (Fig. 4b). The inlet annuli and exit plenums of the six sectors in SCoRe are physically separated using liquid-metal heat pipe dividers. These dididers passively cool the reactor sector experiencing a loss of coolant. In such case, the dividers transport the fission power generated in that sector to the circulating lithium in the adjacent sectors (El-Genk et al., 2005). Figure 4a shows a radial cross-section of SCoRe-S11 used in the present investigation. Each sector in SCoRe-S11 is loaded with 171 fuel pins for a total of 1026 pins in the core, which nominally generate up to 4 MWth for 7 – 10 years.

Coolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletRx Vessel DomeSectors Divider

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Rx Vessel Dome

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Liquid Metal HeatPipe Dividers

BeO RadialReflector

BeO / B4CControl Drums

Rx Core withSingle UN Fuel Pins

Rx Vessel DomeSectors Divider

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Coolant Exit Coolant Inlet

Mo-14ReVessel

Coolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletCoolant Exit Coolant InletRx Vessel DomeSectors Divider

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Rx Vessel Dome

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Rx Vessel Dome

Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus

Liquid Metal HeatPipe Dividers

BeO RadialReflector

BeO / B4CControl Drums

Rx Core withSingle UN Fuel Pins


Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus



Coolant Exit Plenum

LM Heat Pipes

Coolant Annulus


Mo-14ReVessel

(a) Radial cross-section of SCoRe-S11 core. (b) Coolant flow in SCoRe-S11 core.

Figure 4. Radial and axial cross-sections of SCoRe-S11 core (not to scale).

V. DynMo-TE As indicated earlier, each primary and secondary lithium loop in SCoRe-TE has an EM pump (Fig. 5). The

primary and secondary loops are thermally coupled in a SiGe PCA and in the pumps’ TCA. Each TAC provides electrical power for a pair of EM pumps in the secondary and primary loops (Fig. 5). The secondary loop transports the rejected heat from a PCA and a pumps TCA to a rubidium heat pipes radiator panel. The 6 primary loops transport the thermal power generated in the reactor sectors to the six PCAs and six pumps TCAs. The. PCAs have built-in redundancy by using multiple parallel strings of SiGe Thermoelectric Converter Assemblies (TCAs) to provide a nominal load voltage of 450 VDC at nominal steady state operation. The SiGe unicouples in the pumps TCAs are connected electrically in parallel to provide the highest electrical current to the pumps. Each half TCA provide power to a primary loop or a secondary loop EM Pump. The nominal electric current supplied to each pump is typically a few thousand amperes at a few hundred mV.

DynMo-TE (Fig. 5) is comprised of a number of coupled physical models of, namely: (a) reactor thermal-hydraulics and kinetics, (b) primary and secondary lithium loops, (c) accumulators in the primary and secondary loops, (d) PCAs and pumps TCAs, and (e) the segmented radiator panels. These models are detailed next. (a) SCoRe-S11 Model: It couples a six-points kinetics model of the reactor to the reactor thermal-hydraulic model

(Fig. 5). The former calculates the reactor fission power subject to the external reactivity insertion ar a user specified rate and the temperature reactivity feedback for liquid lithium and UN fuel, and the Doppler reactivity feedback, when applicable. These reactivity feedbacks are functions of the transient changes in the fuel and lithium temperatures and in the fission power in the reactor. The changes in the fuel and coolant temperatures as well as in other core structures depend on the lithium coolant flow rate and the rise in its temperature in the reactor. Both are determined from solving the coupled momentum and energy balance equations in a pair of primary and secondary coolant loops (Fig. 5).


6

(b) Secondary and Primary Loops Models: These models are coupled thermally in the PCAs the pumps TCAs (Fig. 5). Each pump is powered by a half-TCA. A PCA and the pumps TCA are mounted between the secondary and primary loops. The constitutive equations in the models of the primary and secondary loops are those of the overall momentum and energy balance. The models accounts for the transient changes in the lithium pressure, thermal inertia of circulating liquid lithium of the piping in the secondary and primary loops and the rubidium heat pipes radiator, the energy balance equation accounts for the rate of rejected thermal power and removed by the circulating lithium in secondary loops for the PCAs and the pumps TCAs. These models also account for the rate of heat rejection by the radiator panels and the parasitic heat losses, the transient changes in the thermal inertia and pressure of the lithium in the primary loop, the thermal inertia of the structure, fuel, and coolant in the reactor, and the PCA and pumps TCA. The overall energy balance in the primary loops accounts for the transient change in the reactor fission power, rate of heat transfer to the secondary loops in the PCAs and pumps TCAs, thermal and electrical parasitic losses, and the lithium coolant pressure and thermophysical properties in the primary and secondary coolants loops. The transient flow rates lithium in the primary and secondary loops are calculated from solving the coupled energy and momentum

balance equations in these loops at the points where the determined pressure loss -flow rate demand curves intersect the calculated pressure head-discharge characteristic curve of the EM pumps.

(c) EM Pump Model: This model calculates the pressure head-discharge rate characteristic as a function of temperature and time during a transient, the dimensions of the flow ducts in the pumps, the thermal and electrical resistivities of the wall materials and liquid lithium coolant, and the supplied DC voltage and current by the pumps TCA (El-Genk, Buksa and Seo, 1987). The EM pump use a permanent magnet that is thermally insulated from the lithium coolant ducts and maintained well below the Curie point of the magnet material. The magnet remains saturated by the secondary magnetic flux generated by the electric current passing through the pump ducts. Such magnetic field also generates an opposing electric potential that is overcome by the electric potential supplied by the pumps TCA.

(d) Radiator Model: The radiator model discretizes the inlet and outlet lithium flow channels in the radiator panels into small axial sections. Each section provides thermal energy by convection from the flowing lithium to a number of rubidium heat pipes with C-C armor and C-C fins. In each section of the channels, the radiator model solves the coupled momentum and energy balance equations for the flowing lithium and calculates the decrease in its temperature in the inlet and exit channels. It also calculates the temperature drops in the channels walls and in the structure of the evaporator section of the heat pipes. The radiator model is coupled to a heat pipe model, which calculates the vapor flow and temperature drop in the rubidium heat pipes and in the condenser structure to the base of the C-C fins. The heat pipe model also calculates the heat pipes sonic,

Magnet thermal model

Primary H

X m

odel

TE model

Secondary HX

model

EM pump model

Primary coolant loop model Secondary coolant loop model

Primary EM pump model

PCA

model

Pumps TC

A m

odel

Magnet thermal model

EM pump model

Secondary EM pump model

Reactivityinsertion

Feedbackreactivity

Reactorcontrol

SCoRe-S11thermal-

hydraulics model

Reactor kineticsmodel

Reactor model

Segmented heat pipes

radiator modelPrim

ary HX

model

TE model

Secondary HX

model

Accumulatormodel

Electric load

Accumulatormodel

+ -

+ -+ -

Magnet thermal modelMagnet thermal modelMagnet thermal model

Primary H

X m

odelPrim

ary HX

model

TE model

TE model

Secondary HX

model

EM pump model

Primary coolant loop model Secondary coolant loop model

Primary EM pump model

PCA

model

Pumps TC

A m

odel

Magnet thermal modelMagnet thermal model

EM pump model

Secondary EM pump model

Reactivityinsertion

Feedbackreactivity

Reactorcontrol

SCoRe-S11thermal-

hydraulics model


Reactor model

Reactivityinsertion

Reactivityinsertion

FeedbackreactivityFeedbackreactivity

ReactorcontrolReactorcontrol

SCoRe-S11thermal-

hydraulics model


Reactor model

Segmented heat pipes

radiator modelPrim

ary HX

model

Primary H

X m

odel

TE model

TE model

Secondary HX

model

Accumulatormodel

Accumulatormodel

Electric load

Accumulatormodel

Accumulatormodel

+ -

+ -+ -

Figure 5. Simulation model of SCoRe-TE NRPS:


7

capillary, entrainment, and incipient boiling limits, to ensure a sufficient nominal operation design margin. The radiator model calculates the effective fin efficiency for rejecting waste heat into space and the local effective radiation view factor. The radiator model calculates the pressure drops in the lithium flow channels and across the orifices in the perforated dividers between the inlet and exit channels, and the change in the coolant flow rate along the channels. More details on the design of the radiator heat pipes are available in another article in these Proceedings (Tournier and El-Genk, 2005).

(e) PCA and Pumps TCA Models: The basic building block in both models is a transient performance model of the SiGe unicouples (El-Genk, Seo and Buksa, 1987). The cross-sectional areas of the n- and p-legs are optimized for maximum efficiency at the nominal steady-state operation conditions of the SCoRe-TE NRPS. In the PCAs, the SiGe unicouples are connected in series in 4 parallel strings to form a TCA, and the TCAs are connected in series to form the PCA and deliver the system’s nominal electrical power of 112 kWe at a nominal load voltage of 450 VDC. For additional redundancy, the six PCAs in SCoRe-TE NRPS are connected in parallel. In order to provide the very high current needed to operate the pumps the optimized SiGe unicouples in the pumps TCAs are connected in parallel in two-series strings. The electrical power generated in the PCAs and the pumps TCAs, and the corresponding terminal voltages and currents are calculated as functions of time during the startup and other operation transients. (f) Accumulator Model: To accommodate the volume changes in the liquid lithium flowing in the secondary and primary loops during a transient operation, each loop is equipped with a bellow type accumulator (Figs. 5 and 6). The accumulator model in DynMO-TE is similar to that reported by El-Genk, Lapin, and Seo (1988). It consists of an outer cylindrical chamber made of 0.6 mm thick Nb-1Zr and thermally insulated. The chamber encloses bellows with 0.254 mm thick Nb-1Zr wall, and a Nb-1Zr compression spring (Fig. 6 The cavity above the bellows is filled with inert gas to support the spring and the bellows in adjusting the

lithium pressure in the loops. In SCoRe-TE NRPS each primary and secondary loops hold 42.5 and 46.5 liters of liquid lithium, respectively. The accumulators in the primary and secondary loops are designed for minimum mass and operating at < 70%vol capacity and < 30 kPa at the average nominal lithium temperatures in these loops of ~1223 K and ~ 800 K, respectively. The optimized primary loop accumulator weights 10.9 kg and is 34.3 cm high and 33.5 cm in diameter (Fig. 6). The secondary loop accumulator weights 5.9 kg and is 28.2 cm high and 26.2 cm in diameter. The accumulator model in DynMo-TE calculates the transient changes in the volumes and pressure of liquid lithium in the loops and the compression length of the metal spring. It also accounts for the changes in the stiffness of the metal spring and bellows and the pressure of the inert gas in the accumulators. The thermal model that is coupled to the hydrodynamic model of the accumulator calculates the average temperature of the accumulator and accounts for the thermal masses of the stored liquid lithium and of the bellows, the metal spring, and the chamber wall in the accumulator.

VI. Startup Simulation The dynamic simulation of the SCoRe-TE NRPS (Fig. 1) is carried out using DynMo-te delineated in Fig. 5. The

output of DynMo-TE includes all system parameters as a function of time during the startup transient. These include: (a) reactor thermal power, (b) load electric power and voltage, (c) the rejected thermal power and surface temperature of the radiator, (d) the lithium flow rates, pressures and temperatures in the primary and secondary loops, (e) the characteristics of EM pumps as function of the terminal voltages and electrical currents supplied by the pumps TCAs, (f) the temperatures in the PCAs and pumps TCAs, (g) the total and feedback reactivities in the reactor core, (i) the rise in and exit temperature of the lithium coolant flowing through the reactor core, and (j) the pressure losses in the primary and secondary loops.

Figure 6. Accumulator at startup, nominal, and maximum capacity positions during startup transient.

Helium gas-Filled cavity

ExpandableNb-1Zrbellows

CompressionSpring

(Nb-1Zr)Nb-1Zr

wall

Liquid coolantFrom loop

∆Xn

Hsp

Hb

Dsp

Db

D

H

Hb

(a) Initial position (b) Nominal operating position (c) Maximum capacity position

Thermal insulationHelium gas-Filled cavity

ExpandableNb-1Zrbellows

CompressionSpring

(Nb-1Zr)Nb-1Zr

wall

Liquid coolantFrom loop

∆Xn

Hsp

Hb

Dsp

Db

D

H

Hb

(a) Initial position (b) Nominal operating position (c) Maximum capacity position

Thermal insulation


8

The startup procedures delineated in Fig. 7 are for the SCoRe-TE NRPS initially at 600 K and fully covered with thermal blanket. Since thermal losses from the insulated system to space are not capable of establishing a temperature difference across the pumps TCAs to circulate the liquid lithium in the primary and secondary loops, auxiliary batteries are used for a short time during the startup transient to power only the EM pumps in the secondary loops. The resulting convective cooling develops a temperature differential across the pumps TCAs but unless the reactor is started to also raise the primary lithium temperature, the developed temperature difference in the plumps TCAs would not be large enough to disconnect the auxiliary batteries and connect the pumps TCAs to power both the primary and secondary EM pumps. The startup procedures simulated in this paper are delineated in Fig. 7 and discussed next.

A. Startup Procedures The present startup simulation assumes

equilibrium space sink temperature of 10 K, and radiator emissivity of 0.9. The initial system temperature of 600 K is ~ 150 K higher than the melting temperature of lithium (454 K). Before launch, and the lithium in primary and secondary loops is thawed on the launch pad to sufficiently high temperature (e.g., ~ 800 K) using electric heaters. The stowed SCoRe-TE NRPS is covered with thermal insulation blanket during launch and until the system is successfully inserted into orbit. This blanket minimizes the thermal losses from the system during launch and ensures that its temperature of does not drop below 600 K. The insulation blanket is also kept on during the first 300 s of the startup procedure, before jettisoned into space (Fig. 7).

During startup in orbit, the electrical load is disconnected and the only secondary EM pumps are initially operated using auxiliary batteries. The electrical load is connected when the system reaches nominal operating temperatures. The TCAs begins to power both the primary and secondary EM pumps when the developed temperatures difference across the pumps TACs is large enough to supply electric currents to the secondary pumps that are at least 1 A higher than those provided by the auxiliary batteries. At Such instance, the auxiliary batteries are disconnected from the secondary EM pumps.

The startup procedures in Fig. 7 last for a total of ~ 8,000 sec (2 hrs and 15 minutes) to reach nominal operating conditions and supply nominal electric power to the load. The sequential actions taken during the startup procedures are indicated in Fig. 7 are:

(a) At startup, SCoRe-TE NRPS is initially at 600 K, the TCAs are connected to the primary EM pumps, and the secondary EM pumps are powered using auxiliary batteries, which provide each secondary EM pump with 4 A at ~1 mV DC.

(b) External reactivity is inserted into the reactor core by rotating the BeO/B4C drums in the radial reflector e (Fig. 2a rotated are an constant angular speed to maintain a reactivity insertion rate of 1.0 cent/s. This phase ends after a total external reactivity insertion of $0.80, and the rotation of the control drum ceases momentarily for about 1920 s, during which the reactor almost reaches steady state low power operation.

(c) At 300 seconds into the startup transient the insulation blanket is jettisoned into space, a process that takes 30 second to complete.

(d) The resulting increase in heat losses from the radiator following action (c) increases the temperature difference across the pumps TCAs to power the pumps (i.e. providing > 5 A to each EM pump) and the auxiliary batteries are disconnected from the secondary EM pumps 20 s after removing the insulation blanket or 350 s from the beginning of the startup procedures. These actions increase the circulation rates of the liquid lithium in the secondary and primary loops and load open circuit voltage. The reactor thermal power and other system parameters reach steady state ~ 2000 s from the initiation of the startup procedures, or 1650 s after the TCAs began powering the EM pumps in the primary and secondary loops (Fig. 7).

Figure 7. SCoRe-TE NRPS Startup procedures.

0

20

40

60

80

100

120

140

160

180

200

0 2000 4000 6000 8000

Resume reactivity insertion (t = 2000 s)

1.0

cent

/ s

0.02 cent /

s

Connect electrical load (t = 7500 s)

End of reactivity insertion(t = 7000 s)

Disconnect batteries and connect pumps TCAs (t ~ 350 s)

Remove radiatorblanket (t = 300 s)

Connect auxiliary batteriesto secondary EM pumps

Time (s)

Ext

erna

l Rea

ctiv

ity In

serti

on (c

ents

)


9

(e) The second phase begins 2000s after the initiation of the startup procedures by resuming the outward rotation of the control drum slower than during the first phase such that to maintain a steady external reactivity insertion rate of 0.02 cent/s. This phase ends 7000 s after initiating the startup transient, when the total reactivity insertion reaches $1.00 (Fig. 7).

(f) After the system reaches steady state, 500 s latter, the electrical load is connected, causing a brief transient (Fig. 7). After an additional 500s, SCoRe-TE NRPS reaches steady state nominal operation.

The next section presents and discusses the calculated changes in the various operation parameters of SCoRe-TE NRPS during the startup procedures delineated in Fig. 7.

B. Changes in Operation Parameters during Startup Transient The reactivity feedback in Fig. 8a is the sum of those of the UN and liquid lithium in SCoRe-S11 core. At the

beginning of the startup procedures the reactor thermal power of 0.01 Wth is produced by the radioactive decay of the fuel in the reactor core (Fig. 8b). Also, the entire NRPS is at a uniform temperature of 600 K (Fig. 8c). The reference temperature for the reactivity feedback in the reactor core is 300 K. After about 1000 s from the initiation of the startup procedures the reactor thermal power increases rapidly and peaks at 1100 kWth before decreasing also rapidly as the reactor temperature increases (Figs. 8b and 8c) and the negative reactivity feedback kicks in. The reactor thermal power reaches ~ 300 kWth (point 3 in Fig. 8b) when the temperature of the liquid lithium exiting the reactor core is ~ 640 K (point 3 in Fig. 8c). This happens 2,000 s after the initiation of the startup transient (Fig. 7). At such time, the total reactivity is nearly zero, as the external and feedback reactivity become equal in magnitudes, but with opposite signs (point 3 in Fig. 8a); at such time the reactor thermal power almost equal that rejected by the radiator, and the open circuit voltage of the PCAs is ~ 80 VDC (point 3 in Fig. 8d), but the load electric power is zero because the load is disconnected. This point ends the first phase of the startup procedure delineated in Fig. 7.

Figure 8b shows the rejected thermal power by the radiator increasing rapidly, immediately after the insulation blanket is removed (point 1 in Fig. 8b), to a peak of ~ 1100 kWth (point 2 in Fig. 8b). It then decreases precipitously to its lowest value of ~ 200 kWth, before increasing again as the reactor thermal power begin rising (Fig. 8b)..

-2.0

-1.5

-1.0

-0.5

0

0.5

1.0

1.5

2.0

0 2000 4000 6000 8000

(5)

(4)

(3)

(1)

(a)

Feedback reactivity

Total reactivity

External reactivity

Rea

ctiv

ity ($

)

500

600

700

800

900

1000

1100

1200

0 2000 4000 6000 8000

5700 s

(5)(4)

(3)(c)(1)(2)

Radiator HPs are sonic-limited

Radiator exitRadiator inlet

Core in

letCore ex

it

Lith

ium

Tem

pera

ture

(K)

0

500

1000

1500

2000

2500

3000

3500

0 2000 4000 6000 8000

(5)

(4)

(3)

(2)

(b)Radiator heat re

jection

Reactor

(1)

(1) Remove radiator blanket(2) Radiator HPs become sonic-limited(3) Resume reactivity insertion(4) End of reactivity insertion(5) Connect load

Time (s)

Ther

mal

Pow

er (k

W)

0

100

200

300

400

500

600

700

800

900

0 2000 4000 6000 8000100

110

120

130

140

150

160

170

180

190(5)(4)

(3)(1)

(d)ClosedcircuitOpen-circuit

Time (s)

Load

Vol

tage

(V)

Ele

ctric

al P

ower

(kW

e)

-2.0

-1.5

-1.0

-0.5

0

0.5

1.0

1.5

2.0

0 2000 4000 6000 8000

(5)

(4)

(3)

(1)

(a)

Feedback reactivity

Total reactivity

External reactivity

Rea

ctiv

ity ($

)

500

600

700

800

900

1000

1100

1200

0 2000 4000 6000 8000

5700 s

(5)(4)

(3)(c)(1)(2)

Radiator HPs are sonic-limited

Radiator exitRadiator inlet

Core in

letCore ex

it

Lith

ium

Tem

pera

ture

(K)

0

500

1000

1500

2000

2500

3000

3500

0 2000 4000 6000 8000

(5)

(4)

(3)

(2)

(b)Radiator heat re

jection

Reactor

(1)


Time (s)

Ther

mal

Pow

er (k

W)

0

100

200

300

400

500

600

700

800

900

0 2000 4000 6000 8000100

110

120

130

140

150

160

170

180

190(5)(4)

(3)(1)

(d)ClosedcircuitOpen-circuit

Time (s)

Load

Vol

tage

(V)

Ele

ctric

al P

ower

(kW

e)

Figure 8. SCoRe-TE NRPS performance during startup procedures.


10

During the second phase of the startup procedures, during which external reactivity is inserted in the reactor core at a steady rate of 0.02 cent/s, both the reactor thermal power and the rejected power by the radiator increase linearly (Fig. 8b). The difference between their values represents the electrical power supplied to the EM pumps by the TCAs, the parasitic electrical losses in the PCAs and TCAs, and as the thermal losses from the piping and structure of the system. At the end of this phase of the startup procedures the total external reactivity insertion is $1.00 and the reactor thermal power peaks at 2.86 MWth (point 4 in fig. 8b). Also, the temperatures of liquid lithium in the primary loop at the reactor exit and in the secondary loop at the inlet to the radiator reaches ~ 1180 K and ~ 770 K, respectively (point 4 in Fig. 8c), and the open circuit voltage of the NRPS is ~ 870 VDC. After an additional 500 s from the end of the external reactivity insertion, the external load is connected (point 5 in Fig. 8d) causing the load voltage to precipitously decrease to 450 VDC and the load electrical power to increase almost instantaneously to a peak of ~ 130 kWe, then dropping and experiencing minor transients before reaching a nominal steady state at 112 kWe at 450 VDC (Fig. 8d). At such condition, the nominal steady state thermal power of the nuclear reactor is 2.44 MWth and the temperature of liquid lithium exiting the reactor core is 1179 K.

B. Changes in Liquid Lithium Pressure and Temperatures Figures 9a and 9b presents the changes in the pressures and subcoolings of the liquid lithium at selected points in

the primary and secondary coolant loops of SCoRe-TE during the startup procedures delineated in Fig. 7. The lithium pressures are the lowest at the inlets of the primary and secondary EM pumps (Fig. 9a). The increase in the pressure in the primary and secondary EM pumps equals the pressure losses in the primary and secondary loops, respectively (Fig. 9a). As this figure indicates, during the second phase of the startup procedures and at nominal steady state operation condition the pressure of liquid lithium exiting the secondary EM pumps is higher than that of the lithium at the inlet of the primary EM pumps. However, the pressure of the lithium exiting the primary EM pumps is higher than exiting the secondary EM pumps. Figure 9a shows that although the pressures of the lithium exiting the secondary and primary EM pumps and entering the primary pumps increase with time during the startup transient that of the lithium entering the secondary EM pumps decreases, except after connecting the external load it increase some to its steady state value of ~ 14 kPa. Such a decrease in lithium pressure is because of the rise in pressure caused by the increase in the average temperature in the secondary loops is smaller than the increase in the pressure losses in the radiator panel. These losses increase as the temperature difference across the pumps TCAs increases (Fig. 8c) and hence both the electric current supplied to the pumps and flow rate of liquid lithium in the secondary loops.

Figure 9b presents the calculated subcooling of liquid lithium at selected points in the primary and secondary loops. The highest liquid subcoollings are those that at the location of the accumulator and at the inlet of the EM pumps in the primary and secondary loop, respectively. The results confirm that there is sufficient operation margin in the lithium temperature in the primary and secondary loops during the startup and at nominal steady state operation (Fig. 9b). At the latter, the liquid lithium in the primary and secondary loops is ~ 300 K and 600 K, respectively, providing a large margin for an increase in the reactor thermal power as a results of increasing the load demand beyond nominal.

5

10

15

20

25

30

35

40

45

50

0 2000 4000 6000 8000

(a)(5)(4)

(3)(1)

Secondary looplosses

Primary looplosses

Primary

pump e

xit

Secondary pump exit

Primary pump inlet

Secondary pump inlet

Time (s)

Lith

ium

Pre

ssur

e (k

Pa)

200

300

400

500

600

700

800

900

0 2000 4000 6000 8000


(b)(3)

(4)

(1)

(5)

Primary accumulator

Primary pump inlet

Radiator inletSecondary pump inlet

Time (s)

Lith

ium

Sub

cool

ing

(K)

5

10

15

20

25

30

35

40

45

50

0 2000 4000 6000 8000

(a)(5)(4)

(3)(1)

Secondary looplosses

Primary looplosses

Primary

pump e

xit

Secondary pump exit

Primary pump inlet

Secondary pump inlet

Time (s)

Lith

ium

Pre

ssur

e (k

Pa)

200

300

400

500

600

700

800

900

0 2000 4000 6000 8000


(b)(3)

(4)

(1)

(5)

Primary accumulator

Primary pump inlet

Radiator inletSecondary pump inlet

Time (s)

Lith

ium

Sub

cool

ing

(K)

Figure 9. Lithium pressure and subcooling in SCoRe-TE NRPS.


11

C. Flow Rates and EM Pumps Currents Figures 10a shows the changes in the flow rates

of the liquid lithium in the primary and secondary loop of SCoRe-TE NRPS during the startup procedures in Fig. 7. The flow rate in the primary loop is always lower that that in the secondary loop because the temperature rise across the reactor core is always lower than the temperature decrease in the radiator panels. The specific heat of lithium is high but almost independent of temperature. At steady state nominal operation the rise in the lithium temperature in the reactor is 36.3 K versus only 27.3 K in the radiator panels (Fig. 10), and the flow rates of the lithium in the primary and secondary loops are 3.1 kg/s and 3.96 kg/s respectively. Figure 10b shows that current supplied to the primary and secondary EM pumps by the TCAs increase with time during the straps transient, as a result in the increase in the temperature difference across the TCAs (Fig. 10c). The electric current to the primary pumps is slightly higher than that of the secondary pumps, except near the end of the startup and at nominal steady state operation they are almost equal in value (~ 3150 A), and the temperature difference across the SiGe thermoelectric unicouples in the PCAs and the pumps TCAs are also very close at ~340 K.

D. Flow Rates and EM Pumps Currents Figure 11 list most of the steady state nominal

operation parameters of SCoRe-TE NRPS at the end of the startup transient and after the electrical load is connected. At such condition, the accumulators in the primary and secondary lithium loops are full to only 63% and 67% of capacity, leaving large enough operation margin for increase following an increase in the load electrical demand beyond nominal or 111.5 kWe, at 450 VDC. The lithium pressures at the inlets of the accumulators in the primary and secondary loop are 28.0 and 29.4 KPa, respectively, at which the lithium is ~ 300 and 600 K subcooled, respectively (Fig. 9b).

The nominal reactor thermal power is 2.86 MWth and net efficiency of SCoRe-TE NRPS is 3.982%. This efficiency accounts for all electrical and thermal parasitic losses in the PCAs and the electrical power supplied to the EM pumps both in the secondary and primary loops. The latter is almost the same for each primary (405.3 We at 3,163 A) and secondary (404.3 We at 3,175 A). Thus, the total electrical power supplied to and consumed by 12 EM pumps, 6 each in the primary and the secondary loops is 4.857 kWe, which when added to the load electric power of111.5 kWe, increase the effective system efficiency to 4.07%. This efficiency is lower than those of the PCAs and EM pumps TCAs (4.3%). The efficiency of the EM conduction pumps in the secondary loops of 51% is significantly higher than that of the pumps in the primary loops (37.8%) and the nominal pressure head of the later (22.1 Kpa) is ~ 10% lower than that of the former. Figure 10 also show that nominal heat rejection load per radiator panel is 451 kWth and a total of 2.706 MWth for the entire radiator. The nominal pressure losses in each of the sectors in SCoRe of

0

1

2

3

4

5

0 2000 4000 6000 8000

(5)(4)

(3)

(2)

(1)


(a)

Primary l

oop

Secon

dary

loop

Lith

ium

Flo

w R

ate

(kg

/ s)

0

500

1000

1500

2000

2500

3000

3500

0 2000 4000 6000 8000

(5)(4)

(3)(2)

(1)

(b)

Seco

ndar

y pum

p

Prim

ary p

ump

Pum

p C

urre

nt (A

)

0

50

100

150

200

250

300

350

400

0 2000 4000 6000 8000

Main PCAPump TCA

(5)(4)

(3)(2)

(1)

(c)

Time (s)

∆T

(K)

Figure 10. Changes in EM pumps operation parameters during startup procedures.


12

11.3 KPa is the largest by far in the primary loops, but slightly lower than that of the circulating liquid lithium in the radiator panel (12.2 KPa). The next section present the changes in the operation parameters of SCoRE-TE NRPS following a 50% decrease in the load demand, followed by a similar increase to the nominal load electric power.

E. Response to Changes in Load Demand

30

50

70

90

110

0 500 1000 1500 2000 2500 30001

2

3

4

5

(a)50%

100%

2.2%

3.9%

Load

Dem

and

(%)

Sys

tem

Effi

cien

cy (%

)

1165

1170

1175

1180

1185

1190

0 500 1000 1500 2000 2500 30002500

2600

2700

2800

2900

3000

(c)

2570 kW

2860 kW

Rea

ctor

Exi

t Tem

pera

ture

(K)

Rea

ctor

Pow

er (k

Wth

)

400

500

600

700

800

900

0 500 1000 1500 2000 2500 300040

60

80

100

120

140

(b)

744 V

450 V55.8 kWe

111.6 kWe

Time (s)

Load

Vol

tage

(V)

Elec

trica

l Pow

er (k

We)

30

32

34

36

38

0 500 1000 1500 2000 2500 30003.0

3.1

3.2

3.3

3.4

(d)

Time (s)

Rea

ctor

∆T

(K)

Lith

ium

Flo

w R

ate

(kg

/ s)

Figure 12. SCoRe-TE NRPS response to a change in load demands.

SCoR

e-S 11

Rea

ctor

2,86

0. k

Wth

EM pump

PCA

1 of 6 primary loops

HX

HX

HX

HX

+ – – +

+ –

1142.9 K

1144.6 K

111.5 kWe @ 450 V DC

797.7 K, 32.4 kPa, 3.96 kg/s

771.7 K

Qrad = 451 kW

TCA Rb

Heat pipes radiator

1179.2 K, 31.0 kPa, 3.1 kg/s 770.5 K

405.3 We, 3163 A 404.3 We, 3175 A

1142.9 K, 27.0 kPa

770.5 K

Load

% 4.3 =TEη

Accumulated3.0 kg, 6.8 liters

(63%)

Accumulated1.56 kg, 3.2 liters(67%)

14.3 kPa

∆P = 11.3 kPa ∆P = 12.2 kPa

29.4 kPa28.0 kPa

% 4.3 =TEη

37.8% kPa, 1.22 ==∆ EMEMP η 51% kPa, 2.25 ==∆ EMEMP η

SCoR

e-S 11

Rea

ctor

2,86

0. k

Wth

EM pump

PCA

1 of 6 primary loops

HX

HX

HX

HX

+ – – +

+ –

1142.9 K

1144.6 K

111.5 kWe @ 450 V DC

797.7 K, 32.4 kPa, 3.96 kg/s

771.7 K

Qrad = 451 kW

TCA Rb

Heat pipes radiator

1179.2 K, 31.0 kPa, 3.1 kg/s 770.5 K

405.3 We, 3163 A 404.3 We, 3175 A

1142.9 K, 27.0 kPa

770.5 K

Load

% 4.3 =TEη

Accumulated3.0 kg, 6.8 liters

(63%)

Accumulated1.56 kg, 3.2 liters(67%)

14.3 kPa

∆P = 11.3 kPa ∆P = 12.2 kPa

29.4 kPa28.0 kPa

% 4.3 =TEη

37.8% kPa, 1.22 ==∆ EMEMP η 51% kPa, 2.25 ==∆ EMEMP η

Figure 11. Nominal parameters of SCoRe-TE NRPS.


13

Figures 8c and 10a present the changes in the temperature of the heat rejection radiator panel and in the secondary coolant flow rate to the radiator as functions of time during the startup procedure. During the first 300 s of the scenario, the radiator exit temperature decreased slowly to 596 K, due to the heat losses through the insulation blanket. Following the removal of the blanket, the radiator panel temperatures decreased precipitously by ~ 35 K (Fig. 8c), as the radiator heat rejection increases to 1.05 MWth (Fig. 8b). This high heat rejection rate at such low temperature causes the rubidium heat pipes in the radiator to become sonic limited.

Figure 10c shows the changes in the electric currents supplied by the pumps TCA. Figure 10c shows the change in the temperature drop across the SiGe unicouples in the pumps TCA during the startup procedure. As this figure indicates, after removing the insulation blanket of the radiator, a large enough temperature difference is developed in the TCA, generating high enough currents to operate the EM pumps without the need for auxiliary batteries.

As the radiator temperature decreases further, the heat rejection decreases also rapidly (Fig. 8b), since the power throughput of the sonic-limited heat pipes is proportional to the saturation pressure of rubidium. This decrease in heat rejection causes a decrease in the temperature drop across the pumps TCA thermoelectrics (Fig. 10c). As a result, both the current supplied by the TCAs to the pumps and the mass flow rates in the loops decrease (Fig. 10 a and 10b). At 1,100 s, however, the ∆T across the thermoelectrics, the pumps currents and the mass flow rates reach a minimum value and start increasing rapidly (Fig. 10), reaching values > 1.0 kg/s, due to the rapid increase in the TCAs hot side temperature. Following the initial reactivity insertion of $0.80, the reactor’s thermal power increases exponentially and finally reaches significant levels, peaking at 1.05 MWth (Fig. 8b), raising the coolant temperature rapidly in the primary loop (Fig. 8c), to 680 K. This increase in reactor temperature causes an increase in the negative feedback reactivity, and a decrease in the total reactivity below zero (Fig. 8a), and the reactor thermal power decreases (Fig. 8b).

After a waiting period following the first reactivity insertion of $0.80, the reactivity insertion is resumed at time 2000 s, but at a much slower rate of 0.02 cent/s (Figs. 7 and 8a). At this rate, the reactor temperature increases at a ramp of only 6 K/mins. During the following period and to the end of the external reactivity insertion process (7000 s from start of the procedure), the reactor thermal power and temperatures increased at steady rates, while the net reactivity in the reactor core is nearly zero (i.e. the external and temperature feedback reactivities are equal in magnitude). Such increase in the reactor temperature increased the temperature differential across the PCAs, increasing the PCAs open circuit voltage. It also increased the temperature differential across the pumps TCAs, and the electrical currents and powers supplied to the pumps, increasing flow rates of primary and secondary coolants commensurately (Figs. 8 and 10).

Figures 9a and 9b show the coolant pressure and subcooling at different locations of the primary and secondary loops. As explained earlier, the pressure in the loops is controlled by the accumulators for the excess liquid volume, due to the thermal expansion of the liquid as it heats up from fusion to operating temperature. The larger the excess volume in the loop, the larger the accumulator pressure since the excess volume pushes against the spring, bellows and gas of the accumulator chamber (Fig. 6). As shown in Fig. 9a, the pressures increase steadily in the primary loop during heatup, due to the larger temperature in this loop (up to 1179 K at the end of the startup procedure). Similary, the pressure increases with time at the exit of the secondary pump. However, the pressure at the inlet of the secondary loop, the location of smallest pressure, decreases with time (Fig. 9a). This is because the pressure losses in the secondary loop increase faster than the increase in the loop pressure level due to the thermal expansion of liquid.

Figure 9b shows the coolant subcooling in the primary and secondary loops. The pump inlets are the cold leg locations of lowest pressure in the loops, while the radiator inlet and the primary accumulator (or primary inlet of main PCA) are the locations of the hot legs where the pressure is lowest; these locations are those of potential lowest subcooling. As shown in Fig. 9b, the smallest subcoolings occur at the inlet of the secondary pump for the secondary loop, and at the primary accumulator (or inlet of the PCA) in the primary loop. The accumulators of the SCoRe NRPS were designed to ensure a minimum of 200 K subcooling at any location in the primary and secondary loops.

Figure 11 shows a line diagram of a single primary loop and its coupled secondary loop. The nominal, steady-state operation parameters of the reactor and other components of the SCoRe NRPS are shown in this figure.

(g) Summary and Conclusion Presented and discussed are the effects of the type of the liquid metal coolants in the primary and secondary

loops of a SRPS with SiGe PCAs, separate SiGe TCAs for the primary and secondary EM pumps, and rubidium heat pipes radiator, during the startup procedure from an initial temperature of 500 K to steady-state nominal operation. The startup procedure lasted for a total of 3 hrs and 47 minutes. The investigated coolant combinations in the


14

primary and secondary loops are Li/Li, Li/Na, Li/NaK-78, Na/Na and Na/NaK-78. The startup simulations are conducted using a Dynamic simulation Model for Thermoelectric SRPSs (DynMo-TE), developed on the SIMULINK® platform. The SRPS investigated has no single point failure since it employs SCoRe-S11, a fast-spectrum reactor with six sectors, each with a separate liquid metal primary loop, secondary loop, and Rb heat pipes radiator panel.

For the coolant combinations investigated, the nominal electrical power varied in a narrow range from 145.2 to 151.8 kWe, the load voltage varied from 441.5 to 451.9 VDC, and the system efficiency varied from 4.97% to 5.09%. These performance parameters are attainable at reactor powers of 3330 to 3407 kWth and relatively low reactor exit temperatures of 1223-1249 K. Results indicated that the secondary, not the primary coolant, strongly impacts the nominal steady-state operation of the SRPS. The values of the load electrical power and voltage for the Li/NaK-78 and Na/NaK-78 systems are very close, but the reactor exit temperature in the former (1223 K) is 25 K lower than in the later. Because of the likely need for an auxiliary power system to initially thaw the lithium, and the added complexity of using radiological gas separators for this fluid, the Na/NaK-78 cooled SRPS may be a better choice. The load electrical power and voltage for this SRPS could be increased to 150 kWe and 450 VDC, by simply increasing the total external reactivity insertion from the value used in the present analysis ($1.49) by only 2 cents (to $1.51), at the expense of operating at a slightly higher reactor exit temperature of 1260 K.

Even though the preliminary screening of working fluids for the SRPS indicates that Li has much lower pumping requirements than NaK-78 and Na, the present results show that a SRPS designed for a specific combination of working fluids (in this case Na/Na) could still yield comparable performance with other liquid metal coolants. This is because the different thermal-physical properties of the fluids cause the components of the SRPS to operate at different temperatures and coolant mass flow rates, somewhat mitigating the differences in properties. Such results could not have been obtained using a simple engineering recipe. The present results of the startup of a fully-thawed SRPS with liquid metal-cooled reactor and SiGe PCAs illustrate the usefulness of developing a modular, dynamic system model. Such a model is necessary for understanding system operation, designing and integrating components, developing safe startup procedures, and for developing and testing schemes for adaptive and autonomous operation and control of SRPSs.

Acknowledgments This work is funded by the University of New Mexico’s Institute for Space and Nuclear Power Studies. The

opinions expressed in this article are solely those of the authors, and have neither been endorsed by nor reflect an official position of NASA.

References Barrett, M. J., and Reid, B.M., “System Mass Variation and Entropy Generation in 100-kWe Closed-Brayton-Cycle Space

Power Systems,” Proceedings of the Space Technology and Applications International Forum (STAIF-2004), AIP-CP-699, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2004, pp. 445-452.

El-Genk, M. S., Buksa, J. J., and Seo, J. T., “A Self-Induced Thermoelectric-Electromagnetic Pump Model for SP-100 Systems,” in Space Nuclear Power Systems 1986, Vol. 5, edited by M. El-Genk and M. Hoover, Orbit Book Co, Malabar FL, 1987, pp 161-172.

El-Genk, M. S., Hatton, S., Fox, C., and Tournier, J.-M., “SCoRe – Concepts of Liquid Metal Cooled Space Reactors for Avoidance of Single-Point Failure,” Proceedings of the Space Technology and Applications International Forum (STAIF-2005), AIP-CP-746, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2005, pp. 473-484.

El-Genk, M. S., Lapin, S. N. and Seo, J. T., “A Dynamic Model for the Aft Accumulator of the SP-100 System,” in Space Nuclear Power Systems 1987, Vol. 7, edited by M. El-Genk and M. Hoover, Orbit Book Co., Inc., Malabar, FL, 1988, pp. 375-382.

El-Genk, M. S., Liscum-Powell, J., and Pelaccio, D., “Bimodal, Low Power Pellet Bed Reactor System Design Concept,” in Proceedings of 11th Symposium on Space Nuclear Power and Propulsion, AIP-CP-301, edited by M. El-Genk, American Institute of Physics, New York, NY, 1994, pp. 1535-1548.

El-Genk, M. S., and Paramonov, D. V., “An Integrated Model of the TOPAZ-II Electromagnetic Pump,” J. Nuclear Technology, Vol. 108, No. 2, 1994, pp. 171-180.

El-Genk, M. S., and Rider, W. J., “Reliability and Vulnerability Studies of the SP-100 Dual-Loop Thermoelectric-Electromagnetic Pump,” Journal of Propulsion and Power, Vol. 6, No. 3, 1989, pp. 305-314.

El-Genk, M. S., and Saber, H. H., “Performance Analysis of Cascaded Thermoelectric Converters for Advanced Radioisotope Power Systems,” Energy Conversion and Management, Vol. 46, 2005, pp. 1083-1105.

El-Genk, M. S., Saber, H. H., and Caillat, T., “A Performance Comparison of SiGe and Skutterudite Based Segmented Thermoelectric Devices,” Proceedings of the Space Technology and Applications International Forum (STAIF-2002), AIP-CP-608, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2002, pp. 1007-1015.


15

El-Genk, M. S., and Seo, J. T., “SNPSAM - Space Nuclear Power System Analysis Model,” in Space Nuclear Power Systems 1986, Vol. 5, edited by M. El-Genk and M. Hoover, Orbit Book Company, Malabar, FL, 1987a, pp 111-133.

El-Genk, M. S., and Seo, J. T., “Analysis of the Transient Behavior of Thermoelectric Generators,” in Space Nuclear Power Systems 1986, Vol. 5, edited by M. El-Genk and M. Hoover, Orbit Book Company, Malabar, FL, 1987b, pp 187-198.

El-Genk, M. S., and Seo, J. T., “Parametric Analyses of the TE SP-100 System Performance,” in Space Nuclear Power Systems 1987, Vol. 7, edited by M. El-Genk and M. Hoover, Orbit Book Company, Malabar, FL, 1988, pp 399-408.

El-Genk, M. S., Seo, J. T., and Buksa, J. J., “Load Following Characteristics of SiGe/GaP Thermoelectric Generators and their Response to External Heating,” Journal of Applied Physics, Vol. 61, No. 5, 1987, pp. 2059-2064.

El-Genk, M. S., and Tournier, J.-M., “SAIRS – Scalable AMTEC Integrated Reactor Space Power System,” Progress in Nuclear Energy, Vol. 45, No. 1, 2004, pp. 25-69.

El-Genk, M. S., and Tournier, J.-M., “DynMo: Dynamic Simulation Model for Space Reactor Power Systems,” Proceedings of the Space Technology and Applications International Forum (STAIF-2005), AIP-CP-746, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2005, pp. 1005-1020.

El-Genk, M. S., and Xue, H., “Decay Heat Removal by Natural Circulation in an SP-100-550 kWe Power System for a Lunar Outpost,” Journal of Nuclear Technology, Vol. 100, No. 3, 1992, pp. 271-286.

El-Genk, M. S., and Xue, H., “Two-Dimensional Steady-State and Transient Analysis of Single Cell Thermionic Fuel Elements,” Journal of Nuclear Technology, Vol. 108, No. 1, 1994, pp. 112-125.

El-Genk, M. S., Xue, H., and Paramonov, D., “Transient Analysis and Startup Simulation of a Thermionic Space Nuclear Reactor System,” Journal of Nuclear Technology, Vol. 105, No. 1, 1994, pp. 70-86.

Harty, R. and Mason, L., “100-kWe Lunar/Mars Surface Power Utilizing the SP-100 Reactor with Dynamic Conversion,” in Proceedings of 10th Symposium on Space Nuclear Power and Propulsion, AIP-CP-271, Vol. 2, edited by M. El-Genk and M. Hoover, American Institute of Physics, New York, NY, 1993, pp. 1065-1071.

Hawley, J. P., “Water Immersion Safety for SNAP Reactors,” Atomics International Report NAA-SR-11361, September 1967.

King, J. C., and El-Genk, M. S., “Spectral Shift Absorbers for Fast Spectrum Space Nuclear Reactors,” Proceedings of the Space Technology and Applications International Forum (STAIF-2005), AIP-CP-746, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2005, pp. 285-294.

Liscum-Powell, J., and El-Genk, M. S., “Neutronic Analysis and Design Optimization of the Pellet Bed Reactor for Bimodal Applications,” in Proceedings of 11th Symposium on Space Nuclear Power and Propulsion, AIP-CP-301, Vol. 3, edited by M. El-Genk, American Institute of Physics, New York, NY, 1994a, pp. 1548-1563.

Liscum-Powell, J., and El-Genk, M. S., “Options for Enhanced Performance of Pellet Bed Reactor Bimodal System,” AIAA/SAE/ASME/ASEE, 30th Joint Propulsion Conference and Exhibit, Paper No. AIAA 94-3167, Indianapolis, IN, 27-29 June 1994b.

Marriott, A. T., and Fujita, T., “Evolution of SP-100 System Designs,” Proceedings of the 11th Symposium on Space Nuclear Power and Propulsion, CONF-940101, Vol. 1, edited by M. S. El-Genk and M. D. Hoover, American Institute of Physics, New York, NY, 1994, pp. 157-169.

MatLab, “MATLAB® 7.0.1,” http://www.mathworks.com/products/matlab, accessed September 2004. Metzger, J. D., El-Genk, M. S. and Parlos, A. G., “Model-Reference Adaptive Control with Selective State-Variable

Weighting Applied to a Space Nuclear Power System,” J. of Nuclear Science and Engineering, Vol. 109, 1991, pp. 171-187. Metzger, J. D., and El-Genk, M. S., “Application of a Model-Reference Adaptive Controller with Selective State-Variable

Weighting to an SP-100 Space Nuclear Power System,” J. of Propulsion and Power, Vol. 8, No. 5, 1992, pp. 1093-1102. Paramonov, D. V., and El-Genk, M. S., “Development and Comparison of a TOPAZ-II System Model with Experimental

Data,” Journal of Nuclear Technology, Vol. 108, No. 2, 1994, pp. 157-170. Paramonov, D. V., and El-Genk, M. S., “Analysis of Ya-21u Thermionic Fuel Element,” Journal of Nuclear Technology,

Vol. 116, No. 3, 1996, pp. 261-269. Paramonov, D. V., and El-Genk, M. S., “Test Results of Ya-21u Thermionic Space Power System,” Journal of Nuclear

Technology, Vol. 117, No. 1, 1997, pp. 1-14. Poston, D. I., “Nuclear Safety Calculations for Heatpipe Power System Reactors,” Proceedings of the Space Technology and

Applications International Forum (STAIF-2002), AIP-CP-608, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2002, pp. 748-758.

Rider, W. J., Parametric and Transient Analyses of the SP-100 System, M. S. Thesis, The University of New Mexico, Albuquerque, NM, August 1989.

Simulink, “SIMULINK® 6.1,” http://www.mathworks.com/products/simulink, accessed September 2004. Thieme, L., and Schreiber, J., “Advanced Technology Development for Stirling Converters,” Proceedings of the Space

Technology and Applications International Forum (STAIF-2004), AIP-CP-699, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2004, pp. 432-439.

Tournier, J.-M., and El-Genk, M. S., “Reactor Lithium Heat Pipes for HP-STMCs Space Reactor Power System,” Proceedings of the Space Technology and Applications International Forum (STAIF-2004), AIP-CP-699, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2004, pp. 781-792.

Tournier, J.-M., and El-Genk, M. S., “Liquid Metal Heat Pipes Radiator for Space Nuclear Reactor Power Systems,” in these Proceedings of the 3rd International Energy Conversion Engineering Conference (IECEC-2005), American Institute of Aeronautics and Astronautics, 2005.


16

Truscello, V. C., and Rutger, L. L., “The SP-100 Power System,” in Proceedings of the 9th Symposium on Space Nuclear Power Systems, AIP-CP-246, edited by M. El-Genk and M. Hoover, American Institute of Physics, Melville, NY, 1992, pp. 1-23.

Wright, S. A., and Sanchez, T., “Dynamic Modeling and Control of Nuclear Reactors Coupled to Closed-Loop Brayton Cycle Systems Using SIMULINKTM,” Proceedings of the Space Technology and Applications International Forum (STAIF-2005), AIP-CP-746, edited by M. S. El-Genk, American Institute of Physics, Melville, NY, 2005, pp. 991-1004.

Startup and load-following transients of a thermoelectric space reactor system with no single point...

Documents

Transcript of Startup and load-following transients of a thermoelectric space reactor system with no single point...