ML20214W748

Midwest Fuel Recovery Plant Technical Study Rept

ML20214W748
Person / Time
Site:	05000268
Issue date:	07/05/1974
From:	GENERAL ELECTRIC CO.
To:
Shared Package
ML20214W733	List: ML20214W730 ML20214W748
References
1516, NUDOCS 8706160168
Download: ML20214W748 (51)

Text

,z .

3

..o- u ;;

1 l

l l

MIDWEST FUEL RECOVERY PLANT l l

L TECHNICAL STUDY REPORT l

JULY 5,1974 l-

\

i -

l 1

1

, _ _ ;.N l

1518 8706160168 840708 .

PDR ADOCK 0 % g 8 G E N E R A L ,' E LECTRIC

. . . . _ . _ m . . -

a c:

TABLE OF CONTENTS

} E.a.p .

I- INTRODUCTION AND BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1'

$  : A. Background and Reasons for the Study- . . . . . . . . . . ._ . . . . . . . . . . . . . . . . . . . . - 1 B. Objectives of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 C. ' The Nuclear Fuel Cycle; Purpose of the MFRP . . . . . . . . . . . . . . . . . . . . . . . . . . 3

- D. The G E Aquafluor Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 E. - Plant Description . . . . . . . ...................................... 10 H CORPORATE STUDY ORGANIZATION AND PROCEDURE . . . . . . . . . . . . . . . . . . . . . . 14 A. Advisory Board Established ...................................... 14 B. . Investigation Teams EstabVahed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 C. In-Depth Technical Reviews Conducted and Evaluated . . . . . . . . . . . . . . . . . . . . . 16 III OBSERVATIONS AND CONCLUSIONS ON FUNDAMENTAL TECHNICAL 4 PROBLEMS ................................................... 18 A. Radioactivity in the Uranium Nitrate Process Stream Requires Re mote Ope ration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

18

' B. Fission Product Deposition in the Uranium Conversion and Purification Syste m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

- C. Inoperable Vital Canyon Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

1. Critical Processes and Equipment Description . . . . . . . . . . . . . . . . . . . . . . . . 21

, a. UNH Calciner and Related Equipment . . . . . . . . . . . . . . . . . . . . . . . . . 21

b. Solids Transfer System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-

c. Fluorinator and Related Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2. Proble m H istory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

a. UNH C alc ine r Feed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

b. Bed Caking and Calrod Burnout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

- c. Solids Transfer from Calciner to Fluorinator . . . . . . . . . . . . . . . . . . . . 32

d. Erosion and Plugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

e. Other Problems . . ...................................... 34

- D. Plant Configuration and Close Coupling of Process Equipment . . . . . . . . . . . . . . . 35 IV OUTLOOK FOR ACCEPTABLE SOLUTION OF MAJOR PROBLEMS . . . . . . . . . . . . . . . 37 V APPRAISAL OF OVERALL PLANT OPERATING CAPABILITY . . . . . . . . . . . . . . . . . . 40

' 42 VI - ALTERNATIVE PROCESS FLOW SHEET AND PLANT CONFIGURATION . . . . . . . . . . .

VII - ' APPENDIX .................................................... 44 i i 4

4- -- <-e,---- --,-.-----e e m-, -- - ,y ----r,-s ---m - - - - -

---m------- . - - , , - - , ,----~y,m- -- e w -s , w v. v vw.r><-y -e,- y - - - ,

MIDWEST FUEL RECOVERY PLANT TECHNICAL STUDY REPORT JULY 5,1974 I INTRODUCTION AND DACKGROUND A. Background and Reasons for the Study j i

The start-up of the General Electric Midwest Fuel Recovery Plant (MFRP) (Figure 1-A-1) l has been extended many times. Original estimates were for mid-1970, but construction and equipment delays extended that date a year. Equipment testing actually started in November 1971, at which time it was anticipated that hot start-up would follow in mid-1972.

Numerous equipment failures and operating problems were encountered that made it impossibir to operate the plant. Testing and modification of individual pieces of equipment continued. In September of 1972 cold uranium was introduced into some pieces of equipment for the fi'.'st time. By that time, the hot start-up date had been extended to April 1973, and recer_tly ic October 1974, which date now seems impossible to meet because of additional recent equipment failures. It is now clear that the basic concept as originally conceived of a simplified decontamination process incorporating remotely operated and maintained solids handling and waste disposal systems has developed many unforeseen technical problems.

As originally conceived in 1964, Corporate management was expected to invest $20 million in the development and demonstration of the Aquafluor* plant and process. In 1968 the Nuclear Energy Division requested and obtained Board of Directors authorization to construct a demon-stration plant with an annual output of 300 metric tons at a cost of $36 million. In 1970, an additional $8 million was authorized to cover overruns, for a total of $44 million to complete plant construction.

Because of the equipment failure problems mentioned earlier, it has been necessary to purchase replacement equipment and make modifications to the plant. The high level of installation and start-up effort has built up on-site employment to about 160 people, with an aggregate annual payroll of about $4,000,000.

By mid-1974, including casks for transporting spent fuel and pre-operational start-up expenses, General Electric will have expended $64 million in the Midwest Fuel Recovery Plant, which continues to be a nonoperational facility.

In March of 1974, with the projected hot start-up date half a year away, Mr. R.H. Jones, Chief Executive Officer of General Electric, requested Dr. C.E. Reed, Senior Vice President -

Corporate Studies and Programs, to make a corporate review of the technical and operational capability of the MFRP. This study was requested because loading the MFRP with irradiated fuel represents a critical step. In view of the serious problems encountered during the three-year start-up period, it was felt that this step should not be taken until after an independent high level review had been conducted.

• Service mark of General Electric Company 1

m

,m ff[k l g u

fh, i

,s

~

-% N N g p, i .$, " kT I;l v7:%;eb

~

s

y i

- Q G n 'A .

• gcq@n- ]c4 Z Agu -

x

?

- r u

u-4

_a 3

, i , ,

& 4 / I Y ,$

m :G 1' F,. . ',,"i..94" h-V g sp '

9

_ E.

j h,  : c .

g M .c ~ \, t

- 'd . u-p [  %,,;' A. 2

>v * -

' l-d W\

M O t Q'%

a b i E

N;s

< a m

[!!i

-4 fd i D$h t

1' T),i .

~

g m a pr ' i pg  ;

Z W M31 r 6'y9 fl s

$h k '( b ,

8

, bli q - ,

e,,

d+

fi <an ~<wy u.pu 3,f i ,*

t >Y$ s

[. ~ lA . , 1 8 -

f . . . Yl '. *E

""i ' \y "?

9 c'3 a 4 -

B. Objectives of the Study The objectives of the study were to focus primarily on:

1. An objective examination of the fundamental technical problems of the process and equip-ment.

2. The outlook for solution of the fundamental technical problems.

3. An assessment of the effect of these technical problems on the outlook for the operating capability of the plant.

C. The Nuclear Fuel Cycle; Purpose of the MFRP The fuel cycle shown in Figure I-C-1 indicates the course of uranium fuel during its use and reuse. Natural uranium, as it is mined, is a mixture of isotopes, predominantly U 238 (99.28%) and U 235 )* 235 is fissionable and the U 238 is " fertile," i.e. , capable of conversion to a fissionable isotope (Pu by neukon esorpdon. m U must be present in 239 235 greater than natural abundance for the uranium to be useful as fuel in light water moderated reactors.

The Mine and Mill produces uranium oxide (U38 0 ) in an intermediate form known as yellow-cake, which is shipped to a conversion plant where it is made into uranium hexafluoride (UF 6*

The UF6is transported to an AEC gaseous diffusion enrichment plant where the fissionable isotope (U235) e ntent is increased three to seven times. The enriched UF6 is hn transporkd to the fuel fabricator for processing into fuel rods and fuel bundles, as shown in Figure I-C-2.

At the fuel fabrication plant the mixture of U and U is topes is converted from 235 238 uranium hexafluoride (UF6) to uranium dioxide (UO2 ) which is a brown powder. The powder is pressed and sintered to a ceramic-like fuel pellet. The pellets are ground to size (typically about 1/2 in diameter and 1/2 in. long) and seal welded into thin walled zirconium alloy tubes about fourteen feet long. These fuel rods are thoroughly tested and assembled into fuel bundles ready for insertion into a reactor.

In the energy-producing nuclear fission reaction occurring in the reactor, U Hssions to 235 a wide variety of lower atomic weight elements, many of which are highly radioactive -- com-monly known as fission products. In addition, a small part of the U in the reactor is con-238 verted by neutron absorption to fissionable plutonium (Pu239). Pu 239 can also be used as a fuel in power reactors. The fuel bundles are removed from the reactor before all of the U is 235 consumed. The reactor design is such that approximately 25% to 35% of the fuel bundles are removed each year and, after a suitable cooling period, transported to a recovery plant (also known as a reprocessing plant). The purpose of the recovery plant is to separate and de-contaminate unburned uranium and plutonium from the fission products and to convert radio-active wastes to forms suitable for safe long-term storage.

The recovered uranium, reduced in fissionable uranium (U235) content, yet greater than natural ore, must now be returned to the diffusion plant for re-enrichment prior to being fabricated into new fuel bundles. The recovered plutonium is stored for future use in nuclear fuel.

3

.,.=

4 NUCLEAR FUEL CYCLE Mill- , Conversion to -

Mine -

Uranium -

Uranium Hexafluoride ,

Ore Extraction . . Yellow Cake UF 6-1r I Enriched Enrichment Cascades d O UF 6 Cylwers UO2 Pellets.in U 235 Fmm, to 2 -. 6 M ds Zirconium Rods AEC Operated n.

k,

• Enriched Partially Fuel Enriched !

[ Bundles -UF 6-4 1r Utility Midwest Fuel Recovery ---DPlutonium -

Power Spent Fuel Plant ---> Neptunium Reactors Bundles

+ +-

High Low Activity - Activity -

Waste Waste

COMPONENTS OF BWR LIGHT WATER REACTOR O ,f 3 Ir g a E5 E8 l

= l ,

._p 1/2" p c_ Mb .-

14 ft ,

5 - f a

o URANIUM s 'sm~s s

N y ,V i, ,

Y 'f DIOXIDE

$ FUEL PELLET I

'\~

's % /p N

• s wjh[/'

%_/

1r  %

% J J.) A 9

%v' %_- _a d FUEL BUNDLE CORE OD 1 CORE = 3293 Mw (THERMAL) 1 CORE = 764 Bundles x 49 Rods / Bundle = 37,436 Rods 1137 MwE (NET ELECTRICAL) = 37,436 Rods x 288 (Approx) Pellets / Rod ~ 10.8 Million Pellets

,n's Fer N . "1) = 37,436 Rods x 14 Ft./ Rod = 524,104 Ft. = 99.26 Miles More Recent Designs are 8 x 8 = 764 Bundles x .1892 Metric Ton / Bundle = 144.5 Metric Tons l

Recovery The fuel in a commercial light water reactor requires removal and reprocessing at periodic intervals because of loss of reactivity due to burnup and accumulation of fission product poisons or when loss of fuel integrity is threatened due to irradiation damage, corrosion or mechanical damage to the fuel bundles. At the time of discharge, the fuel contains valuable fissile materials. It is the function of reprocessing to recover these materials in a form suitable for fuel refabrication. In general, reprocessing entails separation of fuel material from cladding and assembly hardware, decontamination of residual uranium and plutonium from fission products and separation of the uranium and plutonium.

Decontamination processes yield recovered uranium and plutonium with activity levels at or near background levels for the respective product materials. High decontamination of the uranium product is required because the product is normally to be recycled to a gaseous dif-fusion plant for re-enrichment. The high activity wastes (often referred to as high level wastes) from this process must be concentrated, dried, encapsulated and stored until they can be sent to an AEC site for permanent storage. The low activity wastes (often referred to as low level wastes) must be concentrated, stored and finally shipped to a suitable burial ground.

Several processes have been developed under AEC programs for the recovery of special nuclear materials from the nuclear fission reaction products. These processes have been adapted for use in reprocessing spent fuels from power reactor plants currently in operation or under construction. There are two types of processes; the first is based entirely on aqueous chemistry. The second is a combination of aqueous and fluorine dry chemical process-ing (mixed process).

Re-enrichment The recovered uranium hexafluoride (UF ), which still contains more U than does 6 235 natural uranium, is returned to the AEC enrichment cascades where it is processed with the material coming from the mines and other sources to re-enrich the U content by stripping 235 out some of the U238. The enriched UF6 is then ready for conversion to UO2 "t " I"'I fabricator's plant.

D. The GE Aquafluor Process The Aquafluor Process was developed by the General Electric Nuclear Energy Division to reduce reprocessing cost by innovative combinations of chemical processes and by improve-ments in design of reprocessing canyons. These canyons are costly massive concrete structures with walls several feet in thickness designed to shield highly radioactive chemical processes. One goal was to reduce costs sufficiently so that smaller reprocessing plants could be economic and could be located in areas relatively near to groups of nuclear power plants, thereby reducing the transportation problems and costs and facilitating compliance with regulatory requirements.

Simplification of the chemical process flowsheet was accomplished by taking advantage of the decontamination potential of the fluoride volatility process for the uranium, and of the ion 6

exchange process for the plutonium. While the established AEC approach for reprocessing nuclear fuels is to remove fission products and to separate uranium by three or, more recently, two cycles of solvent extraction, the origina: AEC work on decontamination with the dry fluoride process showed some promise of accomplishing uranium decontamination with no solvent extraction step. Work done by GE indicated that two cycles of ion exchange could decontaminate the plutonium product, also with no solvent extraction step. In the light of this background, the initial Aquafluor concept envisioned the ultimate in simplicity by eliminating all solvent extraction. Ilowever, as a result of additional development work in the 1960's, and in order to provide a margin of conservatism, it appeared desirable to use one cycle of solvent extraction to eliminate the bulk of the fission products upstream of the ion exchange and dry fluoride processes.

As now conceived, the Aquafluor Process, by using the decontamination capability of the ion exchange for plutonium and exploiting the relative volatility differences of uranium hexafluoride and fission products fluorides, eliminates the second and third cycle of solvent -

extraction from the traditional process. The flowsheet is also designed to recover neptunium because of its potential value. The uranium output is in the form of hexafluoride, the required form for use in the AEC diffusion enrichment plant. The process also facilitated conversion of all radioactive liquid wastes to the solid form, reducing the storage area required and storage costs. With no radioactive liquid effluents, the process eliminated undesirable envi-ronmental effects, thereby minimizing licensing and site location problems. Simplification of the canyon design was accomplished primarily by developing a concept of mounting process equipment on plugs with many of the process connections brought out of the canyon through the plug so that they could be disconnected manually. This approach eliminated a large number of remotely installed jumpers which are bulky and require considerable space. The resulting re-duction of canyon size and investment was expected to be a major factor in reducing repro-cessing costs.

A brief description of the Aquafluor Process follows (refer to MFRP Flow Chart, Figure I-D-1).

The first step in the fuel recovery process is the mechanical disassembly of the fuel bun-dies taken from the fuel storage area (indicated on Figure I-D-1 as Fuel Storage and Shearing).

Groups of rods are then sheared into short segments and the uranium, plutonium and fission products are leached from the cladding using nitric acid in a semicontinuous dissolver tray.

The leached cladding hulls are rinsed and monitored to verify that all the above materials have been dissolved. Finally, the hulls are emptied into a cladding cart with the rest of the

! external hardware for transfer to a water-filled cladding storage vault, an underground structure of reinforced concrete lined with stahdess steel.

The dissolver solution containing the uranium, plutonium, neptunium, and fission product wastes is sent to the solvent extraction step where more than 99.0% of the fission product wastes are removed from the process stream (Solvent Extraction). These highly radioactive wastes are combined with other high activity wastes, concentrated, and then calcined into a dry solid oxide form (Illgh Level Waste Processing). This is canned in stainless steel 7

s

co MFRP FLOW CHART FUEL STORAGE SHEARING SOLVENT EXTRACTION Pu. Nr ION-EXCHANGE LOAD OUT IT'E . , , , ._,

^=----N = ' - ' "

J..".,-

.n .; s-- ... . .

-- ---i; <m.;:- .-

,=;o

~ ~ ' - >

D , .,

- s:.

[ ~~.

,3.=. T.

g o .

C_M,:a -j '

v. A'l m n C"

p

.- Q 0":, fr ,,

-m. d.. tt Qet:v ~a y/* Pll g ..

a-.- p r,%p:.

an g . . , .y ygc.s + wl , '~l*[- 3

-,.A

~,d

~^ W

• t" 4.-,

'C- g-k %.L

!i 1 1 G~ Q p ,,

.~ 4 c.".

' ' e,e, Ts 1 L- .  ::

er-("='.=. 's

• ~

' Q- - _

'"^f - e 7 e:. Q .:n .

W,

, L.  :,h 3.,:..  %,

~

1,..

_. . . y .,. --

.- _T -a Q,,_.

i L , ,,,' "" t t LC kj ,qq ,

b.

U_-- - M m

F2 GENERATION

, , URANIUM CONVER3lON m- ::;

l di f,,-..D AND PURIFICATION /LDADOUT g "-f, gr.,. ~

• ~~~~kr W:

• - r ,at:\

~ ~ ' -

1 5: . ,

-. - ~

3 a 2 .: . l

- . :;. 4.:, _ gu -

p 3 4_4, '_$ Y

t. ~g [.-

aN h .._ h .; g /;{,t p

1. ~

1. D 73 ;

_ _ r.

i .

. _p t .. .7 p,,,,,.,.,,, ..,

x".

Am,p '

s. g l

,,,,,,, lJ 9

am. t w

• =. s~,,y_~

m;r r..g 1

m,_,-

=, a ..-

a; "i*"T .. _ rt" I b _

5' { ',1 H4{y~~-**

,, _ , , , , ,,,,,,.,9 OFF-GAS TREATMENT h 7"4 .Jt4'

• JJ,,, g

_~_ 7 . m. Cn" 6 Wl

-/

W '~ /

d H L_,7 p.,

e' . I i aq -

,, 4

= -- '. a. 4g ti iiin LOW LEVEL e HIGH LEVEL WASTE PROCESSING WASTE PROCESSING h= 7g,:c .,, La ' ""

, -<t

, = - ,

Y b J

. = - .is 9,,

n -

grs .

=U .<%-

l- - . = -

.9, ,,Q rr ma.== ...

g-F' m

.x,.

g.. ' Yd' e f 4 d '"

I# y n. ,.,.,,.

s -- -_

. .,y p , m s. -~ f v. l w.~ -- .., .. :-

A c. ~ 4, W,. /**"" .;. .kM, ,

'G '

a

.h ' gj; ,,

~ (0f .

_' -Q 5.01, a

_. 8

" y P LJ e'. m N,':' .- -

w pL- =-

^

% ~

.:r .,c..- .

c =,,::.r

containers which, after a suitable monitoring period, are transferred to the high level waste storage basin. Here they are stored under water for several years awaiting transfer to a federal repository.

The aqueous process stream bearing uranium, plutonium, neptunium and a small amount of remaining fission products then passes through two successive ion-exchange units, where the plutonium and neptunium are extracted as purified nitrate solutions (Pu, Np, lon-Exchange Load Out). These solutions are then concentrated and loaded into ten-liter AEC-approved containers.

After further concentration, the uranium-bearing process stream is calcined to UO 3"" I fluidized solids reactor calciner (Uranium Purification Load-Out). The UO3is then sent to a fluidized solids reactor fluorinator where it reacts with elemental gaseous fluorine (F Genera-2 tion) to form gaseous UF , leaving behind the less volatile fission product fluorides. The UF 6 6

is passed through a sodium fluoride trap, a magnesium fluoride trap and a final distillation-purification step before being loaded into standard cylinders for shipment.

The low activity aqueous radioactive wastes from the process are concentrated into a near-saturated solution and then pumped into a double-lined reinforced concrete vault (Low Level Waste Processing). After cooling, the major portion of this waste precipitates and forms a solid salt-cake. As the waste solidifies, the alkaline liquid remaining on the surface continues to be recycled through the process for evaporation of the water. Thus there is a minimum amount of low-level wastes in liquid form in the vault at any time. The low activity dry chemical wastes resulting from the fluorination portion of the process are stored in another lined, reinforced concrete vault.

The waste sarage vaults at MFRP were sized to provide storage space for several years of production. For each metric ton of fuel processed, there are about 15 cubic icet of clad-ding waste, 2 to 3 cubic feet of high activity waste,10 cubic feet of low activity waste, and 10 cubic feet of dry waste.

All of the steam, cooling water, and process condensates are recycled in the process.

Much of the nitric acid also is recovered and reused, which eliminates the requirement for disposing of substantial quantities of low level liquid waste. No potentially contaminated liquid wastes are discharged from the plant.

The gaseous effluent from the process stream is chemically scrubbed and then filtered to remove traces of particulates and halogens, and it is finally combined with ventilation air and l passed through a deep-bed sand filter for final filtration before discharge to the atmosphere.

The small quantities of radioactive tritium and krypton associated with the fuel are discharged to the atmosphere. The resulting concentrations are well below safe levels established by the A. E. C.

In summary, the innovative combination of solvent extraction, ion exchange and fluoride decontamination, along with a new type of canyon design was expected to result in an advanced type of licensable reprocessing plant lower in cost and flexible in location because of its smaller size and complete elimination of contaminated liquid effluent wastes. The resulting economy in plant investment and operating costs was expected to make a significant contribution to the reduction in overall nuclear fuel cycle costs.

9

E. Plant Description General The Midwest Fuel Recovery Plant is within a security area of about 25 acres located at the north edge of an approximately 890-acre Company owned site. The site is near Morris, Illinois, immediately south of, and directly adjacent to, the site of the Dresden Nuclear Power Station. All radioactive material handling related to the operation of the MFRP is within the security area.

The sccurity area encompasses all process and service facilities, including the process buildirg, the fluorine generation building, the utility building, the warehouse and shops building, the administration building and the chemical storage facilities.

Process Building The process building and its appurtenances contain all radioactive materials on-site except for the recovered uranium which is stored in approved shipping containers prior to shipment.

The process building, excluding the cask handling, fuel and waste storage areas, is a relatively small structure, about 125 feet long x 80 feet wide x 120 feet high, with up to 20 feet of this last dimension below grade. The principal processes in this building are detailed in Figure I-E-1.

Feed Preparation and Cladding Disposal Facilities are provided for receiving spent fuel in casks via either truck or rail. The structure over the cask decontamination area is of steel frame and insulated siding construction of adequate tightness to maintain proper ventilation control. The fuel unloading and storage basin area are stainless steellined, concrete structures below grade with steel frame and insulated siding above grade. The basins are water filled to provide cooling of the spent fuel and to provide shielding during fuel unloading, handling and storage. There are gates between the basins to permit transferring material from one basin to the other.

The fuct bundles are unloaded from the shipping casks and stored in the fuel basin until ready for processing. At that time they are transferred one at a time to the mechanical cell where they are chopped up and the spent fuel, fission products, plutonium, etc. are chemically separated from the structural parts of the fuel bundle.

The mechanical cell is about 50 feet long x 13 feet wide x 26 feet deep. This cell contains the equipment for fuel bundle disassembly, UO2 core dissolution and the high-level waste packaging operations. These operations are accomplished by operators using specialized tools, equipment, and manipulators (masterslave and powered) while viewing the work through five-foot thick shielding windows.

Fuel Recovery The process canyon area contains equipment for processing the highly radioactive materials.

This equipment is located in cells designed for remote operation and equipment replacement 10

. ,1

~~

' T; E /:

i j

{:

g  : .

k

![j . .

t '( t ',-

I / .' ' ,

,, l j

W > ,t

.* k - . )

I-  ?

-; j;

,/ ,

. .. ,A lin s :I zi h;y ) b b w: ,

. .i kl 7 1

, 'I -

. , J ') j  !!  !!

o

' ~

f%

i g

\qi s hj -

__,f l'

\

% J

- i l o t- 5 g . m g_ . m

,A

• ) ' k U-Yp <

'\

't s' 7-f ,

l g j,

f. _A . .

N s ,1 0 ll' f.;b Y-l _ f. .

ll} '

e

, k 1.; '$l.

< er . .s.

J.

- y

' ;l kh 3

'3ll iH

.. ; . f E)*)

h-  ; . .-  ? . . .

w$ A+~ - i E .u:.a Figure I E t

with no personnel access permitted. This part of the plant is constructed with thick walls and roof of heavily reinforced concrete. The intercell walls are designed to provide ventilation control and process isolation as required. Sheet metal cell covers are used to control air flow in the process area.

The chemical processing cells occupy most of the space in the process canyon area. The cells opposite the mechanical cell are about 42 feet deep with the other cells being 32 feet deep.

In general, the chemical processing equipment is supported from the cell walls. liighly reliable equipment is mounted in the lower part of the cells; equipment that requires remote i maintenance and replacement is placed for easy access from the overhead crane. All instrumentation elements located in the cells that are essential to process control and safety are designed to be remotely operable and maintainable.

A decontamination cell is provided for internal and external decontamination of process equipment prior to storage or disposition in a vault or removal from the building for shipment -

to a licensed disposal area. The decontamination cell is stainless steel lined; the floor area is about 7 feet x 91/2 feet.

A labyrinth area at one end of the process canyon area provides for performance of crane maintenance and for transfer of replacement equipment into, and decontaminated equipment out of the process canyon area. The labyrinth providcs full shielding from the process equipment in the canyon area for personnel who may be performing maintenance in the area. All equip-ment transfers are through an opening in the maintenance area. In addition, the labyrinth area is equipped with doors to provide ventilation control and to prevent inadvertent personnel access to the canyon.

The canyon is surrounded by multi-level gallery areas which house the equipment for final product processing and loadout, as well as aqueous makeup, personnel control lobby, process control area, mechanical cell operating area, Pu and Np product storage areas, laboratory, counting room, operating and laboratory offices, and service and supply galleries.

Equipment for final processing of product streams and for loadout and packaging of these materials is in a reinforced concrete structure enclosed in hooded areas with appropriate ventilation for safety and containment. Some equipment, where required for protection of personnel, is lightly shielded with concrete walls and remotely operated. Equipment in these areas, however, is designed for contact maintenance under special control conditions.

The canyon service crane is operated from a separate gallery on top of the process building.

The gallery extends the length of the process canyon area and is equipped with shielding windows for direct observation of the canyon area. Provisions are made for use of remote TV to supplement the windows for viewing. This is expected to permit performance of equipment maintenance and replacement functions in the canyon area without the potential for exposing the crane operator to high radiation doses.

Dry Chemical Waste Treatment Dry chemical wastes from the fluorinator and following purification processes are dumped into a dry chemical waste vault.

11

Illgh Level Waste Treatment liighly radioactive fission product wastes from the solvent extraction step, and other aqueous wastes of particular concern, are concentrated and calcined into a powder which is canned in the mechanical cell and stored in a storage pool.

Low Level Waste Treatment Low level aqueous radioactive wastes from the process are concentrated to a near-saturated solution and then pumped into a double-lined reinforced concrete vault.

On-Site Waste Storage in general, the waste storage vaults have sufficient volume to handle wastes from several years operation at anticipated processing rates, with p.ovisions for expansion as required.

The storage basin for canned high level wastes is immediately adjacent to, and of the same type construction, as the fuel storage basin. This basin is serviced by the same crane that services the fuel basins. This basin is also water filled to provide radiation shielding and cooling of waste containers. The basin water is monitored to detect increased activity levels that would be indicative of leaking containers and is treated to maintain the temperature and radioactivity concentration at acceptable levels.

Off Gas Treatment The small amount of gas evolved from the process passes through a scrubber, reactor and in-line filter before being combined with the building ventilation exhaust stream, which is then drawn through a very large sand bed filter and discharged up the plant stack.

The ventilation system performs the normal function of fresh air supply and personnel com-fort control, as well as the safety functions of controlling radioactive contamination within the plant. These functions are accomplished by use of a once-through system in which the air supply is filtered, washed, and heated, and then distributed to the various process building zones through a supply duct with pressures controlled to assure airflow from zones of no or slight contamination to zones of potentially higher contamination. The building is constructed and scaled as required to assure that all air is discharged from the building through the high efficiency, graded sand filter and the 300 foot high stack.

There are two ventilation supply units, each having a capacity of 16,000 cubic ft./ min

(= 2/3 of the normal ventilation air flow). Each supply unit has intake control dampers, a fan, filter, preheater, and air washer. In addition, there are three ventilation exhaust fans, each having a capacity of about 50% of the normal airflow.

There are backflow filters in the ventilation openings between zones of potentially differing con-tamination levels. These are high officiency filters which are provided to protect persormeland to prevent contamination spread in the extremely unlikely event of momentary airflow reversal.

Other On-Site Supporting Facilities Warehouse and Shop Dullding A single-story, steel-frame-and-insulated-siding structure of approximately 4000 sq. ft.

houses necessary on-site shops and provides space for receipt and storage of process materials, 12

spare equipment, etc. It is also used to store loaded UF6 pr duct containers awamng off-she shipment.

Utility Building A separate building of approximately 3500 sq. ft., similar in construction to the warehouse and shop building, houses the steam, water, electrical, and air service facilities for the site, as well as clothing change room, lunch room, and other personnel facilities.

Fluorine Building Elemental fluorine for conversion of UO 3 F in the fuel recovery process is generated 6

at the MFRP site in facilities which are separated from the main process building. The structure which houses the fluorine generation equipment is of steel frame, insulated-panel siding construction, on a concrete slab foundation at grade. The main structure is approxi-mately 50 feet in length x 36 feet in width and contains a fluorine cell and a maintenance area, separated by removable access panels, as well as a hydrogen equipment room, a fluorine equipment room, and an electrical equipment area, flydrogen fluoride storage facilities are located outside of the fluorine generation building at an appropriate distance and are provided with weather protection.

13 1 .-

II CORPORATE STUDY ORGANIZATION PROCEDURE A. Advisory Board Established In March 1974, as a first step toward evaluating the technical capability of the Midwest Fuel Recovery Plant, Dr. C. E. Reed established and convened the following Advisory Board of experts to make a preliminary survey and scope of the technical problems. 'Ihe Board members were chosen for their previous in-depth experience in radio-active chemical processes, decontamination and maintenance of equipment and the pertinent types of chemical engineering and process technology in use at MFRP (see Appendix for biographical data).

Advisory Board Dr. C. E. Reed Chairman Senior Vice President Corporate Studies and Programs Mr. A. W. Robinson Secretary Staff Executive Corporate Studies and Programs Dr. R. H. Beaton Vice President and General Manager Energy Systems and Technology Division Power Generation Business Group Mr. D. E. Debacher General Manager j Silicone Products Dusiness Department Dr. W. H. Reas Manager-Nuclear Process Development Vallecitos Nuclear Center Nuclear Energy Products Division Dr. R. D. Richards Manager-International Business Development International Operations Department Nuclear Energy Products Division Dr. S. Seltzer Manager - Intermediates Manufacturing Silicone Products Business Department Mr. W. D. Webster General Manager - Overseas Nuclear Projects Department Nuclear Energy Products Division Following on-site briefing by plant management personnel, the Board inspected the MFRP and observed the nature of the many start-up equipment failures and process operating problems.

The following five general areas of major concern were identified for detailed study to develop, in depth, the nature and basic causes of the problems and the outlook for their solution:

• Process Chemistry and Control e Overall Process Operations and Equipment 14

i e Fluidized Solids Reactors and Solids Processing e Remote Maintenance and Operability e Quality Assurance and Safety Dr. Reed established on March 13 a field office at the Morris Plant, from which during the next 11 weeks he supervised the organization, direction, and coordination of the studies of these problems.

D. Investigation Teams Established Experts with special related prior experience were chosen to make in-depth observations, l study and analysis in each of the five areas of major concern identified by the Advir ory Board I f (see Appendix).

Investigation Teams l Process Chemistry and Control Dr. D. II. Ahmann Chairman Manager - Engineering Neutron Devices Department Dr. S. Lawroski Argonne National Laboratory Mr. D. W. Lillie Liaison Scientist Power Generation Business Group Corporate Research and Development Overall Process Operations and Equipment Mr. E. D. Ilaines Chairman Manager - Machinery & EquipmentOperation Corporate Consulting Services Dr. N. J. IIawkins Consultant - Engineering Design & Technology Corporate Consulting Services Dr. R. D. Richards Manager - International Business Development International Operations Department Nuclear Energy Products Division Mr. V. R. Cooper (outside consultant)

Formerly Manager-Materials Engineering Lalmratory Corporate Itescarch and Development Dr. K. W. Norwood (outsido consultant)

Formerly Consultant-Engineering Operations and itesources Carlorate Consulting Services 15

Fluidized Solids Reactors and Solids Processing Dr. S. Seltzer Chairman Manager-Intermediates Manufacturing -

Silicone Products Business Department Dr. J. M. Dotson Manager-Chemical Engineering Vallecitos Nuclear Center Mr. M. J. Wynn Process Design Engineering Silicone Products Business Department Remote Maintenance and Operability

, Dr. R. H. Beaton Chairman

% Vice President & General Manager Energy Systems and Technology Division Power Generation Business Group Mr. W. K. MacCready (outside consultant)

Formerly General Manager-Chemical Processing Department Hanford Atomic Products Operation Mr. L. L. Zahn, Jr.

Manager-Fuel Recovery Product Operation Advanced Technology Department Quality Assurance and Safety Dr. M. C. Leverett Chairman Manager-Nuclear Safety Assurance Nuclear Energy Products Division Mr. M. L. Turner Manager-Quality Assurance Dreeder Reactor Operation Dr. C. W. Smith Manager-Licensing and Transportation Advanced Technology Department C. In Delth Technical Reviews Conducted and Evaluated From March 25 to April 30, the teams made in-depth technical reviews at the MFRP and the Vallecitos Laboratories of the following aspects of the problems in each of the major areas investigated:

1. Illstory of problems encountered in the past.

2. Current problems and identified but unsolved problems.

3. Anticipated problems, in early May,' 1974, a second meeting of the Advisory Doard was convened at which time the observations of the review teams were presented by the team leaders and discussed and evaluated by the group Conclusions were drawn concerning the technical capability of the pre-sent process and equipment in the MFitP, including an assessment of the plant's overall operating 16

capability. 'Ihese observations and conclusions are outlined in the next sections of this report. The report has been reviewed by, and its observations and conclusions endorsed by the Advisory Board and by those chairmen of the investigating teams who were not Advisory Ik>ard members.

1 17

III OBSERVATIONS AND CONCLUSIONS ON FUNDAMENTAL TECIINICAL PROBLEMS 3e following sections outline the fundamental technical problems of the Aqualluor Process.

The most fundamental problems of the Aquafluor Process and the MFRP derive from a process flowsheet which inherently requires final decontamination of the uranium to be effected by a series of calcination and fluorination processes carried out in a remote operation and maintenance mode in fluidized solids reactors and associated equipment.

Sections A and D outline the reasons why remote operation and maintenance are required in the final uranium conversion and purification (decontamination). Sections C and D describe the detailed nature of the operating problems.

A. Radioactivity in the Uranium Nitrate Process Stream Requires Remote Operation De radiation level of the uranium nitrate stream flowing to the uranium conversion and purification / load-out process (Figure III-A-1) prohibits contact operation and maintenance on these processes, all of which have had to be designed for remote operation.

De decontamination of the irradiated uranium and plutonium as determined by laboratory tests in the aqueous processes is shown in Table 1. Estimates of decontamination by the uranium fluoride purification system are based on work at other laboratories and are considered conservative, ne total amount of radioactivity, given in curies per metric ton of uranium (C1/MTU), is included for a typical load of spent fuel to provide an indication of the residual activity in the process streams. The number of curies has only an approximate relation to radiation level, since it indicates the rate of disintegration of unstable atoms (1 curie = 3.7x 1010 disintegrations /sec) but does not distinguish between the hard-to-shield neutron or gamma radiation and the less penetratirg beta or alpha radiation. it should be noted that although a separation of 99.99% of the radioactivity is indicated for the solvent extraction cycle, between 400 and 500 curiesper metric ton still remain in the intermediate product stream. Correspond-ing radiation levels are too high to permit contact operation and maintenance and necessitate canyon shielding.

B. Fission Product Deposition in the Uranium Conversion and Purification Systems Certain of the fission product fluorides present in the uranium nitrate stream can be expected to plate out on the inside walls of the equipment throughout the uranium fluoride purification system leading to levels of radiation which reinforce the need for remote operation and maintenance of this equipment.

ne uranium fluoride purification system has been widely studied in the AEC national laboratories and has been applied at the AEC Paducah plant for conversion of reprocessed uranium and removal of traces of fission products. Through this work, the understanding of the preparation and handling of UF 6 was well developed. De volatile fluoride system has been shown to be capable of very high decontamination factors.

ne problem elements to be removed from the UFO are the fission products, ruthenium (Ru), niobium (Nb), and technetium (Tc); and the transuranics. plutonium (Pu), and neptunium (Np). Separations techniques assure that the UF 6 pMuct win be satisfactorily decontaminated, 18

URANIUM CONVERSION AND PURIFICATION /l.OAD-OUT

.EE3 m Trees

1. %L rw 3 cb- _, . m.

_. ., o_.

i-0 'dai$ , ,q h6 ,,, ,,,_

r. . .

_2_ _ , ,gm g FIL TE R ' E413 '

5.r A, $ - .' I. 45

_g0p d L -- - f, } C

r_ a 53 ya

Dj i t I,_ n.., nw

.m := v _

~ *

.yg T N%. ""

,c...- -

dh. T ]f e-wur, . .g c. .,,,n

.._..a j ,Ag 7d 'P acW t s,,'A m , l' r.*? J,*c;/3 H  :. M=_T -

Ml

• ssa clo%

.<m 3 arstua T _

L sr I//i-(

p \

l [: . r- s r

_ .. .., ,, o irt . . g P L'T -F - - - - 0 So"4'?; *

i. J _y ffior,

@' T r

,t... :r

. 23 g

A-becla=>

<e W

+3 l t- . -!D Jb Een

< < s r. ( v 7" ""

l l w a ==rt ur. ,ut air ,or _

M,s

.-446 V I

o TABLEI EXPECTED DECONTAMINATION BY AQUAFLUOR PROCESS (For 24,000 Megawatt Days per Metric Ton - 160 Day Cooling)

Solvent Extraction Ion Exchange Volatile Fluoride Radioactivity Expected Residual Pu & Np Product U Product Element C1/MTU DF (a) Ci/MTU DF DF Total Fission Products 4.5 x 10 6

- 104 - 450 103 to 104 > 10 4

5 4 Zr 3.5 x 10 5 x 103(b) - 70 1 x 104(d) > 10 5 3 4 Nb 6. 6 x 10 5 x 10 (c) - 130 1 x 103(e) > 10 4 3 3 Ru 3.2 x 10 1 x 10 - 32 2-10 x 10 > 10 Tc 11 100 .1 U (5000 PPM)

Pu < (0.01 PPM)

Np < (O.01 PPM)

(a) Decontamination Factor 3 (b) Range of observed DFs was 0.2 to 5 x I 22 (c) Range of observed DFs was 1 to 50 4 x 10g with with higher the higher DF achieved DF achieved by Nb-H2by O2 Nb-H Otoaddition addition feed to feed.

(d) Range of observed DFs was 1/2 to 5 x 10 gith higher DF achieved by addition of oxalic acid in scrub of ion exchange (e) Range of observed DFs was 0.2 to 1.5 x 10 with higher DF achieved by addition of oxalic acid in scrub of ion exchange

+

i x

- _ '9

' \\

but residual radioactivity can be expected to build up in the system so that it must be maintained remotely.

'Ihe Pu and Ru compounds will be removed effectively by the sodium fluoride (NaF) sorption trap; the Np and Tc compounds will be removed by the magnesium fluoride sorption trap, and the residual Ru and Tc compounds will be removed in the distillation step. Buildup of radio-activity contamination can be expected to be most severe prior to the NaF sorption trap. Solid i fission product fluorides, like zirconium, can be expected to be entrained in the process gas and deposited in the piping prior to the NaF trap which will act as a filter. In addition, a substantial fraction of plutonium and ruthenium can be expected to deposit as nonvolatiles.

After the NaF trap, the principal activity will be due to Ru; because of the instability of its fluorides, Ru can be expected to be plated throughout the system up to the UF6 Vaporizer where final traces will be removed. This problem is not critical except as it may affect decontamination of process vessels for disposal.

C. Inoperable Vital Canyon Equipment

^

. The radioactivity from the process stream and fission product deposition discussed in the

., previous sections, III-A and B, results in a highly radioactive Uranium Conversion and i ~$ Purification Process (Figure III-C-1). For these reasons, it is essential that the system l

achieve very high reliability and that any problems which may arise can be dealt with through

} , remote maintenance and repair. Equipment testing and trial runs during the past two years I- have proved that the UNH calciner, solids transfer system and fluorinator cannot be operated and maintained remotely. A description of the design and operation of these units follows, with a summary of the many unsuccessful efforts that were made to solve the fundamental problems.

1. Critical Process and Equipment Description (Figure III-C-1)

a. UNH Calciner and Related Equipment (Figure III-C-2)

The uranium nitrate hexahydrate (UNH) raffinate solution from ion exchange (V332)

(see Pu, Np lon Exchange Load-out System, Figure I-D-1) is concentrated in the UNH concentrator (V401) to provide the feed solution for the UNH calciner (V404), which converts the UNH to UO3 . The UNH calciner employs steam as the principal fluidizing medium to limit the volumc of off-gas to be handled in the off-gas clean up system. A j concentrated solution (approximately 900 gm U/ liter) of UNH in 1 to 2 M HNO 3 is fed into the calciner by means of an air-operated atomizing nozzle. The solution employed has a freezing point of approximately 100 C and a boiling point of approximately 125 C.

Fluidized solids reactor UNH calciners have been operated employing feed concentrations ranging from 300 to 1600 gm U/ liter. The concentration employed is as high as is considered practical for remote operations, thereby permitting the maximum attainable processing rate for the calciner which is limited in size by nuclear criticality considerations. Originally the feed system employed an air-lift feed to a syphon-type atomizing nozzle. Problems in feed reliability ultimately were attributed to flashing in the feed nozzle caused by the elevated temperature and the combined reduced 21 m

d l

URANIUM CONVERSION AND PURIFICATION 1

FEED ppggagy l

I f Pr-"/"r-- -

E m3k h run y. +MO 1

e C)t,r =,..=

Au= =

= mo .

FL Nm j E4 3 eF x 1, + - . ..o,

a. Am rux n

vm

..W

_~j;g E A ----+)

i Y ,g,,,

m

" v..o, comETrLim o

=7  ! E-

• f'

=

__l c%lv'?a cENOE'"

?:N I

$_ catch 44u o vs- cm p.2o =

a a

w*-

yrg r:4 g (-

\ /

gL =

=

i t..,s e r L

M')

r2 + _.

+ r

-.v INATOR i

E-.is M ';- - --- >;0 r

=

=

i I' '

L.L. WASTE Mu 4

.-. 6 U CHEM AL v&utt v

MODIFIED CALCINER V-404 e5AG CUT I PRESG b FI L.TEFES WEED \M %l TT 7 T- TEMP II I '

1 r Fl.U)DITtMG l GAS \MLET A'l l l_Li> , 2 C- 2 O O.D.-+

\ h'- \ \ q\ \ "

)

UQg ADD ,

10"1.D. *-

TEMfD

- TEMP

- 1=RE.SS

- ~ " '

F"-

l CALCIMER FEED ' III "

IMLET ** MP l ll I l l TutEm HEETEE l l;j -khED MEA'TERS g l 7EMPH lll l l

TEMP l

IIl U 1

7ggp - ---

~

F L O) D \ 7_l M d>

CUTLET TO GAS A1E VE.YOE REMOTE CCUN 1E.47O W CUTLET h.1 O'd ~K L.E Figure III-C-2 23

pressure effee.s of the syphon feed nozzle and operation of the calciner at reduced pressure. Operation of the calciner at reduced internal pressure is considered an essential aspect of containing radioactivity within the shielded area. De calciner feed system is being modified to utilize a pressure pot system (V447) and a pressure-type atomizing nozzle is being considered to eliminate the feed nozz.le flashing problem.

The calciner is operated at approximately 3300C which produces adequately dry

- UO3 with suitable particle size control. Heat for the successive dehydration and denitration operations is provided by a combination of external and internal electric resistance heaters. The original design of the external heaters (35KW) employed a clam-shell design, utilizing nichrome ribbons and was intended to be remotely re-placeable. Remote replaceability proved unattainable in actual practice, and band-on resistance heaters were substituted to permit a controlled distribution of the heat flux and a higher heat flux in the zone immediately above the feed point.

De internal heaters originally installed were hairpin type units (56KW maximum) which were limited in height to the zone below the feed point to avoid potential cake formation on the heaters and consequent heater burnout, hese heaters proved quite susceptible to burnout due to the impracticality of providing adequate protection against localized high temperatures. He calciner was modified to permit the use of bayonet heaters designed to facilitate replacement of the heaters (by contact maintenance) in the ovent of heater failure. These heaters were designed for a max-imum power of approximately 70KW, but were wired to limit the applied power to about 35KW, thereby limiting the sheath temperature to a level approximately 300 F below the level at which damage to the sheath might occur. This operational limit was determined by means of thermocouples attached to three of the heaters to permit measurement of sheath temperature under operating conditions. Some of the revised internal heaters were extended above the feed zone to provide additional heat in this area. These heaters were located in regions that would not be in close contact with the feed spray.

Particle size control in the calcination process was the subject of considerable study at GE's Vallecitos Nuclear Center and, to the extent defined was found to be a complex relationship among feed UNH and HNO 3 concentrations, feed rate, and operating temperature. Individual droplets go through sequential steps of dehydration and denitration which result in droplet shrinkage, increasing stickiness, a subsequent porous state and finally a less friable mature partieto Bed temperature and droplet size affect the life of various particle stages relative to the free path of the droplet in the fluid bed. Particle size control is also affected by attrition within the bed and led to the installation of attrition jets within the bed to permit control of oversize particles, should the need arise. De calciner is equipped with multiple thermocouples both on the vessel wall ared projecting slightly into the bed to permit control of operating temperature and i:e give indication of poor fluidization should such problems arise. He bed diff9rential pressure measurement can also be used for assessing the quality of

24. 11uidipR,a.

he calciner is equipped with sintered metal filters in the enlarged disengaging section to prevent excessive carry-over of UO3to the offgas system (V403) and to assure an adequate supply of seed particles for controlled particle growth.

He steam fluidizing gas is preheated prior to entering the calciner by heat exchange from the disengaging section. His heat exchange is achieved by wrapping the inlet line around the disengaging section for several turns and metal bonding the pipe to the disengaging section wall. ne resultant steam temperature entering the fluidized solids bed is approximately 130 C.

De fluidizing gas inlet enters the calciner by means of an external ring manifold feeding four orificed lines to the lower part of the bottom cone. he airveyor air returns to the UO3 draw-off line immediately below the bottom cone. Initial operation of the fluid bed is achieved by adding a seed bed of commercial UO .

3 Operation of the UO3 fluid bed is complicated by the sticky phase through which the particles pass and by the tendency of UO 3 to form hydrates at lower temperatures and as the availability of liquid water increases. Reasonable control of the feed atomization and the feed rate, relative to the heating rate are required to prevent cake formation due to wetting of the particles at lower than normal temperatures. Dis-solution of the bed cakes can be accomplished with nitric acid, but is relatively slow because of the limited contacting efficiency. Upon hydration, UO expands with the 3

result that any hydration occuring in confined spaces produces extremely dense cakes.

This behavior was encountered in pressure sensing and chemical addition lines due to the pressure fluctuation within the calciner and the moisture content of the purge gas.

De problem was mitigated by heating the lines for a considerable distance back from the calciner body.

De general behavior of UO3particles in the presence of moisture complicates the operation of the calciner and all of the solids handling equipment. For smooth oper-ation of fluidized solids reactors, the powder must flow as if it were a fluid. Materials that are sticky or that tend to pack exhibit unacceptable flow properties. The fact that moisture, in the form of steam, is present in the calciner requires that all surfaces that might come in contact with the powder must be kept over 200 C and all purge gases for pressure taps must be kept as dry as possible. Upsets in operations in which the steam flows into the instrument sensors will often cause plugging.

! b. Solids Transfer System (Figures III-C-3 and III-C-4)

The UO3 f rmed in the UNH calciner is conveyed continuously to the fluorinator j where it is reacted with elemental fluorine. De transport system must provide a reasonably controlled and predictable rate of UO 3 transfer and must provide for isolation of the incompatible atmospheres in the calciner (steam) and the fluorinator (fluorine, UF ).

6 The original design provided a variable speed screw conveyor (ME404) for the UO 3 transport (Figure III-C-3). Vallecitos tests had confirmed the adequacy of rate control and that a modest buffer pressure of air would permit effective isolation of the calciner and fluorinator atmospheres. He unit was designed for 25

SCREW CONVEYOR ME 404-1 m 14 ' - 2 "  :-

OUTLET NOZZ LE

- i > - -

.g i REMOTE CONNECTOR NonLE

-- DRWE SHAFT 2 &* WAt.t a .

• o- .;

esi:=2 I m CR1VE MOTOR FLulDlZlNG AIR LINES FROM V404 9 / CALCINER I

FLUOR 1MATOR ~

/

,,j . ..

~ .

I 3"SCH 40 PIPE -

l l4*-4 1~

i i Figure Ill-C-3 26

UO 3 AIRVEYOR ELECTRICALLY HEATED $ MONITORED "

AIR ACTUATED VALVES $ DELUMPER AIRVEYOR OPERATION :

@ u

~~

UO3 IS PNEUMATICALLY TRANSFERRED TO A COLLECTOR ,

PIPE. WHEN COLLECTOR FILLS,UO 15 RECIRCULATED - 74 TO CALCINER. COLLECTOR DISCH ES TO BATCH PIPE WHICH DISCHARGES TO FLUORINATOR ON AN @ ,

ADJUSTABLE TIME CYCLE.

b 14'- O '3/ 6 h 93 G e W @

@-8 \g3 ll ,

9 y

@ 2"CALCINER CONNECTOR

~ , @ 3"DELUMPER

@ 3" BALL VALVE h ' ,,

@ CLEANOUT PORT

@ AIR SUPPLY

[2 3"

/4 @ RETURN TO CALCINER y p g "j @ 4' COLLECTOR PIPE (0.16 L1TERS) a~ @l' BALL VALVE

@ 4' BATCH P!PE (3.6 LITERS)

@ 2"FLUORINATOR CONNECTOR

@ PRESSURE TAP

@ ELECTRICAL SUPPLY

@ JET

@ PURGE AIR Figure III-C-4 27

rensote emplacement under the calciner utilizing a spring-loaded connection to effect a seal. In practice, this seal proved unsatisfactory and required redesign. In addition, the screw conveyor did not remain at temperatures high enough (> 200 C) to prevent hydrate formation. As a result UO 3hydrate was formed in the confines of the screw conveyor rendering the unit inoperable. While redesign of the screw conveyor to achieve a satisfactory seal and adequately high temperatures was possible, the extreme difficulty in removing the dense UO3 hydrate cake, once formed, dictated an alternative approach.

A dense phase air-transport system (UO 3 airvey r - Figure III-C-4) was designed to replace the screw conveyor. nis airveyor was connected to the calciner by means of a modified Hanford-type connector head which provided an adequate seal at the bottom of the calciner. A commercial grinder ns incorporated in the feed line to the airveyor to prevent the entrance of particles or cake fragments too large to be con-veyed. A ball valve was also provided to prevent " choked" feed to the airveyor. An air-operated jet was employed at the feed point of the airveyor with the UO "E 3

lifted approximately 14 feet to a series of collection chambers discharging to the fluorinator. Transport air, excess UO and elutriated UO fines are returned to the 3 3 UNII calciner and form a portion of the fluidizing gas. UO3 is collected in the upper chamber which will hold approximately 40 Kgs of UO 3

. Small increments (15 Kg) of UO3 are drained to the lower chamber and then subsequently discharged into the fluorinator. no fluorocarbon-lined plug valves isolating the chambers are interlocked to prevent them from being opened simultaneously. Dry air purges at the bottom of each chamber facilitate UO3 discharge and also provide a buffer pressure to assure isolation of the systems. De airveyor is instrumented and electrically heated to prevent hydrate formation. In addition, a clean-out port is provided below the jet to facilitate clean-out of any oversize materialin that region. Operation of the airveyor can be monitored by means of pressure taps on the airveyor and by measure-ment of the calciner bed level.

c. Fluorinator and Related Equipment (Fluorinator V407 - Figure III-C-5)

The UO 3 is converted to UF6 by reaction with elemental fluorine. De reaction takes place at elevated temperature (>425 C) and is highly exothermic, proceeding directly to UF 6without formation of identifiable intermediates. A fluidized solids reactor is most suitable forproviding the reaction time required, and to permit dissipation of the heat of reaction. The reactor is of monel construction (5/8 inch thick) which is protected from reaction with fluorine by the formation of a tightly adherent fluoride film. At sufficiently elevated temperature, the monel can react destructively with fluorine, for example burning a hole through the vessel wall.

The initial process flowsheet called for use of a granular fused alumina bed from V406 to act as a diluent for fluorination reaction. Parametric studies were made to define the limits of UO3 and fluorine concentration to prevent overheating and sintering of the bed. Since the UO3particles are about twice as dense as the fused 28

FLUORINATOR V407 OFFGAS TO FILT ER n

O OUTLET n -

% [OFFGAS \

CYCLONES V4075 y03 2 5'O. D. == INLET REMOTE OPERATING I

U VALVE FLulD121NG d C

3 G AS INLET U c-RETURN ,- c =

=1( 37['O. D.

._)

25'- l j,. V407 3

\ / A FLUIDizlNG GAS ,,

TU S TEM P.

~

~== ~ ~ ] y 19[0.D. '

HEATER,  ;- '

l i

TEMR N L ---- U PRESS. _ _

PRESS. "

l T EM P.

WAST E Figure III-C-5 29

alumina, the potential for bed segregation with resultant high UO3concentration near the fluidizing gas inlet was of concern. Later development studies showed that the use of more dilute fluorine (20% vs. 40% by volume of fluorine) would permit satisfactory operation regardless of UO 3 e ncentration. Elimination of the diluent alumina pre-vents formation of potentially troublesome sinters but the increased bed density from l l

use of a pure UO3 bed Will require upgrading of the fluidizing gas recycle compressor l through substantial additional development work in order to achieve satisfactory oper- l ation.

With the present interim test flowsheet, UO3 is admitted to the fluorinator (V407) in increments of approximately 15 Kgs where it reacts with the fluidizing gas consisting of approximately 20% fluorine in air. Heat is initially provided by means of an external furnace to raise the bed temperature to >425 C. Subsequent to initiation of reaction, the heat of reaction is removed by a steam-water fog passing through coils metal-bonded to the lower cylindrical section of the fluorinator. The reaction gas passes through an enlarged disengaging section where a cyclone (V407S) serves to remove any entrained unreacted UO 3 r other particulate contaminants. The collected powder is periodically recycled to the fluidized zone of the fluorinator.

He reaction gas subsequently passes through an off-gas clean-up system (V410 and F411) and then to the primary cold traps (E412, 413, 414) where 95% of the UF 6

is removed. A small portion of the primary cold trap off-gas is bled off to the sec-ondary cold trap (E415, 416) to maintain the system pressure balance and to remove the oxygen formed in the reaction. He bulk of the gas stream passes to the suction of the recycle compressor (C420) where it is combined with pure elemental fluorine from the fluorine plant and returned to the fluorinator as the fluidizing gas. ne fhaorine concentration in the fluidizing gas is continuously monitored in the discharge line from the compressoc. ne rate of production of fluorine provides a secondary check on the fluorine concentration. De fludizing gas is preheated by heat exchange with the disengaging section of the fluorinator by means of a helical coil metal-bonded to the disengaging section. The resultant temperature of the fluidizing gas entering the j fluorinator is approximately 170 C.

A bottom outlet on the fluorinator permits periodic discharge of any accumulated solids (primarily iron fluoride from the iron added upstream as a chemical reagent and nonvolatile fission product fluorides) to the dry chemical waste vault (V446). To prevent this stagnant section of line from filling with UO3 , a limited volume of nickel particles are added to the fluorinator after each waste discharge.

The fluorinator is equipped with multiple temperature elements and bed differential pressure taps to permit assessment of the quality of fluidization.

The fluidizing gas inlet to the fluorinator is presently a single point entry in the waste discharge line below the bottom cone. Since the bed is not heated below the cylindrical section, this section is nonreactive, but contributes to overall system pres-sure drop. Development tests are underway to provide the basis for a more efficient gas distribution system.

30

2. Problem History

a. UNH Calciner Feed.

In 1966, with the start-up of a 2 in, fluidized solids reactor pilot plant at Vallecitos, the problems associated with the feed to the calciner were identified. A particle size distribution which would result in a fluidizable bed requires a specific combination of-solution concentration, nozzle configuration and heat flux. A mismatch results in a bed of very fine particles which will not fluidize properly, or in a caked bed. Dese problems were studied in some detail through 1966 and 1967 and an optimum feed composition and temperature range were identified in which the feed nozzle would not plug and the particle size distribution in the bed would be satisfactory for fluid-ization. From 1967 to 1970 the effect of various impurities and the parameters for control of particle size were studied in great detail. Because the feed nozzle con-figuration and feed composition proved to be critical, a 4 in. fluidized solids reactor was built to test scale-up. Tests on this equipment continued through 1973, with efforts concentrated on producing the proper particle size distribution in the fluidized solids reactor by control of feed composition, temperature, pressure and nozzle configuration.

From the first run at MFRP with UNH solution until February 1974 many attempts were made to operate the calciner gravity feed system with little success. The high viscosity of the feed solution and the narrow temperature range in which the feed had to be held caused plugging in the feed line. In some cases the nozzle could be cleared remotely, but often the line had to be removed and cleared manually. Dere were many problems in remotely assembling the feed nozzle, including proper seating to prevent powder leakage. The feed jumper has been removed so frequently in the past year that there is serious concern that the seats have been permanently damaged, making good remote seating almost impossible without redesign.

In February 1974 the feed system was completely revised to a pressure pot system with some success, but other problems persist, particularly nozzle plugging.

At the present time, the best available solution to the problem is to change the feed composition (lower UNH concentration and higher nitric acid concentration). But this would require a major change in the UNH calciner vessel to increase the heat flux to evaporate the additional water. It would also require some changes in the nitric acid balance in the plant and a line to add nitric acid to the concentrated UNH feed.

No satisfactory solution has been developed for the UNH calciner feed problem.

Work continues both at Vallecitos and MFRP to develop alternate solutions which would minimize flow sheet and equipment changes centering around nozzle performance,

b. Bed Caking and Calrod Burnout Early in the development history of the UNH calciner it became obvious that improper feed concentrations, feed rates, temperature and pressures could cause caking of the calciner bed. While attempting to define the operating parameters for producing particles of the proper size Vallecitos was also defining the ranges in 31

l

[

j which caking would occur. By September 1969 they had defined the proper operating conditions in the 2 in, fluidized solids reactor, but between January and June 1971,

- they experienced many caking problems in the 4 in. reactor.

In general, caking could be tolerated because the cake could be dissolved out in nitric acid. However, this reduces time operating efficiency (T.O.E.) and the frequency cannot be high. In initial plant trials caking was frequent because nozzle plugging caused many upsets in operations. By June 1973 13 of the 15 calrods had failed because of overheating, possibly due to caking on the heaters. In February 1974 the calrod units were replaced with bayonet-type heaters less susceptible to overheating and possibly replaceable by contact maintenance. By April 1974 two of the new calrods had failed, but could not be removed, even with contact maintenance.

Present alternate solutions to theproblem fall into three categories:

1. Changes in the heat source: in-bed combustion, radio frequency, heat, or heat transfer fluids. In-bed combustion would require major flow sheet changes.

2. Changes in calrod design to prevent overheating when the bed is caked -- a major development project.

3. Improvements in operating controls to prevent any bed caking. The pressure feed system has helped because the reduction in frequency of feed nozzle plugging has reduced the number of upsets which cause bed caking. However, it must be recognized that pressure operation complicates containment of radioactivity.

c. Solids Transfer from Calciner to Fluorinator The canyon design and construction is such that the solid granular UO I#U" 3

bottom of the UNH calciner must be conveyed into the top of the fluorinator. The original screw conveyor intended for this function could not be kept from plugging because of the physical properties of the UO 3 particles.

By September 1973, after months of development work at Vallecitos, a new type of conveyor, using a continuous flow of fluidizing air was installed in the plant. His airveyor system had worked well in the laboratory, but has plugged in plant trials and continued in-plant testing has not yet achieved operability objectives.

In May 1974 a pressure upset in the fluorinator caused fluorine to leak past the airveyor block valves and into the UNH calciner, producing corrosion damage through the system. Even the limited 9 months of testing had abraded the seats of the block valves.

The present design of the airveyor jumper is such that the plant has been unable to remotely reconnect the airveyor to the calciner and fluorinator. A redesign is necessary, since remote replacement of this jumper is essential for removal of the UNH calciner and the fluorinator, as well as the airveyor itself.

d. Erosfon and Plugging Almost every test run made in the plant has had some equipment or sensor line failure due to plugging or erosion or has required hands-on operation to continue the run. He monthly operating reports record the following incidents:

32

4 .

UNII Calciner - Airveyor November 1971 - Damaged sampler nozzle December 1972 -

Plugged instrument lines

- Screw conveyor plugged January 1973 - Calciner and conveyor plugged with material of consistency of concrete February 1973 - Plugged instrument and solids sampler lines had to be cut open and drilled out; heaters and insulation installed on instrument and transfer lines March 1973 - Solids sampler line modified following unsuccessful efforts to unplug instrument lines plugged because of hydrate formation I April 1973 - Run terminated when a hole eroded through the cone section of the calciner May 1973 - Manual rodding required through drain valve _

to start flow of solids June 1973 -

Plugs formed inside calciner at feed port required manual rodding July 1973 - Drain line rodding resulted in dump of half of bed to cell floor --

^

August 1973 -

Airveyor plugged whe:n calciner fluidizing medium was switched from air to steam September 1973 - Solids sampler modified again to prevent plugging October 1973 -

Airveyor plugging mechanism investigation showed bridging and packing of powder upstream of transfer jet November 1073 -

In-plant testing of airveyor suspended because of bridging and packing. Design modifications required January 1974 -

Airveyor dump valve modified to prevent plugging above the valve and in the instrument taps (which required entrance to cell to unplug)

Fluorinator May 1973 -

Plugging of UO3 addition lines l

February 1974 - Plug in drain of offgas filter line; special tool devised for remote line unplugging In almost every one of the incidents listed, a repair would not have been possible in the remote maintenance mode. Of equal significance is the fact that the calciner's longest run before a plugging or caking failure has been only 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br />.

33

e. Other Problems Even with the continuing process and equipment development program carried out at the Vallecitos Laboratories, it was anticipated that a multiplicity of problems would remain to be discovered on full-scale equipment at the MFRP prior to the introduction of irradiated fuel. It was only after considerable in-plant testing, starting t -

in September 1972 with cold uranium, modifications, redesigns and retesting that the fundamental technical problems discussed above were singled out. Many other prob-lems were identified, and either have already been solved, or are judged to be l

amenable to solutions. The following are examples of such problems relating to the three pie:ces of equipment considered above:

The UNH Calciner ne UO Conveyor 3

ne Fluorination and UF6Ilandling Systems UNH Calciner Failed magnetic flowmeter Excessive fine particles made in the bed Improper nozzle seating Acid leaks on the feed jumper High failure frequency of traps in process steam system Dump valve failure Uns,atisfactory particle size control UO3 Conveyor Screw System Unsatisfactory seal system Necessary replacement of driveshaft Unsuccessful starting Alignment problems Binding of floating screw Airveyor Difficult and unpredictable transfer of very fine powder Binding ball valve after limited operation Increasing difficulty of sampling UO3 as particle size decreases Fluorination and UF IIandling 6

Insufficient space heating capacity Cold trap refrigeration difficulties Fluorine sampling difficulties High pressure drop across secondary cold traps Oil in line from F 2compressor to fluorinator 34 i

liigher compressor pressure required because of change in bed Broken compressor shaft l

L Inadequate compressor clearance Bed sintering Inadequate reliability of the refrigeration units Ii Unsatisfactory performance of valves in the primary UF6 cold traps

}

Temperature excursion resulting in hole burnout in piping junction 1 between fluorine gas inlet and bed drain line.

l. D. Plant Configuration and Close Coupling of Process Equipment The processing systems were compactly designed to improve product recoveries and enrich-l ment segregations and to reduce inventories to facilitate accountability. It was assumed that

_ there would be need for processing a specific batch of a customer's fuel and returning the uranium, plutonium and neptunium in that specific batch of fuel to the customer. It was f realized that this would necessitate a cleanout of the plam between contracts and a separate accounting for each order. Fissionable material accountability and nuclear safety design criteria were thought to be easier to control by reducing retention time and line sizes between steps in the process. nrough compact design, investment costs were expected to be lower, resulting in reduced reprocessing costs.

To achieve these concepts, the MFRP was designed as a closely coupled operation with little or no storage capacity between the various unit operations and subsystems, which complicates the recycling and reworking of cif specification process streams. Unit testing to date has indicated that several subsyrtems with many pieces of equipment are too closely coupled. For example, the solvent extraction system with its four pieces of major equipment will normally be marginal in performance until equilibrium conditions are established by recycling the output until it meets specifications. During the recycle period, there can be no feed to the next step in the operation so the remainder of the pl ut wol be down until the solvent extraction system comes to equilibrium.

The ion exchange systems with six pieces of major equipment separate the plutonium, neptunium and uranium into individual streams for further processing. The present systems are thought to be marginal with respect to their ability to accept feed surges which would result in off-specification products and would choke the system during a normal run.

When the feed starts to the uranium conversion and purification system, which has about 20 pieces of major equipment, there will be a period of adjustment until each piece of equip-ment in that system comes to equilibrium and produces a within-specification product. Mean-while, off-specification material will have to be recycled and no real production will be made.

This same condition of interdependence of one system on another is true for the other s' stems in this plant. By design, there is a minimum of intermediate storage capacity. He equipment in the process main flow stream is in series from fuel bundle to products -- uranium, plutonium, and neptunium. This main stream is also directly affected by any malfunctions of the supportint, systems such as Fluorine Generation, High Activity Waste and Low Activity Waste.

35

Perturbations in steady state flow caused by equipment failure in one part of the plant will be reflected through other parts of the plant. Insufficient buffer storage between the present units of equipment will necessitate shutting down these units when the feed supply is exhausted, which necessitates a restart when it returns. Out-of-specification material can be expected to be generated during these shutting down and starting up periods.

'Ihe failure rate of equipment experienced during testing in the past two years indicates that the time operating efficiency of the present plant will be extremely low because of the anticipated small percentage of the time when all equipment would be simultaneously operating correctly and because of the large amount of productive time that will be wasted on recycle of off-standard product streams between the major systems.

Of several subsystems, the solvent extraction cycle alone may take up to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to stabilize; hence, as subsystems must be started sequentially, any start-ups disturbing the solvent extraction system may at times extend stabilization periods for the whole plant to several times 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

On the other hand, an idea of the time that the whole system can operate between forced outages is indicated by the fact that the two longest runs on the UNH calciner to date have been 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> each -- and this is only one of many series elements of the process, even though in all probability the most difficult to operate reliably. It thus appears that the time required to stabilize the process and obtain useful output may well exceed the mean time to failure. If this should be the case, it would be difficult to be able to run long enou$ to obtain some output, and the Time Operating Efficiency (T.O.E.) would be close to zero.

36

IV OUTLOOK FOR ACCEPTABLE SOLUTION OF MAJOR PROBLEMS Many of the problems of plugging, erosion, and caking normally encountered in fluidized solids reactors were anticipated, and special equipment designs were developed through which it was ex-pected to work out these problems during the plant start-up period. Plant test results to date have indicated that all such problems are intensified by the high density and difficult flow properties, caking properties, and plugging properties of uranium nitrate, uranium oxide and their partially hydrated forms. For example:

A. UNil Calciner Feed Plugging. The narrow range of feed concentration which is suitable for the particle growth required in the calciner results in a narrow window of operating conditions (temperature concentrations and pressures) at which a liquid phase can be maintained in the feed nozzle. Excursions in temperature, pressure, concentration, feed rates, nozzle alignment and effective diameter will cause plugging of the nozzle or caking of the bed. Even if the operating window is enlarged, frequent shutdowns for transient upsets can be expected. Even after much effort to effect their control, it has been found impossible to hold these variables within the ranges necessary for satisfactory operation. Numerous solutions have been investigated for these problems, but none has been found satisfactory. For example, the remote nozzle clean out device provides only a partially satisfactory solution to the problem. When this fails, a plant shutdown and remote jumper removalis required. While this will not result in serious failures and vessel replacements, it will continue to be a major cause of low Time Operating Efficiency (T.O.E.) under remote operating conditions.

B. UNil Calciner Bed Caking. The major causes of bed caking must be identified and corrected by the time the plant is committed to hot operation. The operating history indicates, however, tr.at the bed is susceptible to caking and can be expected to cake periodically from unanticipated plant excursions and perturbations. This will lead to localized erosion and vessel wall failure and to calrod overheating and burnout. The frequency may decrease as the plant gains experience, but failures will continue to occur in random fashion. In the remote maintenance mode, these fail-ures will require vessel replacement. Adequate capability for decontamination and storage of failed equipment at the now anticipated failure rates is not presently available and it would not be practical to provide it.

C. Operability of Instrument Sensors and Sample Taps. Reliable measurements of bed height are especially important to sustain fluidized solids reactor operation. The measurement depends on clea:ced instrument sensor lines and sample taps, and plugging and failures will occur several times a year. Redundancy can relieve but not eliminate these problems. Alert operators will usually recognize the condition and take corrective action. Occasionally, however, the failure cannot be discerned and the bed height will get out of control. In the UNH calciner, a very low bed will overheat the calrods; a very high bed will overload the filters or even destroy them. In the fluorinator, a very low bed height could cause high temperatures and a fluorine fire; a very high bed height will overload the cyclone and perhaps erode a hole in it. Each of these failures could require vessel replacement in the remote maintenance mode.

D. Erosion of Internals. Fluidized solids reactor operation has shown that generalized erosion of j all of the internals in the fluidized solids bed of the reactor can be a problem. The rate is a 37

function of particle shape and hardness, wall hardness, average gas velocity and configuration

of the internals. Points at which the gas or the bed must change direction of flow are particu-larly subject to erosion since transient eddy currents tend to cause rapid crosion. For example, such crosion failures may be caused by a particular pattern of bed caking which results in a severe change in the local flow pattern. Mechanical problems must be expected in the operation of fluidized solids reactors with complex internal equipment of the type in the UNH calciner.

In similar types of industrial equipment, with complex internal heat transfer and filtration equipment, a two-month run without a shutdown for mechanical reasons is extremely rare.

Frequently, these problems develop because a transient excursion from the normal operation occurs without warning and a complex series of events causes the breakdown. Experience in fluidized reactor calcination at MFRP has shown that all of these types of problems have been worse than originally anticipated. Often, it is necessary to get at some of the internals of the vessel to make repairs. Most such repairs readily accomplished in a non-radioactive system consume extensive time and may be impossible of accomplishment in the presence of radiation, thereby requiring complete equipment replacement.

E. grveyor Failure. Under laboratory conditions, the airveyor worked well. Under actual oper-ation conditions, it plugged shortly after the first test run. After limited service, the isolation valves also leaked because of erosion. The same types of plugging and erosion problems en-countered in the fluidized solids UNH calciner also critically affect the operation of the airveyor.

The equipment presently installed is only temporary, as it cannot be replaced remotely, thus making it also impossible to replace remotely the UNH calciner and the fluormator to which it is connected. The necessary redesign to make it remotable may place constraints on desired changes to minimize plugging and erosion problems. The only obvious solution to these prob-lems is to change the flow sheet to permit installation of the UNH calciner and fluorinator out-side the high radiation areas.

F. Noncontinuous Operation. Continuous processes run well at steady state, but run poorly during start-up and shutdown, when the many combinations of unusual operating conditions frequently produce out-of-specification material, requiring rework. These conditions are particularly true of fluidized solids reactor processes. The worst events can be expected to occur during start-up or shutdown, subjecting the vessels, jumpers, and instruments to unusual transient conditions. Many of the failures will be a result of these transients. The experience of the last three years has shown that such transient conditions can be expected to be particularly troublesome in the present plant which, as pointed out in Section III.D., consists of closely coupled process equipment with very little capability for recycling and reworking the off-speci-fication material to be expected during start-up and shutdown.

The above examples demonstrate that there is a high probability in regular plant operations of failure of the UNH calciner and fluorinator fluidized solids reactors -- especially during start-up and shutdown. These are failures which could be corrected in a normal hands-on operating mode, but which can be expected to result in excessive downtime of equipment and process operation where corrections can only be accomplished through remote operations, as will be required in radioactivity hot operations.

38

Furthermore, it has been demonstrated that many such failures cannot be corrected in the re-mote operating mode, with the only possibility for resumption of operations requiring complete re-moval of the failed equipment and substitution of a spare, which spare, in general, would be subject to the same failure mode. Even with long design and development programs, it is difficult to see -

any reasonably satisfactory solutions to many of these problems which would be compatible with the.

constraint of remote operation and maintenance of such equipment over the life of the plant.

While some of the above problems also exist in the Ifigh Activity Waste calciner, there are mitigating factors which tend to reduce their severity:

1. - Seed material controls the depth and particle distribution of the fluid bed, reducing the caking and erosion problems.- l l

2. A much wider concentration range of feed materials is tolerable.

3. A much wider operating range of temperature and pressure swings is tolerable.

4. There are no internal heaters in the calciner.

5. There is a minimum likelihood of exposure to fluorine.

6. The calciner has been operated for over 1500 hours0.0174 days <br />0.417 hours <br />0.00248 weeks <br />5.7075e-4 months <br /> and most of the problems encountered have been solved.

Thus the likelihood of frequent serious failures of this system is low. Ilowever, the vessel will fall and should be redesigned to facilitate removal and replacement.

(

h 1

39

V APPRAISAL OF OVERALL PLANT OPERATING CAPABILITY The plant cannot be expected to achieve a reasonable production output relative to its original design rating of 300 metric tons U/ year with its present process flowsheet in its present configura-tion.

In making an appraisal of the overall operating capability of MFRP, an attempt has been made to distinguish between:

e The multitude of equipment and process flowsheet problems which have given trouble up to this point, but which might possibly be expected to be corrected with further development work; and e The fundamental problems which in all probability cannot be corrected in any way to insure a plant production capability of reasonable level and continuity within the constraints of the present process flowsheet and plant configuration. These latter barrier type problems have been discussed in Sections III and IV.

Even with all contemplated improvements in the fluidized solids reactors and associated equip-ment used to calcine uranium nitrate to uranium oxide and to fluorinate the latter to uranium hexa-fluoride, it is the considered technical judgment of the experienced technical experts studying the history and outlook for these systems that such systems will continue to be unreliable. Following radioactively hot operation with the constraints of remote operation and maintenance, it is judged that the expected plant downtime and forced outage levels resulting from the barrier type problems with these systems alone would:

e Limit the first year plant throughout to 10-15 metric tons, e Limit subsequent years' throughout to 50-100 metric tons.

From the standpoint of overall integrated plant operation, as discussed in III. D. , there is a substantial risk that in radioactively hot operation, the required time for stabilization of plant process streams may well be of the same order as the mean time to failure, resulting in a time operating efficiency which would limit ultimate plant output to a level substantially less than the 50-100 metric tons estimated.

There is also a serious problem regarding satisfactory storage and disposition of the excessive quantities of radioactively contaminated failed equipment which can be aniticpated from the present process.

Finally, despite best efforts with the present plant configuration, there is significant risk that the plant would suffer a disabling failure during the early years of operation from which it could not recover without a shutdown of extended duration --perhaps measured in years.

It is therefore concluded that the MFRP, with its present flowsheet and present plant configura-tion, should not be committed to radioactively hot operation.

Quality Assurance and Safety The subject of quality assurance and safety was reviewed in detail. It appears that substantial additional work remains to be done in the area of quality assurance, llowever, this finding is not of material importance in view of the conclusion, stated above, regarding the operability of the MFRP.

40

. It is not believed that the plant design presents any significant problems of safety with respect to the off-site public. Ilowever, problems of operability interact with considerations of on-site per-sonnel safety. Process and equipment breakdowns could occur which would not be correctible with-out very protracted shutdowns, consistent with observance of personnel safety requirements.

- Prior to commencing commercial operation of the fuel reprocessing plant, the probability of equipment failure must be reduced to a level where the potential radiation exposure to the operators due to the accidental spread of contamination or associated repair procedures will be within accept-able limits. Additionally the number of types of failures must be reduced to the point where a thorough safety evaluation of each possible type of failure can be made. Only in this fashion can it be assured that no significant safety hazard will be involved.

41 l

VI ALTERNATIVE PROCESS FLOWSIIEET AND PLANT CONFIGURATION As stated in the previous sections, the basic problems of the present plant stem from the fact that final decontamination of the uranium and its conversion to uranium hexafluoride must be accomplished in a series of innovative process steps involving fluidized solids reactors and solids handling technologies which have proven unexpectedly difficult to adapt to the remote operating and maintenance mode required in radioactively hot operations. Continuing the present start-up type effort offers no realistic prospects of achieving sustained plant operation.

The only technically feasible solution to this problem in the judgment of the technical experts is in the development of a new process flowsheet and new plant configuration. The new process flow-sheet would require the addition of a second cycle of solvent extraction decontamination to reduce the radiation level of the uranium nitrate stream leaving the ion exchange process in the canyon to a level sufficiently low to permit contact operation, maintenance and repair on the uranium conversion and purification processes (Figure III. C. 4). One such possible two-cycle solvent extraction flow-sheet is shown in Figure VI-1. The plant configuration would be changed by removing the uranium conversion and purification processes from the present canyon and installing them in a new and separate building. Such changes could provide space in the present canyon for the necessary second solvent extraction cycle, and could also permit the introduction of some additional equipment for surge capacity to improve recycle and rework capability.

While it is believed that such extensive changes to the process flowsheet and plant configuration could result in a new plant with an acceptable time operating efficiency, it is recognized that they represent a complete departure from the original approach. The new flowsheet would have to be

! established and demonstrated. A preliminary estimate has indicated that a minimum of four years would be required for the engineering. construction and start-up of such a new plant, with no allowance for contested regulatory proceedings; and that the costs would be in the range of $90-130 million.

42

jl jj il1 l jIlil l 1 d

e r

P p

p N s

e c H >

ie N C -

C U O N t

n oC T

e v 3 O2 _' 0 -

lo N 3 5 S 3 o X H -

E U s T 2 MW O E" M W a D

1

0. L F 24.O0F L

f 0

E 0  : * "

u P 0 WTB 4,

• 2 E

2%o w

1 T 3 o) 0 1

0t 3F E

C E

NJD LA g

p o

r Ff ; ui,O-X N

u, p g

y MW L

A 7 oF V p X '

0 _' ,

2 0

2 e X

l P =

T _ "

E _'

E *  %

H 0 0 3 _'

S $ O 2 N - F uq f 3 0

0 2 HW O

0 2 " ,H O PU1 3 P

nun %-

W fu L F 0 2

H Mg%5W p

2g 10 0L

- O

- F 3

O PHINV)V)

I U -

0 2

O L

p e

F

+C X

M (u(p 7 PN 0F M O 1L W 1 s

F le WN C U CN

_' A O C C HC H C P

t n

e w

R S to 3

O0 N 5 0

5 F o H 3

- b U s 5

T Mw U MW M

d r 1 C 0 L0 H 4. O 0F L

0 F D 0 - c n e  :

'[

E P0 B4 * -

I F

I CWTW H$O0L 1  ;

~ p g80 g

P NC 2O C

D 3F

'O]

O M e l

l g

p g

l 3.PU0 0 PN 2 l

~ s

)

1

_ " w g $

_ 31 u90-l990l o l g B

SWo H Me L  ; F l

g y 0 P

B4 0 g 0 X,T A,

H.% B LW oO

- Al F

_  : 3 C

O N P. WN ToAO

~

_ ~ H .

HC M F 3

3 O 2 N

H U -W L I

M O 4 M t. L 0 3 0 7 OF O

30 5 N -

1 w W mHM o M

O L

l R 3 F

[c5 $/

e ll Illl I lI lll l

(

h I'

BIOGRAPHICAL INFORMATION - ADVISORY BOARD Dr. R. H. Beaton An alumnus of Northeastern University (BS,1939; Hon.D.Sc.,1967) and Yale University (D.Eng., .

1942), Dr. Beaton worked on the Manhattan Project during World War II. He joined General Electric at Hanford in 1946, where he was in charge of a variety of atomic design and construction programs.

' He was named General Manager of the X-Ray Department's 908 Product Section in 1957; of Spacecraft

. Department in 1963; of Apollo Systems Department in 1964 and was made Vice President and General Manager of the Electronic Systems Products Division in 1968. In May 1974 he was named General Manager of the newly-created Energy Systems and Technology Division in the Power Generation Group.

. Mr. D. E. Debacher

'After graduating from Rennsselaer Polytechnic Institute in 1951 with a degree in chemical engineering,

. Mr. Debacher worked for the Ethyl Corporation before joining GE in 1955 as supervisor of the Pittsfield polycarbonate pilot operation. in 1962 he transferred to Mt. Vernon, Indiana and held various positions in that plant, becoming Plant Manager in 1968, and Manager of the Lexan Products Section in 1970. In

1973 he was promoted to General Manager of the Silicone Products Department.

Dr. W. H.' Reas A graduate of the University of California, Berkeley (BS, Chemistry in 1943 and PhD., Chemistry in 1948),' Dr. Reas joined GE in llanford, starting as a Research Chemist. He was made Head of Product

Mttallurgy, Head of Plutonium Conversion Process Development, Manager of Chemical Research, and

. Manat..r of the Chemical Laboratory. In 1964 he was made Manager-Technical Development in the Fuel Recovery Operation, Nuclear Encrgy Division, and in 1967 he became Manager-Nuclear Processes Development in the Boiling Water Reactor Systems Department.

Dr. C. E. Reed, Chairman Dr. Reed holds chemical engineering degrees from Case Institute of Technology (BS,1934) and M.1.T.

_(ScD,1937) where he taught until joining the Research Laboratory as a research associate in 1942.

After engineering management assignments in the Chemical Division, he became General Manager of the -

- Silicone Products Department in 1952, of the Metallurgical Products Department in 1959, and of the Division in 1959. He was made Vice President in 1962, Vice President and Group Executive, Compo-nents and Materials Group in 1968, and Senior Vice President in 1971. He is a fellow of the American Institute of Engineers and a member of the National Academy of Engineering.

Dr. R. B. Richards

- Dr. Richards graduated from Pennsylvania State University (BS Chemical Engineering,1939; MS,

- Organic Chemistry,1941; PhD 1946). He joined General Electric at the Research Laboratory in 1947 and later that year transferred to the Hanford Atomic Plant, serving in a variety of managerial posi-tions until 1953 when he was appointed Manager-Chemical Separations and Pile Technology. He was transferred to the Atomic Power Equipment Department in 1957 as Manager-Engineering. In 1968 he became Manager-Engineering, Reactor Fuels and Reprocessing Department; in 1971 he was appointed Manager of Special Fuel Programs with Overseas Licensing responsibilities. He assumed his present position, Manager-International Business Development in 1973.

44 l

Biographical Information - Advisory Board, cont.

Mr. A. W. Robinson, Jr. , Secretary After receiving his M.S. E.E. degree from M.I.T., Mr. Robinson joined GE in 1940. He joined the Guided Missiles Operation in 1945, was made Manager-Guidance Engineering in 1953, Manager-Systems Engineering in 1955, Weapon Systems Engineer in 1956 and Manager of Future Growth Study, Advanced Systems Engineering and Space Business Development in 1960. After two years with the Office of the Secretary of Defense, he returned to the Company in 1965 as Manager-Advanced Require-ments Planning Operation, Missile and Space Division. From 1968 to 1969 he was Manager- Aerospace Resources Analysis, from 1969 to 1970 he was Manager-Group Integration Operation in the Information Systems Group. !!e was made Staff Executive in 1971.

Dr. S. Seltzer Dr. Seltzer holds chemical engineering degrees from Cooper Union (BS,1947) and University of Michigan (PhD,1951) where he was also a Teaching Fellou until he joined General Electric in 1951.

After several engineering assignments in the Chemical and Metallurgical Division, he joined Silicone Products Department in 1961 as a Process Engineer, later becoming Process Development Engineer, Engineering Leader, Manager-Process Engineering-Intermediate and in 1969 was made Manager-Intermediates Manufacturing, his present position.

Mr. W. B. Webster After attending the University of Tennessee, Mr. Webster joined the Manhattan Project where he remained until joining the Hanford Works in 1947. After several assignments in the design and con-struction field, he was named Project Engineer for the Dresden Nuclear Power Plant on behalf of the Atomic Power Equipment in 1955. lie then served as Manager of the Cost Estimating Unit, Project Manager for the Tarapur-India Nuclear Power Plant, Manager of the Overseas Power Plants Section and Manager of the Turnkey and Overseas Projects Department. He was named General Manager of the Overseas Nucicar Projects Department in 1973.

45

BIOGRAPHICAL INFORMATION -INVESTIGATION TEAMS Dr. D.11.- Alunatm An alumnus of Iowa State University (BS, Chemistry,1941; PhD Physical Chemistry,1948), Dr.

Ahmann joined General Electric as a Research Associate at the Knolls Atomic Power Laboratory. After holding various managerial positions there, he transferred in 1957 to Vallecitos Atomics Laboratory as Manager-Chemistry and Chemical Engineering; in 1967 he was made Manager-Chemistry and Metallurgy in the Nucleonics Laboratory. Later in 1967 he moved to Cincinnati as Manager-Nuclear Materials and Propulsion Operation and in 1968 he was made Manager-Materials Science and Technology. He assumed his present job as Manager-Engineering in the Neutron Devices Department in 1969.

Mr. V. R. Cooper After receiving a BA in Chemistry from the University of California at Los Angeles and a MS in Chimical Engineering from the University of Michigan, Mr. Cooper joined the DuPont Company at the Experimental Station. Subsequently he was engaged with the production of high explosives and assign-ments to the Manhattan Project at the University of Chicago, Oak Ridge and Hanford, where he condected process development and production assignments on plutonium isolation and purification. He joined

- GIneral Electric in 1947 at the Hanford Atomic Products Operation where he managed development programs leading to continuous separations processes for the isolation and purification of plutonium, uranium recovery, and radioactive waste disposal. He joined the General Engineering Laboratory in 1961 where he held various managerial positions. At the end of February 1974 Mr. Cooper left General Electric and has joined the Electric Power Research Institute where he has been assigned to Fossil Fuel and Advanced Systems.

Dr. J. M. Dotson Ater graduating from West Virginia University (BS, ChE,1948, PhD, ChE 1954) Dr. Dotson joined General Electric in the Silicone Products Department where he was Group Leader in such fields as development of silicone processes, fluidization, fractional distillation, organo-silane chemistry, flame processes, jet grinding, heat transfer, kinetics and thermodynamics and bench and pilot plant develop-ment. In 1967 he joined the Nuclear Energy Division where in his present assignment, Manager-Chemical Engineering he has responsibility for spent fuel reprocessing development, fluidization, fluoride volatility, calcination of uranium nitrates, radioactive waste disposal and uranium conversion processes.

Mr. E. D. Haines A graduate of Newark College of Engineering with a BS, ChE, Mr. Haines joined the DuPont Company in 1941. After two years of start-up operations in acid plants, he was transferred into their fuel develop-ment program for the Manhattan Project at Chicago and later became project engineer for the fuel fabrication area and reactor areas at Hanford. In 1947 he joined General Electric to manage the fuel development section at the Knolls Atomic Power Laboratory (KAPL). In 1950 he organized and managed th:tr manufacturing section which built two nuclear reactors for the Navy in the ersuing five years. In th3 following fifteen years he held a number of managerial positions at KAPL in areas of development shops, plant operations, health physics, radiological engineering and maintenance. In 1968 he joined Manufacturing Engineering Services as Consultant-Plant Engineering and Maintenance and assumed his present position, Manager-Machinery and Equipment Operation in 1971.

46

Biographical Information - Investigation Teams, cont.

Dr. N. J. IIawkins Dr. Itawkins received his education at Johns llopkins University (AB,1948, AM,1949 and PhD in physical chemistry in 1951). lie joined General Electric as a research associate in chemistry at the Knolls Atomic Power Laboratory in 1951 where, after individual research on tritium for a year, he became project leader of the Puf6 program. lie joined the Power Tube Department in 1957 and held vaalous managerial positions, in 1965 he joined Engineering Services where his present position is Consultant-Process Engineering.

Dr. S. Lawroski After graduating from Pennsylvania State University (BS,1934; MS,1939; PhD,1943), Dr. Lawroski joined the Standard Oil Development Company. In 1944, at the request of the Manhattan District, he was loaned to that government project. At the Metallurgical Laboratory he led a group in the develop-ment of a chemical separation process of irradiated nuclear fuels, eventually assuming direction of all this work. IIe returned to Standard in 1946, and in September of that year he was recommended as a candidate for atomic energy training at Clinton Laboratories (now Oak Ridge National Laboratory) where he studied reactor and separations technology until June 1947. The following month he joined Argonne as head of the Process Development Section and Assistant Director of the Chemistry Division. In 1948 he was named Director of the newly-organized Process Development Section. He assumed his present position, Associate Laboratory Director, in 1963.

Dr. M. C. Leverett Dr. Leverett attended Kansas State College (BS, ChE), the University of Oklahoma (MSE) and M.I.T.

(DSc in Chemical Engineering). He held research and management positions at various companies, among them Humble Oil and Refining Company, the University of Chicago Metallurgical Laboratory, Clinton Laboratories, Monsanto Chemical Company, Carbide Carbon Chemical Corporation and Tech-nical Director-NEPA project at Fairchild Engine which position he left to join Geaeral Electric in 1951 in the Aircraft Nuclear Propulsion Department as Engineering Manager. He was later made Manager-Development Laboratories. In 1961 he transferred to Hanford as Consulting Engineer; was made Manager-Research and Engineering before he transferred to Nuclear Energy Division as Manager-Division Safety. His present title is Manager-Nuclear Safety Assurance.

Mr. D. W. Lillie Mr. Lillie is a graduate of Harvard with a BA in Chemistry in 1939. After working with Crucible Steel Company and the U.S. Atomic Energy Commission, he joined General Electric at the Research Labora-tory in 1954. lie became Manager-Process Laboratories Unit in 1959; Manager-Physical Metallurgy Branch in 1961 and was appointed Liaison Scientist for the Construction Industries Group in 1970. He assumed his present position, Group Liaison Scientist for the Power Generation Business Group in 1973.

47

Biographical Information - Investigation Teams, cont.

Mr. W. K. MacCready Mr. MacCready is a graduate of the University of Alabama (BS Physical Chemistry,1934; MS,1935).

He joined the DuPont Company as an analyst in the Jackson Laboratories. After several supervisory a:signments, in 1944 he transferred to their Hanford Works as Operations Assistant Chief Supervisor, 1:aving that position in 1946 to join the General Electric Company as Superintendent in the fuel separa-tions area. In 1948 he was made Assistant Manager-Manufacturing;in 1951 he became Assistaat to \/

General Manager; in 1954 - Manager-Reactor and Fuels Operations; from 1955 to 1956 he was Manager- '

Manufacturing and in 1956 he was named General Manager-Chemical Processing Department, which position he left in 1961 to set up his own consulting firm, MacCready Consultants.

Dr. K. W. Norwood After attending Princeton University (BS, ChE,1955) and the University of Delaware (PhD, ChE,1958)

- Dr. Norwood joined General Electric as an engineer in the Irradiation Processing Department at Hanford and later served as New Production Reactor plant project representative for the Research and Engineering Section. Subsequently he was made Manager-Advance Fuel Engineering and Manager-Reactor Engineering in the N-Reactor Department. In 1963 he joined the Capacitor Department as Manager-Electronic Capacitor Product Engineering, and later became Manager of Engineering and Manager of Marketing. Before joining Engineering Consulting Services as Consultant-Engineering Operations and Resources Studies he was also Manager-RTV Product Section in Silicone Products. In September 1973 Dr. Norwood left GE to join Mechanical-Technology, Inc. (MTI) in Schenectady.

4 4

Dr. C. W. Smith A graduate of the University of Oklahoma (BS ChE,1959) and the University of Santa Clara (J. D. ,1974),

Dr. Smith joined GE at Hanford in 1959. In 1960 he joined the Chemical Process Department as Process Design and Development Engineer, in 1964 he transferred to the Fuel Recovery Operation where he has been successively Engineer-Business Studies, Engineer-Licensing and Regulation and in 1967 he assumed his present position as Manager-Licensing and Transportation.

Mr. M. L. Turner After obtaining a B.E. E. degree at New York University, Mr. Turner joined General Electric as a Project Engineer in the Light Military Electronics Department. Subsequent engineering assignments included Re-entry Systems, Apollo Systems. Mr. Turner then transferred to Systems Engineering in the Re-entry and Environmental Systems, followed by additional Systems Engineering assignments in Special Military Products Department. In 1969 he was named Manager-Quality Assurance for the Re-entry and Environmental Systems Division, and in 1973 transferred to the Nuclear Energy Division where he is currently Manager-Quality Assurance for the Breeder Reactor Operation.

48

R i pl . Biographical Information - Investigation Teams, cont.

..e.

y-Mr. M. J. Wynn

' After graduatioh from the University of Idaho (BS ChE,1965) Mr. Wynn joined the G'e neral Electric

. Company's Manufacturing Management training program and held assignments at the Hanford Works in the N Reactor and Nuclear Fuels sections, as well as assignments in the Ballast Department and the Silicone Products Department. Upon completion of the program, Mr. Wyim was appointed Process

~ Engineer in the Intermediates Process. Engineering Operation at Silicone Products and subsequently was made Production Engineer in the Intermediates Production Operation. He assumed his present position, Process Design Engineer for Fluids, Resins and Chemicals in 1973.

p Mr. L. . L.' .. Zahn

>- - A graduate of Clarkson College of Technology (BS, ChE), Mr. Zahn worked for Standard Oil Develop-ment Company as a Process Design Engineer. In 1952 he joined General Electric at Hanford as a Process Engineer in the evaluation of separations process alternatives followed by assignments in Purex -

L Separations, to Manager of Extraction, Process Design Engineering, Manager Facilities Planning and l --Manager of Purex. He remained at Hanford when GE left, working for Isochem Inc. and later Arco I Chemical Company. With Arco he was successively Manager, . Nuclear Diversification, Manager of Process Design and Manager of Engineering. He then transferred to Oxirane Corporation where he was

.,,- 1 Manager. of Capitol Projects with responsibility for world-wide plant engineering and construction.

-)% . activities which he left to return to GE in April 1974 ar.d his present assignment as Manager of the Fuel Recovery Products Operation.

n

-l1 s 5 q,,

,l

\%'

49

[

.