ML20035D960

Rev 3 to Critically Safety Engineering Rept

ML20035D960
Person / Time
Site:	Claiborne
Issue date:	04/09/1993
From:	LOUISIANA ENERGY SERVICES
To:
Shared Package
ML20035D956	List: ML20035D954 ML20035D960
References
NUDOCS 9304140154
Download: ML20035D960 (83)

Text

{{#Wiki_filter:-. Page 1 of I Louisiana Energy Services I Criticality Safety Engineering Report Push PullInstructions Revision 3 April 9,1993 Remove Insett Tab " List of Effective Pages" & - List of Effective Pages (8 pages) " Table of Contents, " Table of Contents, - pages i - vi - pages i - vi Section 4.0 Section 4.0 - pages 4-i thru 4-56 - page 4-i thru 4-56 Tab " APPENDIX H" Tab " APPENDIX I" l l Tab " APPENDIX J" & Appendix J (6 pages) Notes: 1) Each page of this revision has the month and year of the revision and Revision Number printed in the lower right hand corner of the page. 2)~ The " List of Efective Pages" contains the latest revision and date of the revision afecting the page. 3) All changes or additions to text ofeach document are indicated by a sidebar ( l } in the right hand margin. In the case of deletion of text, the sidebar appears in the right hand margin with a perpendicular line towards the text (-l ) indicating where material was deleted. 9304140154 930409 PDR ADOCK 07003070 C PDR

l LOUISLANA ENERGY SERYlCES CRITICAL 2TY SAFETY ENGINEERING REPORT l LIST OF EFFECTIVE PAGES { m Pace / Fable /Ficure Number

• Revision Number. Date of Revision TABLE OF CONTENTS P i 3, 04 2 /93 P ii 3, (WO9/93 l

P iii 3, 044)9/93 Pv 3,04A)9/93 i P v-3, (WO9/93 P vi 3, 04 2 /93

SUMMARY

2, 03/15/93 SECTION 1.0 p 1-i 2, 03/15/93 p 1-1 2, 03/15/93 p 1-2 2, 03/15/93 SECTION 2.0 p 2-i 2, 03/15/93 p 2-1 2, 03/15/93 p 2-2 2, 03/15/93 l p 2-3

2. 03/15/93 p 2-4 2, 03/15/93 p 2-5 2, 03/15/93 p 2-6 2, 03/15/93 SECTION 3.0 p 3-i 2, 03/15/93 p 3-1 2, 03/15/93

• NOTE:

A Attaciunent = P Pu8e = F Figure = R Reference = T Table = Louisiana Energy Services 1 of 8 April 1993 Criticality Safety Engineering Report Revision 3 s

l LOUISIANA ENERGY SERVICES CRITICAUTY SAFETY ENGINEERING REPORT UST OF EFFECTIVE PAGES Pace /FaNe1 Figure Number

• Revision Number. Date of Revision SECTION 4.0 p 4-i 3, 04/09/93 p 4-ii 3, 04/09/93 p 4-iii 3, 04/09/93 p 4-1 3, M/09/93 p 4-2 3, 04/09B3 p 4-3 3, M/09B3 p 4-4 3, M/09/93 p 4-5 3, M/09/93 i

p 4-6 3, M/09/93 p 4-7 3, M/09/93 p 4-8

3. M/09B3 p 4-9 3, 04/09/93 p 4-10 3, 04/09/93 p 4-11

3. N/09/93 O

p 4-12 3, 04/09/93 p 4-13

3. 04/09/93 p 4-14 3, 04/09/93 p 4-15 3,04/09B3 p 4-16 3,04/09B3 p 4-17 3, M/09/93 p 4-18 3, 04/09/93 p 4-19 3, M/09/93 p 4-20 3, M/09/93 p 4-21 3, M/09/93 p 4-22 3, 04/09/93 p 4-23 3 M/09B3 p 4-24 3, 04/09/93 p 4-25 3,04/09B3 p 4-26 3, 04/09/93 Attaciunent

• NOTE:

A = Page P = Figure F = Reference R = Table T = Inuisiana Energy Services 2 of 8 April 1993 Criticality Safety Engineering Report Revision 3

LOUISEANA ENERGY SERVICES CRITICAUTY SAFETY ENGINEERING REPORT O l UST OF EFFECTIVE PAGES Pane /raNe/ Figure Number

• Revision Number. Date of Revision l

p 4-27 3 M/09/93 l p 4-28 3, M/09B3 p 4-29 3, 04/09/93 p 4-30 3,04/09B3 p 4-31 3,04/09B3 l p 4-32 3,04/09B3 p 4-33 3, 04/09/93 i p 4-34

3. M/09/93 p 4-35 3, M/09/93 p 4-36 3, M/09/93 p 4-37 3, 04/09/93 p 4-38 3, 04/09/93 p 4-39 3, M/09/93 p 4-40 3, 04/09/93 p 4-41

3. 04/09B3 p 4-42 3,04/09B3 p 4-43 3, 04/09/93 p 4-44 3, 04/09/93 I

p 4-45 3, M/09/93 p 4-46 3, 04/09/93 p 4-47 3, 04/09/93 p 4-48 3, M/09/93 l p 4-49 3, M/09/93 p 4-50 3, M/09/93 p 4-51 3, 04/09/93 p 4-52 3,04/09B3 p 4-53 3 M/09/93 p 4-54 3 M/09B3 p 4-55 3,04/09B3 p 4-56 3 M/09B3 Attachment

• NOTE:

A = page p = Figure F = Reference R = O Table T = Louisiana Energy Services 3of8 April 1993 Criticality Safety Engineering Report Revision 3 l l

LOUISIANA ENERGY SERVICES CRITICALITY SAFETY ENGINEERING REPORT LIST OF EFFECTIVE PAGES \\ Pace /1'able/Fieure Number

• Revision Number. Date of Revision SECTION 5.0 i

p 5-i 2, 03/15/93 p 5-1 2, 03/15/93 p 5-2 2, 03/15/93 SECTION 6.0 p 6-i 2, 03/15/93 p 6-1 2, 03/15/93 j l REFERENCES f p R-1

2. 03/15/93 l

l p R-2 2, 03/15/93 ) TABLES TI 2, 03/15/93 l T2 2, 03/15/93 T3 2, 03/15/93 i I FIGURES F1 2, 03/15/93 F 2 (page 1 of 3) 2, 03/15/93 F 2 (page 2 of 3) 2, 03/15/93 i F 2 (page 3 of 3) 2, 03/15/93 F3 2, 03/15/93 F4 2, 03/15/93 APPENDIX A i p A-i 2, 03/15/93 [ p A-1 2, 03/15/93 p A-2 2, 03/15/93

• NOTE:

A Attachment = p page = F Figure = Reference R = Table T = Louisiana Energy Services 4 of 8 April 1993 Criticality Safety Engineering Report Revision 3 1

1 LOUISIANA ENERGY SERVICES l CRITICAUTY SAFETY ENGINEERING REPORT LIST OF EFFECTIVE PAGES l i j l Pace / Table /Ficure Number

• Revision Number. Date of Revision j

i p A-3 2, 03/15/93 I i p A-4 2, 03/15/93 p A-5 2, 03/15/93 9 p A-6 2,03/15B3 F1

2. 03/15/93 j

A 1 (9 pages) 2, 03/15/93 i A 2 (3 pages) 2, 03/15/93 A 3 (2 pages) 2, 03/15/93 A4 2, 03/15/93 l APPENDIX B p B-i

2. 03/15/93 p B-1 2, 03/15/93 p B-2 2, 03/15/93 p B-3 2, 03/15/93 p B-4 2, 03/15/93 p B-5 2, 03/15/93

[ p B-6 2, 03/15/93 A1 2, 03/15/93 I A2 2, 03/15/93 A 3 (3 pages) 2, 03/15/93 i APPENDIX C l I p C-i 2, 03/15/93 p C-1 2, 03/15/93 p C-2 2, 03/15/93 p C-3 2, 03/15/93 A (3 pages) 2, 03/15/93 I

• NOTE:

A Attachment page p = Figure F = Reference R = Table T = Louisiana Energy Services 5 of 8 April 1993 Criticality Safety Engineering Report Revision 3 ,.,,,,,,,.y,. -,m, _._.m.--,.m.., ..g...

+a t i LOUISIANA ENERGY SERVICES CRlTICAUTY SAFETY ENGINEERING REPORT l UST OF EFFECTIVE PAGES Paceffable/Ficure Number

• Revision Number, Date of Revision APPENDIX D p D-i 2, 03/15/93 p D-1 2,03/1563 p D-2 2, 03/15/93 l

p D-3 2, 03/15/93 p D-4 2, 03/15/93 p D-5 2, 03/15/93 p D-6 2, 03/15/93 i p D-7 2, 03/15/93 p D-8 2, 03/15/93 p D-9 2, 03/15/93 p D-10 2, 03/15/93 p D-11 2, 03/15/93 p D-12 2, 03/15/93 p D-13 2, 03/15/93 p D-14 2, 03/15/93 p D-15 2, 03/15/93 { l p D-16 2, 03/15/93 p D-17 2, 03/15/93 APPENDIX E p E-1 2, 03/15/93 p E-2 2, 03/15/93 p E-3 2, 03/15/93 l p E-4 2, 03/15/93 p E-5 2, 03/15/93 p E-6 2, 03/15/93 i R 2 (14 pages) 2, 03/15/93 l l 1

• NOTE:

A Attachment = p page = F Figure = R Reference = T Table = Louisiana Energy Services 6 of 8 April 1993 Criticality Safety Engineering Report Revision 3

i .1 LCUISIANA ENERGY SERVICES CRiitCALITY SAFETY ENGINEERING REPORT i LIST OF EFFECTIVE PAGES Paec/ Table / Figure Number

• Revision Number. Date of Revision APPENDIX F p F-1 2, 03/15/93 p F-2 2, 03/15/93 p F-3 2, 03/15/93 p F-4 2, 03/15/93 A (1 page) 2, 03/15/93 i

APPENDIX G l p G-1 2, 03/15/93 l APPENDIX H p H-1 2, 03/15/93 l p H-2 2, 03/15/93 l p H-3 2, 03/15/93 p H-4 2, 03/15/93 t l p H-5 2, 03/15/93 F1 2, 03/15/93 F2 2, 03/15/93 i F3 2, 03/15/93 APPENDIX I p1-1 2, 03/15/93 p1-2 2, 03/15/93 I F1 2, 03/15/93 I

• NOTE:

A = Attachment p page = F Figure = R Reference i = Table T = i Louisiana Energy Senices 7 of 8 April 1993 Criticality Safety Engineering Report Revision 3

LOUISLANA ENERGYSERVICES CRITICAUTY SAFETY ENGINEERING REPORT l UST OF EFFECTIVE PAGES i Page/ Table / Figure Number

• Revision Number. Date of Revision 1

APPENDIX J l p J-l 3, 04/09/93 pJ-2 3, 04/09/93 i pJ-3 3, 04/09/93 l p J-4 3, 04/09/93 F1 3, 04/09/93 i i i l l i l k l l l i

• NOTE:

A = Attachment P = page F Figure = R Reference = T = Table Louisiana Energy Services 8 of 8 - April 1993 Criticality Safety Engineering Report Revision 3 l

I i LIST OF ABBREVIATIONS, ACRONYMS AND SYMBOLS i BNFL British Nuclear Fuels Ltd. CEC Claibome Endchment Center em Centimeter CMS Centdfuge Monitor System C, Safe Concentration D, Safe Diameter F Degrees Fahrenheit F Fluorine FSRC Facility Safety Review Committee ft Feet g Grams i GEVS Gaseous Effluent Vent System HEPA High Efficiency Particulate Air i hr Hour HVAC Heating, Ventilating and Air Conditioning in Inch k,, K-effective [' kg Kilogram 1 Liter LES Louisiana Energy Senices LWD Liquid Waste Disposal l mbar Millibar M, Critical Mass mrem Millirem l R SdeMm NRC Nuclear Regulatory Commission ORNL Oak Ridge National Laboratory I ppm Parts Per Million psia Pounds Per Square Inch Absolute RCZ Radiation Control Zone RSC Radiation Safety Committee S, Safe Slab nickness TSA Technical Services Area U234 Uranium 234 Isotope U235 Uranium 235 Isotope Louisiana Energy Senices ii April 1993 Criticality Safety Engineering Repon Revision 3

~ \\ UF. Uranium Hexafluoride UO F Uranyl Floride 2 2 V. Safe Volume wt Weight wt% Weight Percent l j i l i i i Inuisiana Energy Services ill April 1993 Criticality Safety Engineering Report Revision 3 1

j LIST OF TABLES l i ) i TABLE 1 Safe Values for Uniform Aqueous Solutions of 5% Enriched UO F i 2 2 TABLE 2 K,, for Square Array of Failed Centrifuges Within a 10 x 10 Array l TABLE 3 Estimated Annual Uranium (kg) Input and Disposition, TSA l i i l I -l l l i i 1 l i I I I E l I I l 4 i i n Louisiana Energy Sen' ices iv April 1993 i l Criticality Safety Engineering Report Jtn'ision 3 I s ...-__._._.;..._.--_,.--__........_,-._.-__...__.__.._...1

f LIST OF FIGURES FIGURE 1 CEC Components Containing Uranium FIGURE 2 Mechanical Flow Diagram Product Vent System F:GURE 3 Process Flow Diagram Cascade Mobile Pump Set and Sample Rig FIGURE 4 Liquid Effluent Systems 1 1 i rrU l O Louisiana Energy Services v April 1993 Criticality Safety Engineering Report Revision 3

APPENDICES i APPENDIX A Centrifuge / Centrifuge Anay Criticality Analysis APPENDIX B Desublimer Criticality and Interaction Analysis l APPENDIX C Contingency Dump Trap Criticality Evaluation APPENDIX D Criticality Event Evaluation APPENDIX E Moderation Control in 30" Product Transit Cylinders (including "BNFI Nuclear Safety of the Storage of 30B Hex Containers on the Capenhurst Site) APPENDIX F Calculation of Minimum Quantity of Water Required to Produce j a Criticality with 5% UO F / 5% UO F 3.5 H O 2 2 2 2 2 APPENDIX G Safe Operating Conditions for UF. Cylinder Storage Areas l APPENDIX H Criticality Safety of Process Pipe Bends and Joints Installed in Centrifuge Uranium Enrichment Plant APPENDIX I Criticality Safety of Liquid Waste Process and Collection Tank Arrays l APPENDIX J Criticality Safety of the UCN Mobile Vacuum Pumpset Used in l Centrifuge Uranium Enrichment Plant i l I i Louisiana Energy Senices vi April 1993 Criticality Safety Engineering Report Revision 3 l

t l l l l TABLE OF CONTENTS l i [ 4.0 ANALYSIS OF SYSTEMS AND COMPONENTS................... 4-1 i 4.1 CENTRIFUGES AND CASCADES............................... 4-1 l t 4.1.1 Normal and Abnormal Operations............................... 4-1 l 4.1.2 Criticality Safety Discussion................................... 4-4 l 4.1.3 S umm ary................................................ 4-6 4.2 PRODUCT CYLINDERS...................................... 4-7 l 4.2.1 General Description......................................... 4-7 4.2.2 Normal and Abnormal Operations............................... 4-7 4.2.3 Criticality Safety Discussion.................................. 4-10 l 4.2.3.1 Criticality Safety Case For A Sincle 30B Cylinder................. 4-10 l 4.2.3.2 Array Analysis.......................................... 4-1 1 4.2.4 S umm ary............................................... 4-13 4.3 DESUB LIMERS........................................... 4-14 i 4.3.1 Normal and Abnormal Operations.............................. 4-14 4.3.2 Criticality Safety Discussion.................................. 4-16 4.3.2.1 Moderation Via HF impurities in ' Die Product Flow................ 4-17 4.3.2.2 Moderation Via HF Gas Produced From Air Inleakage.............. 4-18 4.3.2.3 Moderation Via Water In ne Form Of Hydrates.................. 4-18 4.3.2.4 Safety Of The Desublimer/ Product Cylinder Array................. 4-18 4.3.2.5 Summary............................................... 4-19 4.4 CHEMICAL TRAPS.......................................... 20 4.4.1 Normal and Abnormal Operations.............................. 4-20 4.4.2 Criticality Safety Discussion.................................. 4-21 4.5 CONTINGENCY DUMP TRAPS................................ 4-21 4.5.1 Normal and Abnormal Operations.............................. 4-21 4.5.2 Criticality Safety Discussion.................................. 4-22 l 4.6 VACUUM PUM PS.......................................... 4-23 t j 4.6.1 Normal and Abnormal Operations.............................. 4-23 4.6.2 Criticality Safety Discussion.................................. 4-23 4.7 PR OCESS PIPING.......................................... 4-24 4.7.1 Normal and Abnormal Operations.............................. 4-24 4.7.2 Criticality Safety Discussion.................................. 4-24 4.8 VACUUM PUM P OIL....................................... 4-25 4.8.1 Normal and Abnormal Operations.............................. 4-25 4.8.2 Criticality Safety Discussion.................................. 4-26 Louisiana Energy Services 4-i April 1993 Criticality Safety Engineering Report Revision 3 ~. -

4.9 VACUUM CLEANERS...................................... 4-27 4.9.1 Normal and Abnormal Operations..... 4-27 \\, 4.9.2 Criticality Safety Discussion.................................. 4-27 4.10 CONTAMINATED SOLID WASTES........................... 4-27 4.10.1 Normal and Abnormal Operations............................. 4-27 4.10.2 Criticality Safety Discussion................................ 4-8 4.11 CONTAMINATED AQUEOUS WASTES AND ASSOCIATED PROCESSING EQUIPMENT.................................. 4-30 4.11.1 Normal and Abnormal Operations............................. 4-30 4.11.2 EfDuent Collection Tanks, UF, Handling Area..................... 4-32 4.11.2.1 Normal and Abnormal Operations........................... 4-32 4.11.2.2 Criticality Safety Discussion............................... 4-32 4.11.3 EfDuent Collection Tanks Technical Services Area (TSA)........ 4-33 4.11.3.1 Normal and Abnormal Operations............................ 4-33 4.11.3.2 Criticality Safety Discussion................................ 4-34 4.11.4 Citric Acid Baths And Spent Citric Acid Tank, TSA................ 4-37 l 4.11.4.1 Normal and Abnormal Operations............................ 4-37 4.11.4.2 Criticality Safety Discussion............................... 4-38 4.11.5 LWD Reaction Tank And Precipitation Centrifuge.................. 4-40 b 4.11.5.1 Normal and Abnormal Operations... ........................4-40 \\ 4.11.5.2 Criticality Safety Discussion................ 4-40 4.11.6 Decontamination Effluent Tank And Rinse Water Baths.............. 4-41 4.11.6.1 Normal and Abnormal Operations.. .........................4-41 4.11.6.2 Criticality Safety Discussion............................... 4-42 4.11.7 La u n d ry Tank........................................... 4 -4 3 4.11.7.1 Normal and Abnormal Operations........................... 4 -4 3 4.11.7.2 Criticality Safety Discussion................................ 4-43 4.11.8 Laboratory Tanks......................................... 4-44 4.11.8.1 Normal and Abnormal Operations............................ 4-44 4.11.8.2 Criticality Safety Discussion................................ 4-44 1 4.11.9 Dryer Feed Tank......................................... 4-4 5 4.11.9.1 Normal and Abnormal Operations............................ 4-45 4.11.9.2 Criticality Safety Discussion................................ 4-46 4.11.10 Dryer, Dryer Entrainment Separator, Dryer Condenser, Dryer Distillate Tank 4-47 4.11.10.1 Normal and Abnormal Operations........................... 4-47 4.11.10.2 Criticality Safety Discussion............................... 4-48 4.11.11 Effluent Monitor Tanks And Demineralizers..................... 4-49 f (\\ Louisiana Energy Services 4-ii Aprii 1993 Criticality Safety Engineering Report Revision 3

4.11.11.1 Normal and Abnormal Operations........................... 4-49 4.11.11.2 Criticality Safety Discussion............................... 4-50 4.11.12 30 Gallon Dry Powder Drums............................... 4-51 N 4.11.12.1 Normal and Abnormal Operations.......................... 4-5 1 4.11.12.2 Criticality Safety Discussion..... .........................4-52 4.11.13 Dryer Feed Tank FC:er.................................... 4-52 4.11.13.1 Normal and Abnormal Operations........................... 4-52 4.11.13.2 Criticality Safety Discussion............................... 4-53 4.12 ACTIVE VENTILATION SYSTEMS............................ 4-53 i 4.12.1 Normal and Abnormal Operations............................. 4-53 4.12.2 Criticality Safety Discussion................................. 4-54 4.13 DEGREASER BATHS.......... .......................... 4-55 4.13.1 Normal and Abnormal Operations ....................... 4-55 4.13.2 Criticality Safety Discussion ........................4-56 t i l l l l ( .s Louisiana Energy Se-. ices 4.;;; I Criticality Safety Engineering Report Renss.on 3

4.0 ANALYSIS OF SYSTEMS AND COMPONENTS ? t t r t The systems, components and operations of the CEC were studied to determine where l I accumulations of uranium are possible. These systems, components and operations then were subjected to a criticality safety analysis. Those which are similar and have similar criticality concerns were addressed together. Differences were considered where appropriate. Some { were unique components or operations and required separate analysis. l The following subsections address each of the categories. The normal operations, abnormal i operations, pertinent controls and safety systems are described. Criticality safety is addressed for every aspect, including any postulated means by which criticality is possible. 'Ibe l contingencies regarding process changes, upsets or failures which must occur before a i i criticality is possible are presented and the methods for maintaining criticality safety are j l described. 5 l 1 4.1 CENTRIFUGES AND CASCADES \\ I 4.1.1 Normal and Abnormal Operations j l Arrays of gas centrifuges, called cascades, are used to separate gaseous UF feed, with a natural uranium isotopic concentration,into a product stream enriched in the U235 isotope and a stream depleted in the U235 isotope. This system is further described in SAR Section 6.3.1, Enrichment System. The gas centrifuges used in the CEC are designed to operate continuously with a mean life of i 10 to 14 years. To assure reliable operation, the design incorporates features from many l years of actual operating experience. Each centrifuge has an outer casing which acts as a l vacuum chamber to reduce friction on the spinning centrifuge rotor and also is a barrier for { l flying parts should a centrifuge fail, i.e., " crash." Failures of centrifuges are detected by motor current sensors which initiate messages in the control room. The loads resulting from these infrequent centrifuge failures are restrained by the casing and the floor mounting element (flomel.) The casing is designed so that rotor debris is contained and the flomels are designed so that the casings can not break away from the floor in the event of a centrifuge crash. Gaseous UF. is fed at subatmospheric pressures through piping headers to the assay units, each consisting of seven parallel centrifuge cascades. Each cascade receives feed from the O Louisiana Energy Services 4-1 April 1993 Criticality Safety Engineering Report Revision 3 i' _--_.,_..._.-.__i

header via a feed control system. Deviations from the feed pressure control point are automatically recorded and alarmed. b The centrifuge temperature is controlled by a closed-loop cooling water system and the cascades are housed within enclosures to maintain their temperature stability. Under normal operation there are no wastes or effluents generated in the cascades. 'Ihe cascades are monitored from the central control room. A sophisticated Cascade Protection System detects crashes and process upsets, signals the control room and initiates automatic corrective actions. The following parameters are monitored: A. Cascade header pressure. B. Centrifuge motor power. C. UF, Feed System supply pressure and valve positions. D. Contingency Dump System pressure and valve positions. E. Machine Cooling Water System temperature and flow. F. Product Take-Off System pressure and valve positions. G. Tails Take-Off System pressure and valve positions. If one or more of the support systems fail or the process parameters are outside of the specified limits, the Cascade Protection System initiates actions which establish safe operation or safe shutdown to protect the centrifuges. Deviations from the normal pressure in the cascade headers can affect the optimum process conditions of the centrifuges with respect to enrichment performance and can jeopardize the rotor by causing UF. desublimation inside the rotor. The presence of light gases can cause l centrifuge motor overload. Sometimes immediate evacuation, i.e., " dumping", of the cascade l UF, inventory is required. To achieve unrestricted UF, removal, the cascade control anangement is by-passed to the Product and Tails Take-Off Systems which minimizes flow resistance. If a failure of these dumping systems occurs, a third take-off system, the Contingency Dump System is activated. If a centrifuge fails, i.e., " crashes", nearly all of the rotational energy is converted to heat. This can produce a variety of organic vapors from the organic materials in the rotor. Some of these vapors react with UF at variable rates of reaction. The products of these reactions Louisiana Energy Services 4-2 April 1993 Criticality Safety Engineering Report Rnision 3 r

I l j i I are also varied including gaseous products such as panly fluorinated hydrocarbons and solid products such as UF. and UF - 5 The solid reaction products are generally in a finely divided " dust" which can be dispersed in the cascade piping or in other centrifuges, impairing their enrichment efficiency. A centrifuge crash can upset the process performance, but is not a worker or public safety concern. If a rotor fails, the pressure inside the casing rises rapidly by approximately an order of magnitude, but to a level well below atmospheric pressure. This pressure rise and the ensuing gas velocity effect are used to activate automatic isolation devices which close on pressure and open by gravity and seal the centrifuge from the rest of the cascade. He leak rate of these devices is so small that any remaining reactive gases react within the pipes connecting the centrifuge to the cascade. It requires several days, sometimes weeks, before the centrifuge pressure falls to a value where the isolation devices can open by gravity. In this period all of the remaining reactive gas has reacted and the solid reaction products within the failed centrifuge have agglomerated and settled to the bottom. Tests performed by URENCO have shown that these solid reaction products only contain a few grams of uranium. His is I as would be expected, considering the small UF inventory in a vacuum centrifuge. The unreactive gases pass through the cascade and are eventually discharged through the GEVS. Failed centrifuges are left in place. Their failure is indicated in the control room by the k Centrifuge Monitoring System (CMS). This system scans the motor current and voltage wave forms of each centrifuge and calculates not only the phase angle between these waveforms but also the power drawn by each centrifuge. This data is displayed in the control room using a color graphics system. When a centrifuge failure occurs, there is a pronounced change in its phase angle and power. These changes are detected by the CMS and alarmed on a data logging system as well as on the color graphics system. I Centrifuge casings are designed to maintain their vacuum integrity during and after a centrifuge crash. URENCO experience in Europe (over 30 plant years of operation) has validated the design. One could, however, conceive an event in which a centrifuge crash leads to a small air leak into the machine of magnitude just less than the limit of detection. Such a leak would introduce atmospheric moisture into the machine which would react with j the UF. in the casing and form UO F : 2 2 UF, + 2H O -+ UO F + 4HF 2 2 2 b (w Louisiana Energy Services 4-3 Apnl1993 Cdtically Safety Engineering Report Revision 3

^% It is also chemically possible, but highly unlikely, that the UO F would form as its highest 2 2 hydrate UO F a 1.3 H O (Reference 9). 2 2 2 Again to be conservative it will be assumed that the 4HF molecules are retained in the hydrate matrix, effectively forming UO F

• 3.3 H O for moderation purposes. This is 2 2 2

rounded up to 3.5 H O in the following analysis. 2 4.1.2 Criticality Safety Discussion The maximum allowable product enrichment is 5 wt% U235. The cascade and plant design ensure the enrichment process can have only the designated configuration by providing no interconnections which would allow the enriched product to be routed to the feed system or to allow any other configuration changes to the cascades. Product cylinders are significantly different from feed cylinders and are not readily interchangeable in the process equipment (for example, a special cradle would have to be manufactured in order to support a product cylinder inside a feed autoclave). Therefore, re-enrichment of product by operator error can be excluded. Controlling the feed and product flow ensures that the desired endchment (normal operating margin 0.02 wt%) is attained and that the maximum allowable enrichment is not exceeded. 1 V The administrative and operational procedures used to control product enrichment are as follows: A. Plant production schedules are assessed and optimized. B. An enrichment level is assigned to an assay unit C. A campaign change is authorized in writing. D. Proper control valve settings are calculated using a computer model based on actual operating experience. l E. Operators are instructed in writing to change control valve settings and given predicted values for enrichment and cascade pressures. F. Control valves are adjusted then the settings are independently checked and recorded. Access to the control valve adjusting devices is physically protected. De resulting pressures are compared with the predicted pressures and reported to the computer modeler. G. Within 24 hours, gas samples of the UF product from the assay unit are taken. Actual enrichment is compared with the predicted enrichment and reported to the computer modeler. v Louisiana Energy Services 4-4 April 1993 Criticality Safety Engineering Report Revision 3

H. If minor adjustments are required, the procedure is repeated, after making an appropriate correction to the computer model. I. UF. samples are also taken from individual cascades at least every 6 months to ensure L the stability of the cascade performance and to limit enrichment variations within an assay unit. To be critically safe by geometry, the maximum diameter for an infinitely long cylinder is 25.4 centimeters when filled with a uniform aqueous solution of uranium enriched to 5 wt% U235 at the optimum H/U moderation ratio. The centrifuge casing diameter is less than 25.4 centimeters and therefore, individually, the centrifuges are critically safe by geometry. Since all of the centrifuges are precisely located on the flomels, the interaction of centrifuges in these arrays was also analyzed (Appendix A). These calculations are based on the very pessimistic assumption that various numbers of centrifuges at the pmduct end of a cascade l have crashed and become totally filled with UO F 3.5 H O (see Section 4.1.1). Further 2 2 2 l pessimism is introduced in that these calculations were performed for 6% enriched material to incorporate an additional safety margin. I 1 Two cases of single centrifuge failures filled with UO F

• 3.5 H O in an array of unfailed 2 2 2

centrifuges using the KEN 05A-PC code yielded k of 0.4695 and 0.4781. Other cases, i.e., m y square arrays of failed centrifuges filled with UO F

• 3.5 H O within a 10 x 10 square array, 2 2 2

were also analyzed using the KEN 05A-PC code. The results are shown in Table 2. l The calculated probabilities of occurrence of 2 x 2 and 3 x 3 arrays of failed centrifuges by the end of the 40 year plant life were 7.7 x 10~' and 5.6 x 10, respectively, (see Appendix A). l The criticality accident can thus only occur with the centrifuges under the following highly unlikely, mitigated conditions: A. An array of greater than twenty-five crashed centrifuges occurs in the product stage of j a cascade, j B. 'Ihe crashes allow inleakage of moisture in air or coolant water into each of the centrifuges in quantities which do not trip automatic shutdown systems, and C. All of the centrifuges fill with reaction products from the crash plus reaction of incoming UF with moisture and form hydrated UO F, and 2 2 D. All of the above failures remain undetected. O (N Louisiana Energy Services 4-5 April 1993 Criticality Safety Engineering Report Revision 3 l [

l 1 l l The crash gas dust contains only a few grams of uranium and is not a significant contributor to a criticality accident. He volumetric capacity of each centrifuge is 297 kg of water l moderated uranium. Twenty-five centrifuges at their capacity of 7425 kgU are still suberitical. This is a highly improbable quantity of uranium to be deposited via a process upset. A monthly mass balance is performed on the Process System that would detect a holdup of < 50 kgU. Typical observed centrifuge failure rates are 0.1% per year, equivalent of approximately one per cascade per year. The 10 to 14 crashes which could be expected over the mean life of a cascade would be expected to occur (i) throughout a cascade and not just in the product stage and (ii) without an associated air inleakage. De centrifuges are critically safe by geometry. Square arrays of centrifuges up to twenty-five are critically safe. For twenty-five or more product centrifuges to go critical, the essentially impossible conditions previously described would have to occur. Furthermore, operators would have to deliberately ignore a large number of failed centrifuge signals (from the Centrifuge Monitoring System) in the product part of the cascade, the higher power levels in the cascade and the loss of product UF.. (~ 4.1.3 Summary k The foregoing analysis shows that many independent statistically unlikely events would have to occur before a criticality in an array of crashed centrifuges is possible. Furthermore, the l following administrative barriers would also have to be violated: A. Operators would have to fail to respond to the high power levels in the cascade (high drag on the centrifuge rotors caused by excessive light gas). B. Operators would have to fail to report unusually large numbers of crashed machines in a cascade. C. Operators would have to fail to report clusters of crashed machines in the product stages of a cascade. D. Production planners would have to fail to notice the " loss" of over 7 tons of enriched uranium. l E. Ilealth physics personnel would have to fail to notice the increased y radiation levels resulting from U238 daughter products growing into UO F hydrate postulated to be 2 2 forming in the crashed machines. 1 o Louisiana Energy Services 4-6 April 1993 Criticality Safety Engineering Report Revision 3

F. Operators would have to fail to react to a highly increased venting frequency of the product cylinders. Monitoring of the vent frequency forms part of the criticality safety case for the product cylinders themselves (see Section 4.2). l Although a criticality in arrays of crashed centrifuges can be ruled out on statistical grounds alone, the above also demonstrates that many administrative procedures (barriers) would also have to be violated before it could occur. The double contingency principle is thus more than satisfied. J 4.2 PRODUCT CYLINDERS 1 { l 4.2.1 General Description i i The Product Take-Off System provides continuous withdrawal of enriched gaseous UF. products from the centrifuge cascades, via a train of vacuum pumps into 30 in. diameter product criinders (designated Model 30B) where the UF is solidified. A secondary function is to provide a path for rapidly dumping the cascade's UF contents under abnormal operating conditions. 'Ihere are six Product Take-Off Systems (one for each assay unit, two for each plant unit), each consisting of five product cylinder stations. Each product station consists of a cold chest contaimng a product cylinder, an Air-Cooling System and a Cylinder Weighing System. A Product Liquid Sampling System provides for sampling the contents of all UF enriched product cylinders to verify the precise assay of the U235 and the perity of the UF 'Ibere is one Product Liquid Sampling System for each of the three plant units. The system consists of autoclaves and sample bottles connected to the cylinder by a manifold. i A Product Blending System provides for the blending of contents of two UF. product cylinders to obtain a specific U235 assay which meets customer requirements. There is one central p! ant-wide Product Blending System. It consists of two autoclaves to contain the product cylinders, " donors *, which were selected for blending and five receiver stations to contain blended product cylinders, " receivers". 4.2.2 Normal and Abnorrnal Operations Inside the plant, full an

• empty 30B cylinders are moved on a rail mounted transporter to allow precise location. When cylinders are connected to process piping, the piping which has Louisiana Energy Services 4-7 April 1993 Criticality Safety Engineering Repon kision 3

. l

been exposed to the atmosphere is evacuated to remove air and leak tested using a mobile vacuum pump set. s) Incoming empty cylinders are weighed on an accountability scale, inspected Iw mtenor i contaminants and evacuated. When required for use, an inspected cylinder is transported and loaded into a product cylinder station and connected to the process piping. He cylinder is ] cooled with air at 50 F and brought online with the process. The quantity of UF,in the cylinder is continuously monitored by a Load Cell System. Light gas in the process from niinor quantities of impurities remaining in the feed UF. and small leaks in the process equipment preferentially exit with the product. These gases are vented through a Product Vent System when the cylinder pressure reaches about 450 mbar i (see Section 4.3). The Load Cell System indicates if a cylinder can be put back online after i venting. When a cylinder is filled to capacity it is valved off from the process. He cylinder is disconnected, removed from the cylinder station, weighed on the accountability scale and stomd awaiting sampling. The enriched product in 30B cylinders is transported from the Product Storage Area and (, loaded into a product liquid sampling autoclave. He cylinder is secured to the autoclave by clamps to prevent movements when the autoclave is tilted. Then the cylinder is connected to the sampling manifold. Functioning of the Autoclave Safety Systems is checked. The product cylinder is heated indirectly using air drawn over electrical heaters to the point where the UF. liquefies. During heating the cylinder, the autoclave pressures and temperatures are monitored, automatically controlled and alarmed. The liquid UF, is allowed to homogenize over time via intemal convective currents, then the heater is turned off. He autoclave is tilted about 30 degrees to 1 i pour the liquid UF. into the sampling piping manifold. De volume of each of the three manifold branches is less than the volume of one sample bottle. De autoclave is retumed to the horizontal position and the manifold valves are opened griefly to fill the sample bottles, then closed. The air heater is turned on, the UF. remaining in the manifold vaporizes then condenses in the cylinder. He valve to the cylinder is closed and the autoclave is cooled by circulating water through structurally independent coils on its exterior. s l Louisiana Energy Services 4-8 April 1993 \\ Criticality Safety Engineering Report Revision 3 r

1 i The autoclave is vented to the GEVS and its HF Monitoring System after sampling but prior to opening the autoclave and retneving the sample bottles, so any UF. leaks are detected. s ( After the UF has solidified, the sample bottles are removed and taken to the Chemistry i Laboratory. Then the cylinder is disconnected from the sampling manifold and removed from the autoclave and transported to the Product Storage Area. For blending enriched product, two 30B donor cylinders are loaded into autoclaves and an empty 30B cylinder is loaded into an air cooled cold chest. The cylinders are connected to I the transfer piping. The functioning of the Autoclaves Safety Systems is checked. The donor cylinders are heated using air drawn over electrical heaters until the system is ready to transfer UF.. The gaseous UF. pressure is reduced to subatmospheric pressum using a contru! valve within the autoclaves. All UF outside the autoclaves is handled at subatmosphenc pressure. A prescribed mass of gaseous UF., which produces the desired U235 concentration,is i transferred from the donor cylinders to a cooled receiver cylinder. After continued blending operations have reduced the UF. contents of a donor cylinder to a minimum, the remaining UF is discharged to the Blending Vent System (see Section 4.3). After UF removal the p t donor cylinders are cooled by circulating water through coils on the exterior of the k autoclaves. The autoclaves are vented to the GEVS to check for UF. leaks prior to being opened. The empty donor cylinders and filled receiver cylinders are transported to the Product Storage Area. I Abnormal operation of the product cylinders having relevance to criticality safety involves: l A. Moderating material erroneously being present in an empty 30B cylinder befom it is attached to the plant. B. Moderating material entering the cylinder when it is being filled either with the process gas or via a leak to atmosphere on the cylinder itself. C. An assay unit producing an enrichment greater than 5% l l Scenarios A and B and the barriers interposed to prevent them are discussed in the following section. Control of enrichment has already been discussed in Section 4.1.2. l O Louisiana Energy Services 4-9 April 1993 Criticality Safety Engineering Report Redsion 3 l i i l

= r l 4.2.3 Criticality Safety Discussion j 4.2.3.1 Criticality Safety Case For A Single 30B Cylinder i i Product cylinders filled to the weight limit of 2277 kg of UF enriched to 5 wt1 U235 are } critically safe if moderation controlled. i l l The filling of product cylinders and product blending receiver cylinders is carefully monitored by continuous weighing on load cells to ensure that the weight limit is not exceeded. -j Continuous monitoring of the cylinder pressure provides a check on the purity of the UF and j i therefore, on the absence of moderators. j l Arriving empty 30B product cylinders could potentially contain moderating materials such as l water, oils or hydrates of uranium compounds. Procedures at the CEC therefore require that the following series of inspections be performed on all incoming empty 30B cylinders prior to use: l i A. The empty cylinders are weighed to ensure that the weight is within 10 kg of the j labelled tare weight. These 10 kg include the heel limit (s 11.3 kg) and the scale l l precision of 1 kg. l B. The interior is checked for oil contamination and hydrates by a visual inspection using i l a boroscope. C. The cylinder is then evacuated to a few microns and isolated, and the pressure j monitcred for 5 minutes. If the pressure rise is less than 5 torr over this period then no free water can be present in the cylinder. D. Successful completion of these inspections cenify the cylinder for use. 7 i For the unlikely event that the presence of contaminants in an empty 30B cylinder goes undetected through the initial inspection tests (weighing, visual, vacuum), another independent check is conducted prior to full use in the process. After connection to the plant, a small quantity (about 10 kg) of UF is initially admitted into the cylinder. The cylinder is isolated and the pressure and surface temperature are monitored. The presence of oils or free water would be detected by the pressure rise from the reaction with UF.. If the indicated pressure is not within 5 kPa of the expected UF. vapor pressure at the cylinder temperature, the cylinder is removed from the system. i Louisiana Energy Servicer 4-10 April 1993 Cri:icality Safety Engineering Report Rn>ision 3 I

l i l \\ l l Within the enrichment process the primary potential moderator in the product stream is HF formed from the reaction of UF. with moisture. The HF can enter the process as an impurity l in the UF, feed, as a result of minor inleakage of air into the process or, along with other j reaction products, from crashed centrifuges. l The light gases and HF entering the process via leaks and the gaseous products from crashed centrifuges preferentially enter the product stream. Product Vent Systems (described in Section 4.3) are used to remove these impurities. Product cylinder venting is also required to reduce the back pressure to maintain the efficiency of product collection. This also prevents the buildup of impurities. i Cylinders which requite more than a specified number of vents are disconnected and the 1 sources of the light gases are investigated, identified and corrected. Further details on the rationale behind the inspection tests and control of the number of vents are given in Appendix E. 4.2.3.2 Array Analysis Calculations have been performed by BNFL on an infinite co-planar array of horizontally mounted product 30B cylinders containing 5% enriched UF. All cylinders were assumed to be touching their immediate neighbors along their longitudinal sides and be butted end-to-end. This approach allowed a generic safety case to be analyzed for 30B cylinders rather than considering each product cylinder storage area on a case-by-case basis. The case of a single cylinder being moved in/out/over the infinite array was also analyzed. This simulates actual storage area operations where criticality safety procedures prohibit double stacking and the movement of more than one 30B cylinder at a time (see Appendix G). The procedures also prohibit any product cylinder being placed in a storage area unless it can be shown that either: 1 A. The overall hydrogen to uranium (atomic) ratio does not exceed unity; or B. The H/U ratio exceeds unity, but the cylinder contains less than 2 kg of hydrogen. For further details on the basis for these limits see Appendix E. It must be stressed that the above are the limits used for criticality control purposes. Most l l carichment customers however specify a product purity of 99.5% which is equivalent to a Louisiana Energy Services 4-11 April 1993 Criticality Safety Engineering Repon Revision 3

H/U limit of 0.088 for a full 30B cylinder if the 0.5% impurity is assumed to be all HF. Furthermore, US transport regulations specify a maximum H/U ratio of 0.088 in cylinders of enriched uranium undergoing transportation. LES will, of course, comply with both these requirements in respect of 30B cylinders dispatched from the CEC. Compliance with this latter customer / dispatch limit will be demonstrated on the Liquid Sampling System for each cylinder and is not to be confused with the "in-process" limit of H/U < 1. The calculations were performed using the above cylinder moderation conditions and assuming infinite reflection by water. The reflection calculations were performed at different densities to simulate flooding and snow, the latter providing the worst case. The results showed that the array is suberitical under all operational and weather conditions. The above mentioned criticality analysis is well supported in the literature. A Criticality l Safety Analysis of a 30 in. UF. cylinder is given in Reference 1. His analysis using the i ANISN code indicates a k., of 0.80,0.825 and 0.90 for UF, with enrichments of 4.5%,5.0% and 6% when atomic H/U is at a value of 0.088 (equivalent to - 570 grams of hydrogen in a full cylinder - see Appendix E). The criticality safety of close packed cylinders was experimentally verified by ORNL. An array of seven close packed 30B cylinder filled with 4.5% enriched UF., standing on end was submerged in water. His configuration is the most reactive. The tank tests showed no significant neutron multiplication. Large cylinder array analyses have also been made by ORNL (Reference 1). Such analyses modeled the authorized contents of 30B cylinders as 38 in. diameter spheres, each sphere reflected by 1 in. of water, and each sphere occupying a cubic cell space of 147.9 cubic feet. This cubic cell space corresponds to the effective volume of a 30B cylinder in a close packed i array. The analyses assumed moderation of the UF at H/U = 0.5. The results of the analysis l indicated that suberitical spheres (k,, = 0.222) arranged in a cubic array require 840 such spheres to achieve criticality. A k,, of 0.90 is yielded by 380 spheres in such an array. His number of cylinders may then be conservatively placed in a subcritical, umeilected cubic array. Placing an equivalent number of cylinders in a co-planar array (as used for storage at the CEC) or double stacked array with spacing between rows to accommodate cylinder handling vehicles would have a substantially smaller k,,. Louisiana Energy Services 4 12 April 1993 Criticality Safety Engineering Repon Revision 3 l -~.

He criticality calculations presented here and the above related analyses / experiments demonstrate that criticality cannot occur even in large arrays of moderation controlled product cylinders. s.- 4.2.4 Summary 1 l l For a criticality to occur in a product cylinder, one of the following two chains of independent events must occur-A. Scenario A 1. A moderator is enoneously present in the cylinder when it is delivered to site, and 2. The cylinder weighing test (2 independent weighings with cylinder rotated through 180 degrees) fails to detect the moderator, and 3. The vacuum test fails to detect water, or The visual test fails to detect oil or hydrates, and 4. The on-plant pressure rise test fails to detect the moderator (water or oil only). These multiple unrelated failures adequately satisfy requirements for meeting the l double contingency principle. B. Scenario B \\ l. A significant leak occurs on the product cylinder, and \\ 2. The multiple pressure sensors on the UF supply line to the cylinder fail to detect the abnormal pressure, and 3. The operator violates the criticality safety instruction limiting the number of vents allowed per product cylinder, and 4. Power sensors in the cascades fail to react to the increased pressures in the product take-off lines and to the increased vent frequency. These multiple unrelated failures also adequately satisfy requirements for meeting the l double contingency principle. l l L/ l Louisiana Energy Services 4-13 April 1993 Criticality Safety Engineering Report Revision 3

4.3 DESUBLIMERS 4.3.1 Normal and Abnormal Operations Special Venting Systems are used to remove gaseous impurities from cylinders containing UF.. These systems basically consist of a desublimer vessel,i.e., cold trap, followed by chemical traps and a vacuum pump (see Figure 2). The Cylinder Venting Systems are further described in SAR Section 6.3, Enrichment and Other Processing Systems. Desublimers are used in the Venting Systems for the feed, product and product blending cylinders; cylinders for depleted UF. are not normally vented. Operation of the desublimers is basically the same for venting any UF cylinder. The Feed Cylinder Venting Systems also serve to vent the cylinders containing depleted UF., if required. Each feed cold trap has four tube desublimers because of the more extensive feed venting requirements. De Product and Product Blending Systems have single desublimer vessels. The Feed Cylinder Vent System desublimers are safe from cdticality by diameter (16 in.) with the chemical forms present at enrichments $; 2 wt%. Feed and depleted UF enrichments are always $;0.711 wt%. De Feed and Product Venting Systems are totally isolated from each other so mixing of enrichments is impossible. Therefore, no further consideration is given to criticality analysis of feed desublimers. l Small quantities of air and HF, from the reaction of moisture in the air with UF., enter the enrichment process via small leaks in the system. Dese and any other light gases preferentially collect in the product stream and enter the product cylinders. They also may enter product cylinders during blending operations. De gases are vented when the pressure in a product cylinder reaches approximately 450 mbar. At this pressure, the product cylinder is isolated from the process and vented into a cooled-evacuated desublimer. The desublimer is isolated from the remainder of the Vent System. When the desublimer pressure reaches 50 mbar, the valve to the cylinder is closed and the UF carried over is solidified in the cold desublimer, leaving the air and HF in the gaseous phase. Typically, - I kg of UF. is carried over into the desublimer in each venting operation. De light gases in the isolated desublimer are then vented (down to a desublimer pressure of 2 mbar) through chemical traps, via a vacuum pump, to a GEVS. The chemical traps remove HF, any trace quant ties of gaseous UF and entrained vacuum pump lubricant prior to i venting to the GEVS. The cylinder venting cycle is repeated until the required light gas pressure in the cylinder is achieved. i Louisiana Energy Services 4-14 April 1993 Criticality Safety Engineering Report Revision 3 i __._-,,,4

~ As indicated in Figure 2, the desublimer is a 16 in diameter vessel 17 ft. 5 in. long (19 fL 8 in. flange to flange). The vessel is wound with separate external closed loop heating and cooling coils. Freon is used as the heat transfer medium. The vessel and coils are surrounded by insulation enclosed within an outer metal jacket. The annulus between the desublimer and the outer jacket is blanketed with dry nitrogen to exclude moisture. The process inlet and outlet pipes are fitted with valves plus backup valves. Pressure and I temperature sensors monitor and control the operations as detailed in SAR Section 6.3. l For each cylinder venting sequence, a conservative estimate of 4 kg of UF. carryover is l i logged into the desublimer UF material balance. The carryover is periodically checked by weight decreases in the product cylinders as indicated by load cells. When the desublimer is filled to the administrative processing limit of 100 kg UF., as indicated by the material balance, the desublimer is heated and its UF. contents are transferred to a product collection cylinder. The volumetric capacity of the desublimers is 3700 kg of solid UF.. Abnormal conditions and their consequences include the following: A. A leak in desublimer vessel - The insulated area between the desublimer and the outer shell is blanketed with nitrogen (not a moderator) which would leak inward to the [s lower pressure desublimer. This would be detected by the pressure sensors, the desublimer inlet valve would automatically close and further venting of the cylinder would be prevented. B. Loss of desublimer cooling - Results in the inability to condense UF, at low pressare. Detected by redundant coolant temperature sensors and desublimer pressure sensors. Desublimers inlet valve closes automatically and further venting of the cylinders is prevented. C. Loss of desublimer heating - Results in the inability to transfer desublimer contents to product collection cylinder at operating pressure / temperature. Detected by redundant heater temperature sensors and desublimer pressure sensors. No criticality significance. D. Exceeding the administrative processing limit of 100 kg UF - The operational fill limit is 500 kg UF., Beyond a limit of > 1000 kg, heating the UF. could result in mechanical deformation of internal parts of the desublimer vessel. Weights are known from estimated carryover and cylinder weight decreases. If the limit is exceeded, procedures require slow discharging of the UF, to below 500 kg with moderate heating. ( Inuisiana Energy Services 4-15 April 1993 Criticality Safety Engineering Report Revision 3

E. Loss of nitrogen blanket makeup gas - The annulus chamber would come to equilibrium pressure at temperature. Pressure and flow sensors and alarms indicate the upset condition. No criticality significance. F. Loss of pressure or temperature sensors - The loss of these instruments would interrupt ventmg operations until repaired. The loss of these sensors would be apparent from readouts in the control room. l G. Failure (open) of the outlet valve from a desublimer to..s chemical traps - An alarm would sound and venting of the desublimer would only continue if the manual backtr, valve was not closed. Limit switches detect valve positions. Continued venting fior i a hypothetical double failure would result in some additional carryover of UF, to the chemical traps. H. Failure (closed) of the outlet valve - Light gases cannot be vented from the desublimer until the valve is repaired. I. Failure (open) of the inlet valve from the cylinder to the desublimer - Limit switches detect inlet valve position and alarm is given. Procedures require that the associated manual valve be closed in this event. J. Failure (closed) of the inlet valve - Prevents -ntmg the cylinder to the desublimer until it is repaired. K. Leak in piping between the valves and the desublimer - Introduces air into desublimer O until pressure equilibrium is reached with the atmosphere. This condition would be detected by redundant pressure sensors and the vessel automatically isolated by closing inlet and outlet valves. This condition introduces some moisture from humidity in the air, which reacts with the UF forming HF and UO F (possibly hydrated UO F )- 2 2 2 2 4.3.2 Criticality Safety Discussion The chemical traps, UF pumps and piping are critically safe by geometry. Their criticality considerations are discussed in Sections 4.4,4.6 and 4.7, respectively. This section discusses the criticality considerations for the product desublimer vessels. Calculations (see Appendix B) show that the desublimers are not safe by geometry at the 5% U235 enrichment level (diameter 40 cm, safe diameter: 25.4 cm). Therefore, criticality safety is ensured through moderator control. The moderator of concern for criticality safety is hydrogen, either in the form of water or HF, which results from the reaction of UF. with water. HF can only enter the desublimer as an impurity in the product UF. A direct ingress of water (in liquid form) into the desublimer is Inuisiana Energy Services 4-16 April 1993 Criticality Safety Engineering Repon Revision 3

excluded by design. There is no water piping in the vicinity of the Product Take-off and f Blending Systems. Furthermore, no water can condense or freeze on the thermal insulation t surrounding the desublimer tube because it is encapsulated in a gas tight housing filled with dry nitrogen with 20 mbar controlled (and alarmed) overpressure. Water can only enter the desublimer as moisture via an air leak in the short unjack'eted inlet and outlet pipes of the desublimer. Once inside the desublimer the moisture reacts with the UF., forming UO F and HF. Although there is always excess UF in the desublimers, it 2 2 cannot be ruled out that hydrates of UO F can be formed (Reference 3). For reasons 2 discussed in Section 4... ;alculations on hydrated UO F assume its crystalline form is UO F 2 2 2 2

• 3.5 H 0.

2 The criticality case presented in the following covers: l A. Moderation via HF entering with the product UF.. B. Moderation via HF gas produced from inleaking moist air. C. Moderation via H O in the form of hydrated UO F produced as a result of inleaking 2 2 2 moist air. D. The array of product desublimers and nearby 30B cylinders. g k 4.3.2.1 Moderation Via HF Impurities In The Product Flow HF enters the desublimers with the product cylinder vent gas. Assuming the product vent gas is pure HF at 1000 mbar pressure, the calculated maximum possible quantity of hydrogen entering the desublimer per vent is 33 grams (see Appendix F). The minimum quantity of hydrogen required to produce a criticality is 3100 grams (see Appendix F). No significant quantity of HF can be retained in the desublimer because it is completely removed during desublimer venting. The desublimer venting pressure is < 2 mbar and the vapor pressure of I condensed HF is about 13 mbar at -70 C. Occlusion of HF in desublimed UF does not occur. This lack of retention of HF in the UF has been demonstrated over many years of operation of the desublimers in URENCO plants in Europe. Thus, a safety factor ~100 is f assured and criticality due to moderation from cylinder vent gases is not possible. O Louisiana Energy Services 4-17 April 1993 Criticality Safety Engineering Report Revision 3 l l 1

t 1 i i 4.3.2.2 N1oderation Via HF Gas Produced From Air Inleakage { Should water from moisture in air enter the desublimer via undetected leaks in the short unjacketed inlet and outlet pipes it would immediately react with the UF forming HF. As is the case above, the HF would be removed by venting and present no criticality risk. f 4.3.2.3 N1oderation Via Water In The Form Of Hvdrates t For reasons discussed in Section 4.1, it is assumed that if a leak occurs hydrates of the form UO F *3.5 H O are produced within the desublimer. 2 2 2 l Appendix F shows that 21.5 kg of inleaking moisture can be tolerated without a desublimer going critical. This includes a safety factor of 0.45. If this quantity is assumed to leak in f continuously over the 10 year period between desublimer internal inspections, then the air inleakage rate equates to 2.9 mbar 1/s (equivalent to a 14 mbar/h pressure rise rate). This is much larger than the normal light gas flow from one assay unit of 0.1 to 0.3 mbar 1/s. Due to the strict requirements for leak tightness of all UF, Systems and the high awareness of l the operators regarding any irregularities with the enrichment process, it is highly unlikely _ i that such a large leakage will go unnoticed for such a long period of time. For example, during the standby period of the desublimer (controlled at 2 mbar) the outlet valve would automatically open and close frequently which would be noticed in the central ) control room. Such a leak would be immediately apparent to the operator as he monitors l several identical Desublimer Vent Systems installed side-by-side. Moreover, a pressure rise of 14 mbar/h (see Appendix F) would lead to a pressure increase of over 40 mbar during a regular cooling period. Such a large deviation from the normal pressure level will certainly be noticed by the operator before he switches to the next operational step (standby). It can therefore be concluded that the process design and operational control assures criticality safety under these circumstances. 4.3.2.4 Safety Of The Desublimer/ Product Cylinder Array The product desublimers are spaced 8 ft. center-to-center. No product desublimer is closer than 15 ft. to a product cylinder center-to-center. Because of their length the midpoints of the desublimers are offset by approximately 5 ft. from the midpoint of the product cylinders. Louisiana Energy Services 4-18 Apdf 1993 Cnticality Safety Engineering Report Revision 3

I l Calculations (see Appendix B) based on solid angle considerations show that the offset array is critically safe. i 4.3.2.5 Summary The analysis and calculations performed here for the product desublimers show that they are j not safe by geometry. Their criticality safety is based on moderator control. An analysis of different routes whereby sufficient hydrogen to cause criticality could accumulate in the desublimer has shown that only one such route is credible. This route involves: A. A leak occuning on the desublimer on its unjacketed inlet or outlet pipe, and B. The moisture in the leak leading to the formation of hydrated UO F, and 2 2 C. The leak remaining undetected for - 10 years, when 1. It is over an order of magnitude greater than the normal light gas flow from an assay unit, and 2. The desublimer outlet valve would be cycling "open" and " closed" much more rapidly than the conesponding valves of the adjacent desublimers, and 3. The pressure rise in the leaking desublimer during the cooling down phase would be in excess of 40 mbar when a normal pressum drop occurs during cooldown. The desublimers' monitoring and control system instruments are periodically calibrated and functionally tested. l l Thus, at least two independent unlikely events would have to occur before a criticality could conceivably take place: A. A leak would have to occur at a specific place on the desublimer (the barriers against this are the QA procedures applied during the design, fabrication and installation of the desublimer), and B. Administrative procedures aimed at identifying large leaks in the Desublimer System would have to fail totally for a period of time equivalent to half the service life of the CEC (i.e., many different operators on different shifts would have to fail to follow procedures. This is prevented by periodic LES audits of adherence to procedures). Another mitigating factor against such a criticality is that textbooks on UF. chemistry and criticality in UF Systems do not consider hydrates of UO F to be formed when moist air 2 2 Louisiana Energy Services 4 19 April 1993 Criticality Safety Engineering Report Revision 3

t i l i leaks into UF Systems. This analysis has assumed that a 3.5 H O 'iydrate will be formed. 2 C' ( 4.4 CHEMICAL TRAPS l 4.4.1 Normal and Abnormal Operations 1 Chemical Traps are used at the CEC in Mobile Vacuum Pump Sets, UF. Cylinder Venting Systems and the Contingency Dump System. The basic function of the Chemical Traps is to remove HF, small quantities of entrained UF and vacuum pump oil prior to venting the gases to the GEVS for further treatment. The Chemical Traps in the Mobile Vacuum Pump Sets and the UF Cylinder Venting Systems are similar in function and in criticality considerations. l They are both discussed in this section. Chemical Traps in the Contingency Dump System perform a somewhat different function and have other criticality considerations. They are discussed in Section 4.5. All of the Chemical Trapping Systems are discussed in further l detail in SAR Section 6.3, Enrichment and Other Processing Systems. Mobile Vacuum Pump Sets are used to evacuate and leak test piping and equipment which j has been exposed to UF or has been exposed to the atmosphere and is to be used in UF. I service. These Mobile Vacuum Pump Sets are used to evacuate pumping prior to disconnecting or connecting UF cylinders and prior to and after maintenance activities. They are also used in conjunction with sample rigs for evacuating lines and containers prior to UF. sampling and for removing any light gases during sampling the product and tails UF streams. 6 They also can be used to evacuate the cascades during cascade startup. Basically, the Mobile Vacuum Pump Set consist of a trap with a layer of activated carbon and a layer of aluminum oxide, an aluminum oxide trap, a vacuum pump and an activated carbon trap in series. A typical Mobile Vacuum Pump set is shown on Figure 3. The layered trap and the aluminum oxide trap remove UF,,, HF and any oil back diffusing from the vacuum pump. The pump discharges through the activated carbon trap which l removes any entrained oil. The treated off-gases vent to the GEVS for further treatment. l Failures of components, valves or instrumentation can result in the inability of the Mobile Vacuum Pump Sets to perform their intended evacuation function, but do not present a criticality safety concern for the Chemical Traps. The failed equipment is repaired or replaced. Ck Louisiana Energy Services 4-20 April 1993 Criticality Safety Engineering Report Revision 3

l A criticality safety assessment has been perfonned demonstrating the Mobile Vacuum Pump sets are criticality safe. This assessment is included as Appendix J of this report. Once each year, the Chemical Traps are weighed and changed out based on the quantity of absorbed material. The trap contents are removed in a controlled area by personnel using respiration equipment. The removed material is then packaged and transported to the Radioactive Waste Storage Area of the TSA. There the contents are analyzed to determine the uranium content and the U235 concentration, and placed in containers spaced in a critically safe array. For disposal, the containers are shipped to a Low Level Waste Disposal Facility. 4.4.2 Criticality Safety Discussion All of the Chemical Traps in the Mobile Vacuum Pump Sets and in the UF, Cylinder Vent Systems will have a diameter less than the safe diameter given in Table 1 and are thus safe by geometry. 4.5 CONTINGENCY DUMP TRAPS / ( 4.5.1 Normal and Abnormal Operations s Significant deviations from the normal vacuum in the centrifuge cascade can cause damage to the centrifuges. In the event of excess pressure,immediate evacuation,i.e., " dumping," of the cascade UF. inventory protects the centrifuges. Normally the UF is dumped to the Product and Tails Take-Off Systems. If plant power and standby power are both interrupted, and cascade dumping is required, the Take-Off Systems are unable to receive the dumped UF. because they are in the fail safe condition. In this case, a Contingency Dump System is automatically activated. Each cascade of centrifuges is connected to a dedicated independent Contingency Dump Trap. Although it is expected that this system will be seldom used, if ever, it is required to assure protection of the centrifuges. The system is described further in SAR Section 6.3.8, Contingency Dump System. A Contingency Dump System is provided for each cascade and consists of a Chemical Trap and a surge volume. The Chemical Trap consists of a bed of sodium fluoride (NaF) which during dumping forms a complex with the UF. and HF and removes them from the gas stream. u. Louisiana Energy Services 4-21 April 1993 Criticality Safety Engineering Report Revision 3

The dumped gases pass through the Sodium Fluoride Trap, where the UF. and HF are i O removed, into a primary surge volume. The primary surge volumes of the seven cascades in an assay unit are connected to a common assay unit header which acts as secondary surge volume. Two vacuum pumps and their associated oil traps are connected to this common ] header. Valve interlocks ensure that only one cascade primary surge vessel at a time can be vented to the assay unit header. They also prevent interconnection of cascades and direct connection of the cascades with the vacuum pumps. The secondary surge volume is provided downstream to accommodate the remaining gases prior to venting to the GEVS. The oil trap prevents vacuum pump oil from diffusing back into the system piping. The vacuum pump is i used to evacuate the Chemical Trap and surge volumes and to vent the light gases to the GEVS. i Loss of vacuum is detected by multiple pressure sensors and automatically starts the vacuum pump to restore vacuum. Inability to restore vacuum would temporarily jeopardize the l centrifuges until repairs are made. Since the system is seldom used this is a highly unlikely event. 4.5.2 Criticality Safety Discussion Within the ranges of enrichmcat assumed (product up to 5 wt% U235, tails: up to 0.34 wt% U235) the calculated average enrichment of the UF. in the cascade inventory is always less than 1.5 wt% U235 (see Appendix C). The piping and valve arrangement leading to the Contingency Dump System assures that under all conditions of normal or abnormal operation the enriched product inventory will be diluted by the depleted tails inventory. The Contingency Dump Sodium Fluoride Trap is a vessel with an inside diameter of 54 cm. Using a conservative average enrichment of 1.5 wt% U235 and ignoring the poisoning effect of the NaF the safe diameter is 60.3 cm, which is obtained by dividing the critical diameter f for 1.5% material given in Reference 10 by the 1.12 safety factor from Reference 8. Thus. the criticality safety of the Contingency Dump Trap is assumed by its geometry, and it is located in a framework which maintains it in a critically safe array. l i When the trap contents are discharged they are placed in critically safe containers and stored in critically safe arrays. Criticality safety of the vacuum pumps, surge volumes and piping are discussed in Section 4.6 and 4.7, respectively. Louisiana Energy Services 4-22 April 1993 Criticality Safety Engineering Report Revision 3 l

4.6 VACUU31 PU31PS w 4.6.1 Normal and Abnormal Operations l Vacuum pumps are used throughout the uranium enrichment process for the following purposes: A. To maintain the UF at the vacuum required for proper process conditions and to discharge the product and tails UF streams into cylinders for collection. l B. To evacuate equipment and piping which contacts UF. for planned maintenance operations and after exposure to the atmosphere. C. To obtain samples of gaseous UF.. i D. To remove light gases from product and feed cylinders. E. To maintain a vacuum in auxiliary UF handling systems such as the Contingency 6 Dump System. Under normal process operating condinons all of these vacuum pumps come in contact with gaseous UF.. All except the process pumps contact only trace quantities. Fomblin oil is used as a lubricant in all of the pumps to avoid the reactions which occur between UF and the m standard hydrocarbon lubricants. Over time the buildup of impurities in the oil requires normal maintenance operations to remove the pump, reinstall a new or reconditioned pump and reprocess the oil. Subsequently, the used oil is recovered and recycled after removal of the impurities. The Fomblin oil reprocessing treatment is described in Section 4.8. Abnormal conditions in the process are detected by pressure sensors and result in automatic shutdown of the affected pump to avoid damage to the pump or other process equipment. Failure of the pumps results in automatic shutdown of the system. 4.6.2 Criticality Safety Discussion Criticality cannot occur within the pumps handling product material on account of the small free volumes associated with each pump. The total free volume of each type of pump used in the product and product blending systems is always less than 14 liters. The volume of contained Fomblin oil is much less than that. Both of these volumes are well within the safe spherical volume of 26.9 liters for aqueous solutions of uranium enriched to 5 wt% U235 as given in Table 1. The spacings between the pumps in the Cascade Hall and UF. Handling Area will be such that the safe solid ancle criterion given in Table 1 will not be exceeded. C k louisiana Energy Services 4-23 April 1993 Criticality Safety Engineering Report Revision 3

i t J 4 4.7 PROCESS PIPING ) 4.7.1 Normal and Abnormal Operations Outside of the autoclaves, all process piping, valves, vessels, equipment and pumps which j contain UF operate under vacuum. The autoclaves provide secondary containment for all UF components which operate abovu atmospheric pressure. Desublimation of UF. is possible i at ambient temperature and pressures above 0.725 psia, therefore, all process piping is trace j heated and all valves are contained in heated enclosures if operated at pressures above 0.725 1 j psia. Initially any leaks produced in the process equipment draw atmospheric air into the system. The abnormal process pressures caused by leaks are detected by strategically located redundant pressure sensors and indicated by alarms. Appropriate corrective actions are initiated by the process operator. At certain preset pressure levels corrective actions begin 3 automatically to stop UF flow, isolate equipment or shutdown systems and operations to avoid the release of UF.. J 4.7.2 Criticality Safety Discussion i ( The quantities of UF,, contained throughout the process piping are always small due to the vacuum required for the centrifuge enrichment process. However, none of the product piping is larger than 25.4 cm in. diameter. Furthermore, no adjacent parallel pipe runs for product j material will be pennitted unless they fit into a 25.4 cm diameter envelope. From Table 1, i this is the allowable diameter for an infinitely long cylinder to be critically safe by geometry if filled with a uniform aqueous solution of uranium enriched to 5 wt'7c U235 at the optimum H/U235 moderation ratio. Therefore, all simple process piping is critically safe by geometry I (reference Appendix H for analysis). i Once detailed design of the product pipework systems is complete, a further criticality analysis will be performed of those pipework areas which contain departures from simple i linear pipe runs, e.g., valve frames, hot boxes and areas containing numerous pipe bends and tees. 4 1 4 Q.- Louisiana Energy Services 4-24 April 1993 i Criticality Safety Engineering Report Revision 3

4.8 VACUUM PUMP OIL C/ 4.8.1 Normal and Abnormal Operations UF reacts completely, sometimes violently, with standard hydrocarbon lubricants normally used in vacuum pumps. Fomblin oil, a highly fluorinated, inert lubricant is used in the UF. vacuum pumps to avoid these potential reactions. Fomblin oil also has the advantage of being a very poor moderator. A buildup of impurities, primarily UO F and uranium tetrafluoride (UF,) from the reaction of 2 2 UF with impurities in the Fomblin oil, gradually thickens the oil and impairs the efficiency of the pumps. This requires normal maintenance procedures for removal of the pump and reinstallation of another pump. The UF vacuum pumps (critically safe by volume) are l discussed in Section 4.6. For pump maintenance, the Fomblin oil is drained and collected in the contaminated equipment workshop as part of the pump disassembly process. The oil is collected in safe plastic containers for processing. The containers are labelled for tracking through the process and stored in the Fomblin oil storage area in a critically safe array while awaiting processing. Dissolved uranium compounds are removed from the Fomblin oil by adding anhydrous sodium carbonate (Na:CO ) which precipitates the compounds as sodium uranyl carbonate 3 [(Na:)4UO (CO )3]. The oil is then filtered through a coarse screen to remove metal particles 2 3 and small parts which may be present. These are transferred to the Solid Waste Disposal System (see Section 4.10). The filtered oil is heated to 210-220 F and stirred for 90 minutes to speed the precipitation reaction. Then, the oil is centrifuged to remove UF, [(Na:)4UO (CO )3] and various metallic fluorides. The solids removed are also transferred to 2 3 the Solid Waste Disposal System. Activated carbon is added to the oil and the mixture is heated to 215-220 F for two hours. The activated carbon absorbs any trace hydrocarbons contained in the oil. The carbon is removed by filtering the oil through a bed of 30-80 mesh diatomaccous earth. The resulting sludge is transferred to the Solid Waste Disposal System (see Section 4.10). The recovered Fomblin oil is sampled for uranium and hydrocarbon levels. The impurities in the recovered Fomblin oil sample are extracted into carbon tetrachloride (CCI,) in the O Louisiana Energy Services 4-25 Criticality Safety Engineering Report April 1993 i Revision 3

Chemistry Laboratory and the concentrations determined on an infra-red analyzer. The limits for purity of the oil are: l A. Uranium - 50 ppm by volume or 30 ppm by weight B. Hydrocarbons - 3 ppm by volume or 2 ppm by weight Oil that meets the criteria is reused in the UF vacuum pumps; oil that does not is reprocessed. i The recovered Fomblin oil is stored in plastic containers in the general storage area awaiting reuse. The contamination levels of the Fomblin oil are not high (~ 2 wt% U) because the UF. pumps cannot operate with oil thickened by high levels of impurities therefore, spills or errors during the oil reprocessing will not present a serious contamination problem. 4.8.2 Criticality Safety Discussion Fomblin oil within the pumps is critically safe by spherical volume as discussed in Section ~ 4.6. s After removal from the pumps, the oil is put into plastic bottles which are critically safe by volume (safe volume is 26.9 liters at 5 wt% U235) and stored in a safe planar storage array. Processing of the oil through the Fomblin recovery equipment is a small scale batch process. Following each batch of material processed the Fomblin recovery equipment is stripped down and cleaned out. Specifically, all the filter beds will be emptied and cleaned, the glass processing flasks will be inspected for residues and cleaned if necessary and the centrifuge bowl will be stripped out and the precipitation sludge removed. Criticality Safety will be achieved by limiting the mass of oil, processed per batch, to the double batch safe mass (17.0 KgU). O \\ Exuisiana Energy Services 4-26 April 1993 Criticality Safety Engineering Report Revision 3

i i + 4.9 VACUUNICLEANERS l 4.9.1 Normal and Abnormal Operations Specially fitted vacuum cleaners are used to cleanup small spills of UO F from equipment j 2 2 and surfaces after maintenance operations or other releases of radioactive materials. The exhaust from these vacuums are discharged through HEPA filters. l There are no abnormal operations of the vacuums which could affect safety. Large spills of { l radioactive materials are cleaned-up using special procedures developed for the individual situation. l t 4.9.2 Criticality Safety Discussion t The vacuum cleaners are limited in volume to 26.9 liter and an inside diameter to 25.4 cm. i l These constraints assure that any quantity of uranium enriched to 5 wt% U235 in an l optimally water moderated and reflected arrangement is critically safe. 4.10 CONTA5tINATED SOLID WASTES i 4.10.1 Normal and Abnormal Operations Solid wastes are produced in a number of plant activities and require a variety of methods for treatment and disposal. Due to the differences in handling, storage and disposal requirements f the solid wastes are separated into wet solid wastes and dry solid wastes. Wet solid wastes will have as little free liquid as is reasonably achievable, but have no limit as to the percent l of volume which is liquid. Dry solid wastes will also have as little free liquid as is j reasonably achievable and liquids will not exceed 1% of the volume. All of these wastes are segregated, monitored for radioactive contamination, identified, stored and treated or prepared for shipment. Industrial, hazardous, radioactive and mixed wet solid wastes are segregated. l Industrial wet solid waste is collected in marked receptacles throughout the plant. The types i of industrial wet solid wastes normally generated in the CEC are wet trash, oil filters, resins from the machine cooling water polishers and from the utility water system softener, and miscellaneous wastes. If the waste contains free liquids, it is dewatered before it is put m a receptacle. Industrial dry solid wastes normally generated in the CEC are trash, silica gel desiccants used in refrigerant system dryers, salt used to regenerate the Utility Water Softener l Louisiana Energy Services 4-27 April 1993 Criticality Safety Engineering Report Revision 3 l

t Unit, uncontaminated air filters, and miscellaneous materials. Industrial waste from Radiation b Control Areas is collected in plastic bags and taken to the Radioactive Waste Storage Area t for inspection to ensure it is not contaminated. The clean inspected waste and the wastes from areas outside Radiation Control Areas are stored in dumpsters until transported to a local landfill. I Hazardous wet and dry solid wastes are collected separately in plastic bags within containers. When the plastic bags are full, they are scaled, monitored for external hazardous materials and cleaned if necessary. The bags are taken to the Hazardous Waste Area, identified, labelled, recorded and stored until they are shipped offsite to a licensed Hazardous Waste Disposal Facility. Radioactive wet solid wastes normally generated in the CEC are: wet trash from decontamination operations, Fomblin oil recovery sludge, filter elements and centrifuged precipitate from the spent citric processing route. The Fomblin oil recovery sludge results from the absorption of hydrocarbons on activated carbon and diatomaceous carth and may contain trace amounts of uranium (see Section 4.8). The waste is shipped offsite to a licensed Central Volume Reduction Facility for treatment and ultimately disposed of at a licensed Low Izvel Radioactive Waste Disposal Facility. s The contents of the Spent Citric Acid tank are precipitated in the LWD Reaction tank and the suspended precipitate separated from the remaining liquid in the LWD Precipitation Centrifuge. This precipitate is then collected in safe-by-geometry containers and transferred to the Radioactive Waste Storage Area ready for shipping to a licensed Low Level Radioactive Waste Disposal Facility. Radioactive dry solid wastes normally generated in the CEC are trash from decontamination operations, activated carbon, aluminum oxide, sodium Huoride, pre-filters, HEPA filters, dry power from the LWD dryer and miscellaneous materials. Activated carbon is used to remove UF from exhaust gases in a number of systems. Aluminum oxide is used to remove HF, traces of UF. and vacuum pump oil from exhaust gases in a number of systems (see Section 4.4). NaF is used in the Contingency Dump System to remove UF and any HF from the cascades (see Section 4.5). Vessels containing 6 spent activated carbon or aluminum oxide are removed, packaged and talzn to the Radioactive Waste Storage Area. There the chemicals are removed from the vessel and put k Louisiana Energy Services 4-28 April 1993 Criticality Safety Engineering Repon Revision 3

into separate containers. Each container's contents is analyzed to determine the uranium and w the U235 content. The containers are monitored for extemal contamination, labelled and stored until shipped to a licensed Low 12 vel Radioactive Waste Disposal Facility. [ Contingency Dump Traps are designed to operate for the life of the facility without the necessity to replace the NaF charge. l Pre-filters and HEPA hlters are used in the GEVS, the TSA HVAC system and in fume hoods to remove tmce HF and uranium compounds from the exhaust airstream. When the filters are loaded, as detennined by differential pressure, they are removed and sealed in plastic bags by personnel using respiration equipment. The filters are taken to the Radioactive Waste Storage Arca and measured to determine the quantity and isotopic l distribution of uranium. The plastic bags are monitored for external contamination, cleaned if necessary, and labelled. The filters are sent to a licensed Central Volume Reduction Facility for treatment before disposal in a licensed Low 12 vel Radioactive Waste Disposal Facility. Contaminated metallic wastes are either decontaminated onsite or shipped to a licensed decontamination vendor. The decontaminated metallic wastes are disposed of as regular scrap once the contamination level has been reduced to an acceptance level. Otherwise, the wastes are packaged, labelled and stored until shipped to a licensed Low Level Radioactive Waste i Disposal Facility. Some items may be compacted onsite prior to shipment. All potentially radioactive wastewater from the Precipitation Process, the Decontamination Facility, the Laboratories, the Laundry, the TSA and UF Handling Area floor drains is 6 evaporated in the LWD Dryer. The solid arising from the dryer is in the form of a dry powder which is drummed directly at the Dryer. The drums will be held in the Radioactive Waste Storage Area and the uranium and U235 content determined prior to shipment to a licensed Low Level Radioactive Waste Disposal Facility. Abnormal operations in the handling and packaging of contaminated solid wastes could result in small spills of low level contaminated materials but this presents no significant hazard. 4.10.2 Criticality Safety Discussion Contaminated solid wastes are segregated by type and placed in labelled containers; each limited to both a mass of uranium less than 4.6 kgU and of such dimensions that the overall areal density of less than 4.6 kgU per sq. ft. is achieved. The safe areal density of less than i t Louisiana Energy Services 4-29 April 1993 Criticality Safety Engineering Report Revision 3

~ i -I 4.6 kgU per sq. ft. can be derived from Figure 17 of Reference 10 and applying the mass l safety factor of 2.3. 4.11 CONTAh11NATED AQUEOUS WASTES AND ASSOCIATED PROCESSING l EQUIPMENT 4.11.1 Normal and Abnormal Operations l Aqueous wastes which are contaminated with uranium or potentially contaminated with l l uranium are either put into containers with a volume of 26.9 liters or less and stored in safe arrays or are collected in tanks located in the UF. Handling Area and the TSA. The uranium loading of the tanks is limited on an individual or group basis to less than the safe mass,17.0 l kgU. l A detailed description of the Liquid Waste Handling Systems, including flow diagrams, is given in SAR Section 6:4.14. The Aqueous Liquid Waste Collection tanks and Processing Equipment shown schematically l in Figure 4 include: A. Collection tanks 1, Effluent Collection tanks Units 1,2 & 3 l 2. Effluent Collection tanks, TSA 3. Spent Citric Acid tank f 4. Decontamination Effluent tank j 5. Effluent Monitor tanks j 6. Laundry tank l B. Processing Equipment 1. LWD Reaction tank and LWD Centrifuge l 2. Dryer Feed tank l 3. LWD Dryer Package j 4. Demineralizers l S. Decontamination Citric Baths and Rinse Baths l 6. Degreaser i Each of the Collection tanks is sampled at a frequency appopriate to its predicted uranium l accumulation rate and is fitted with level sensors and alaans. When full, or the Louisiana Energy Services 4-30 April 1993 Criticality Safety Engineering Report Revision 3 l l

l I administrative concentration limits are reached, each tank's contents are mixed using its transfer pump in the recycle mode, or using a tank mixer. The criticality control pdnciple is that each individual tank, or in some cases a group of tanks, is limited to the safe mass of uranium (17.0 kgU). The sections below provide a detailed discussion of the criticality control pdnciple applied to each tank. The process route is shown schematically in Figure 4. The Citric Baths are emptied into the Spent Citric Acid tank. The Spent Citric Acid tank is processed through the Reaction tank and the precipitate centrifuged to produce a precipitate and supernate. The latter is pumped to the Decontamination EfDuent tank. This tank, when not dealing with the centrifuged liquid, is available to accept the decontamination facility I rinse water. Subject to satisfactory sample results the contents of the UF Handling Area Effluent l Collection tanks, TSA Ef0uent Collection tanks, Decontamination EfDuent tank, Laboratory tanks and Laundry tank are pumped into the Dryer Feed tank. The procedure will be managed in relation to tank capacities. The contents of the Dryer Feed tank will be processed c through the Dryer to produce a dry powder product and a distillate. This distillate will be routed to the Effluent Monitor Tanks for sampling and, subject to satisfactory sample results. l discharged to the Sewage Treatment System. The facility to recirculate the Effluent Monitor tanks through demineralizers or to return the contents to the Dryer Feed tank is available. The tanks are located so that the K,, solid angle relationship shown in Table 1 is satisfied. Where necessary the tanks are isolated by shield walls of 1 ft thick,147 lb/cf concrete. Overfilling of any of these tanks is indicated by the level sensors and alarms, so corrective actions can be taken. Proper pump operation is indicated by pressure measurement and overload protection devices. l Each of the tank subsystems and pieces of processing equipment, any unique abnormal operations and their speciDc criticality safety aspects are desc ' bed in detail in the following l sections. A criticality analysis has been performed of the waste liquid process and collection tanks and is included in Appendix I. r l 1 l m Louisiana Energy Senices 4-31 April 1993 Criticality Safety Engineering Report Revision 3 i g. ---- -,m

4.11.2 Effluent Collection Tanks, UF, Handling Area /^ k 4.11.2.1 Normal and Abnormal Operations t l l There are six Effluent Collection tanks in the UF Handling Area. Each of the three plant j units in the CEC has two dedicated tanks located in that unit's efDuent pit. Each tank is l gravity fed and has a capacity of 1060 gallons. The tanks are mounted vertically. l The liquid wastes collected in Units 1,2 and 3 Effluent Collection tanks include: 1 1 A. Water. which collects in floor drains and discharges to the tanks from floor and area washdowns during normal cleaning operations, but not from decontamination operations. B. Water from service sinks used for handwashing. Previous URENCO operating experience has revealed that typical contamination levels in these tanks are usually < 1 ppm uranium. 4.11.2.2 Criticality Snfety Discussion Tank criticality safety is assured by administrative procedures which prohibit the discharge of liquids used to clean-up radioactive spills into drains which feed to these tanks and by administrative limits on the total mass of uranium contained in each tank. tank contents are sampled weekly. A criticality accident could not occur in these tanks during normal operations when only trace quantities of uranium enter the liquid effluents. For sufficient uranium to be available initially requires an accident which results in a large release of UF. specifically from the product cylinders in the UF. Handling Area. This material must then enter the tanks through the drains via a violation of cleanup procedures or a simultaneous water leak. Furthermore, a criticality accident is mitigated by the following: A. Except during blending or sampling the product cylinders contain UF, in solid form under vacuum. A breach of containment would result in a small deposit of solid UF. and solid UO F from reaction of UF with the available moisture in the inleaking air. 2 2 To enter the Effluent Collection tanks these solids would have to be deliberately Louisiana Energy Services 4-32 April 1993 Criticality Safety Engineering Repon Revision 3

I washed into the drains. This operation on a significant quantity of uranium would O generate HF and present an intolerable condition for the operator. Special administrative procedures are used for such cleanup operations and the disposal of the resulting wastes after a release of radioactive material. In over 30 years of operation of URENCO plants in Europe there has never been an accident which has resulted in any significant quantity of uranium being deposited on the floor of any process area. B. The product blending and sampling cylinders are contained within autoclaves whenever their UF. contents are heated or in liquid form. Releases of any kind are contained within the secondary confinement provided by the autoclaves. The autoclave door is interlocked to prevent opening under upset conditions. Instrumentation is provided which detects leaks via the abnormal pressures produced in the autoclaves. If the autoclave door interlock fails, procedures prohibit the operator from opening the door when upset conditions are detected by the instruments. j If all of these contingencies fail and UF. is released to the UF Handling Area it would quickly solidify. It would have to be deliberately washed into the floor drains to enter an Effluent Collection tank. I The application of the double contingency principle is satisfied because multiple unrelated failures are requimd before a criticality is possible. p t( 4.11.3 Effluent Collection Tanks Tecimical Services Area (TSA) 4.11.3.1 Normal and Abnormal Operations There are two Effluent Collection tanks located in the TSA Effluent Pit. The tanks are vertical with cone-shaped bottoms and gravity fed. Each tank has a capacity of 650 gallons. Liquid wastes collected in the TSA Effluent Collection tanks include: A. Floor drains from the Chemistry Laboratory. These are not expected to contain other than trace quantities of urani.1. B. Drains from the Personnel Decontamination Area. These may contain minor quantities of uranium from infrequent use for worker decontamination. C. Drains from the Contaminated Equipment Workshop used for disassembleming equipment which is slightly contaminated. Contains minor quantities of uranium from contaminated components. Louisiana Energy Services 4-33 April 1993 Criticality Safety Engineering Repon Revision 3

i D. Drains from the Radioactive Waste Storage Area. May contain minor quantities of l uranium from surface contamination of waste packaces. Normally no contamination is expected. j E. Drains from other storage areas. Storage is provided for normal plant materials, no contaminated materials are stored in this area and no contaminated efHuent is expected. j F. Drains from the Decontamination Workshop floor. Contaminated liquids from the Decontamination Workshop drain to other dedicated tanks (see SAR Section 6.4.14). l Some contamination could conceivably enter the floor drains due to minor spills of these liquids. Normally no contamination is expected. l G. Drains from the Laundry door. May contain minor quantities of uranium from undetected trace contamination on personnel and clothing. Normally no contamination is expected. H. The LWD Dryer requires a daily Dush. It has been estimated that the quantity of uranium entering the TSA Ef0uent Collections tanks with the flush water is approximately 20 g per year (Reference 13). The total quantity of liquid efDuent discharged to the TSA Effluent Collection tanks is estimated to be approximately 750 gallons per week (Reference 13). i Potential abnormal conditions could include erroneous discharge of sample solutions or Fomblin oil sludge solutions to the drains, erroneous dumping of decontamination solutions into drains or sinks in violation of procedures, leaks in liquid storage containers in the Radioactive Waste Storage Area or leaks in the tanks discharging into the drains. 4.11.3.2 Criticality Safety Discussion A criticality accident cannot occur with the available quantities of uranium in the effluents which are discharged into the TSA EfDuent Collection tanks under normal operations. To study all possibilities for creating an accidental criticality, the total uranium available annually from all sources which enter the TSA in any form was estimated from previous operating experiences (Reference 4). These sources were then studied to determine if a quantity of 3 uranium sufficient for criticality could accidentally enter the TSA Effluent Collection tanks. j The sources and their estimated uranium content are given in Table 3. I 1 l l The uranium which is not shipped out of the TSA unchanged is all in aqueous solutions. apart from the uranium in Fomblin oil. The aqueous solutions are processed in the LWD Dryer to l i N Louisiana Energy Services 4-34 April 1993 Criticality Safety Engineering Report Revision 3

Ning the uranium into the solid state either as dry powder or potassium diuranate. Further details of the LWD Dryer are given in SAR Section 6.4.14. I Table 3 shows that the total annual estimated quantity of uranium from all sources entering the TSA is 378 kg. Of this total,334 kg (~ 88%) enters in solid form as customer samples or i contaminated solid wastes and leaves the TSA unchanged. It is inconcei. ; ; >%t these materials could enter the TSA Effluent Collection tanks in liquid form. l The remaining 44 kg (~ 12%) are either put into solution as a result of processing, or precipitated as solids and sent to the solid vaste disposal system, e.g., the 7 kg of uranium from Fomblin oil." The following is a description of the methods adopted for ensuring that these solutions or the unprocessed Fomblin oil sludge cannot accidentally enter the TSA Effluent Collection tanks. The remaining UF. samples (2 kgUlyr) after hydrolysis (Reference 4) are stored separately in volumetrically safe containers (s 26.91) in a critically safe array pending processing. To assure the criticality safety of the TSA Effluent Collection tanks from inadvertent entry of these solutions and the unprocessed Fomblin oil sludge the following controls are used: w ( A. Procedures for handling, storage and treatment of these solutions and the unprocessed Fomblin oil sludge specifically prohibit discharge to any drains. In addition, normally collected wastes will fill a TSA Eftluem Collection tank in approximately one week. Inadvertent discharge (against pacedures) of these solutions and the unprocessed Fomblin oil sludge to the drains as they are normally produced in this timeframe can not provide sufficient uranium for a criticality. The inadvertent discharge to the drains in quantities sufficient for criticality requires stockpiling (against procedures) almost 5 years worth of normal production of Fomblin oil sludge or approximately 17 years worth of sample solutions. B. The containers of the UO F solutions and the unprocessed Fomblin oil sludge are 2 2 stored in safe-by-shape, non-draining curbed areas. C. The total allowable uranium content in stored sample solutions and the unprocessed Fomblin oil sludge is limited to a safe mass. Solutions fror_ the decontamination of equipment in the Citric Acid Baths (35 kgU/yr) are stored in the Spent Citric Acid tank (see Section 4.11.4). To assure the criticality safety of the TSA Effluent Collections tanks from inadvenent entry of these citric acid solutions the J following controls are used: ~ Louisiana Energy Services 4-35 April 1993 Criticality Safety Engineering Report Revision 3

A. The Citric Acid Baths can only discharge to the Spent Citric Acid tank. The Spent i Citric Acid tank can only discharge to the Decontamination EfHuent tank via the LWD Reaction tank and Precipitate Centrifuge. The Decontamination Effluent tank is connected by common disenarge to the TSA Effluent Collection tanks. The common f discharge pipmg is equipped with back flow presenters to prevent any tank discharging into the outlet manifold from backflowing into another tank. To get citric acid solution into a TSA Effluent Collection tank would therefore require multiple procedural and technical failures to occur simultaneously. See SAR Figure 6.4-34 for the flow schematic. 1 B. The Citric Acid Baths and the Spent Citric Acid tank are located in non-draimng curbed areas. 1 C. The total allowable uranium content in the two Citric Acid Baths combined or the Spent Citric Acid tank is limited to a safe mass (s 17 kgU). i l Wash water from the Laundry is collected in the Laundry tank. To assure criticality safety of j the TSA Effluent Collection tanks from inadvertent entry of the Laundry wash water the following controls are used. See Section 4.11.6 for further details of the Laundry System. l A. Although the Laundry tank is operated to the criterion of not containing more than a safe mass of uranium, it is not expected that anything like this limit will ever be approached. The 2000 gallon Laundry tank is expected to be cycled every 1% weeks and to handle no more than 0.1 kgU in a year (Reference 16). B. The Laundry tank is connected to the TSA Effluent Collection tank via the common discharge manifold as described above for the Decontamination Effluent tank. The piping system is specifically designed to prevent back flow leading to mixing of water between these tanks. To verify the effectiveness of the baniers and controls the U235 concentration of the TSA Effluent Collection tanks is measured when half full and immediately prior to discharge. The application of the double contingency principle is satisfied because of the multiple unrelated failures which are required before a criticality is possible. l Louisiana Energy Services 4-36 April 1993 Criticality Safety Engineering Repon Revision 3

4 4.11.4 Citric Acid Baths And Spent Citric Acid Tank, TSA 4.11.4.1 Normal and Abnormal Operations i l The Spent Citric Acid tank is located in the TSA. The tank is vertical with a cone-shaped bottom. It has a capacity of 650 gallons. Liquid wastes collected in the Spent Citric Acid tank include liquid effluents originating from Citric Acid Baths which are used to i decontaminate components used in the enrichment process. There are two baths: one with a capacity of 450 gallons and one with a capacity of 100 gallons. The large bath is used for decontaminating large components and the small bath for smaller components such as valves. The following procedures are used to control the uranium loading of the Citric Acid Baths: A. Component Consignment 1. All consigning depanments are required to consign items for decontamination via a formalized route. 2. As part of the procedures, the consigning department is required to inspect each component, and to indicate the records which accompany the component, if there is any loose breakdown or blockage. B. Component Receipt Dry Components 1. On receipt of the component into the Decontamination Workshop,its documentation is inspected. 2. If the component is indicated to have no gross contamination or blockages, it is physically re-inspected by Decontamination Workshop personnel to confirm this to be the case. Where the component has concealed surfaces, these are also inspected. If the component passes this second check, it is accepted for decontamination. 3. If the component is indicated to have either a blockage or gross contamination it is forwarded to the Contaminate Equipment Workshop. In the Workshop, f any gross contamination / loose breakdown / blockages are removed using hand tools. Once 'he component has been " cleared" it is accepted for decontamination. (Decontamination equipment and procedures for oily components are described in Section 4.13.) C. Comporent Decontamination i 1. Components will have been released for decontamination, from receipt procedures into the Decontamination Workshop, as described above. Compcnents will be either 1) dry without gross contamination or blockage as jO Louisiana Energy Services 4-37 April 1993 i l Criticality Safety Engineering Report Revision 3

delivered to the Decontamination Workshop; or 2) dry without gross contamination or blockage after having been cleared out by workshop staff or

3) dry without gross contamination or blockage after having been processed through the degreaser.

l 2. The Citric Acid Baths are collectively uranium mass limited to the safe mass, i for criticality control purposes. This is effected by estimating the uranium loading per basket from experience of routinely processing standard plant components on a repetitive basis and by canying out component receipt procedures to ensure that any gross contamination or blockage is removed from components prior to their introduction to the Citric Acid Baths. The Citric Baths are also routinely mixed, sampled and analyzed for U235 content. 3. This procedure has been performed successfully for many years within URENCO. The plant will operate intermittently as required. Daily uranium loadings on working days are anticipated to be in the ten's to low hundred's of grams uranium region compared with the safe mass of 17 kgU. l 4 The bath is emptied before the uranium content exceeds the safe mass limit. Both baths are equipped with ultrasonic agitation, electdc heaters, manually operated spray no7zles and rim vent extraction which exhausts to the GEVS. Recirculating pumps provide mixing for obtaining representative samples. Samples are taken at least weekly, depending on usage and estimated uranium content, and prior to discharge to the Spent Citric Acid tank (see Section 4.11.4.2). Taking 33.7 kg per year uranium arisings in the Citric Acid Baths, from Table 3, and a concentration of 6.35 gU/1 from Section 4.11.4.2 gives a yearly citric acid usage of 53071 or 1401 gallons. The baths are operated to 85% of this limit (see Section 4.11.4.2). The annual citric arisings estimate is 1650 gallons equivalent to an actual uranium concentration of 5.4 gU/1. 4.11.4.2 Djticality Safety Discussion l The Citric Acid Baths are operated to an administrative limit of 6.35 gU/1, which ensures that both Citric Acid Baths combined will contain less than the safe mass (17.0 kgU) for the Spent Citric Acid tank. The estimated quantity of uranium remaining on each component is logged i as the component enters the Citric Acid Baths. This estimate will be based on actual operating experience at the CEC as it becomes available (prior to that on actual operating 1 experience from URENCO plants ii Europe). When the log indicates the quantity of uranium O lxuisiana Energy Services 4-38 April 1993 Criticality Safety Engineering Report Revision 3 l

1 i is 40% of the administrative concentration limit of 6.35 g/1, the bath is mixed, sampled and analyzed. Then the logged value is replaced by the analyzed value. Decontamination i operations are resumed, continuing the log until 70% of the administrative concentration limit is reached, then the bath is sampled again. The logged value is again replaced by the analyzed value and decontamination operations resumed. When the logged value reaches 85% of the administrative concentration limit value the bath is sampled, analyzed and discharged to the Spent Citric Acid tank. The analyzed concentration and calculated total l mass are logged in on the Spent Citric Acid tank log. The uranium loading of the 450 gallon and 100 gallon Citric Acid Baths, considered as a pair, is limited to a safe mass. At the working concentration of the citric solution of 5.4 gU/l the uranium loading of each bath prior to emptying is 9.2 kgU and 2.0 kgU for the large and small citric bath, respectively. t On emptying, the contents of a Citric Acid Bath are pumped to the 650 gallon Spent Citric i Acid tank. A Citric Acid Bath is sampled immediately prior to emptying and the Spent Citric t Acid tank is sampled after every transfer into it and routinely while it holds spent citric acid solution. I The Spent Citric Acid tank is connected uniquely to the two Citric Acid Baths, therefore, e there is no other source of input to the Spent Citric Acid tank. Correct pipework connections j will be confirmed prior to plant startup. l The safety case for the Spent Citric Acid tank is that no more than a safe mass of uranium is allowed within the tank. If filled to capacity with citric solution, at the uranium working limit concentration of the Citric Acid Baths, the uranium loading of the Spent Citric Acid tank will be 13.3 kgU. If for any reason the Spent Acid Citric tank were filled with citric j acid at 6.35 gU/1 the uranium loading would be 15.7 kgU. I i The uranium throughput of the Decontamination Baths is estimated to be 33.7 kgU per year (Reference 14). For the safe mass of uranium to be exceeded in the Spent Citric Acid tank or a Citric Acid Bath, there must be multiple failures of tank management and uranium accountancy procedures in both Citric Acid Baths and the Spent Citric Acid tank. In view of the annual throughput of uranium in the Decontamination Facility these failures would have to occur continually over several months. The double contingency principle is therefore satisfied in respect of the Spent Citric Acid tank and the Citric Acid Baths. I' l i t 1 Louisiana Energy Services 4-39 April 1993 Criticality Safety Engineering Report Revision 3 m.. ..r

f I 4.11.5 LWD Reaction Tank And Precipitation Centrifuge 4.11.5.1 Normal and Abnormal Operations i The TSA is equipped with a 55 gallon LWD Rea; tion tank, KOH Metering System and precipitation centrifuge. l The function of this equipment is to accept batches of aqueous uranic solution into the LWD Reaction tank, to which potassium hydroxide is then added to precipitate the uranium as potassium diuranate. The LWD Reaction tank is provided with agitation to keep the precipitate in suspension. This tank is also provided with a centrifuge through which the contents of the tank are cycled following KOH addition. The centrifuge process is continued until the suspended precipitate has been extracted leaving only liquid in the LWD Reaction tank. Following completion of the centrifuge process the LWD Reaction tank is circulated and sampled for remaining U235 concentration. The LWD Reaction tank contents are then i pumped into the Decontanination Effluent tank. The centrifuge is opened-up and the citric cake is removed. The cake U235 content is determined and the cake is taken to the Radioactive Waste Storage Area where it is stored prior to shipping to a Low Level Waste Disposal Facility. 4.11.5.2 Criticality Safety Discussion There will be four uranium bearing input streams to the LWD Reaction tank / Precipitate Centrifuge System. These streams are: A. Citric solution from the Spent Citric Acid tank comprising 1650 gallons per annum containing 34 kgU (Reference 14). B. Aqueous uranyl solutions from the laboratories comprising 211 x 51 bottles (278 gallons) per annum containing 1572 gU (Reference 15). j C. Aqueous uranyl solutions from the sample bottle cleaning process comprising 1012 gallons per annum containing 1 gU (Reference 14). In the first case, the citric solution is run into the LWD Reaction tank from the Spent Citric Acid tank. In the latter two cases, the solutions are introduced into tie LWD Reaction tank manually. Louisiana Energy Services 4-40 April 1993 Criticality Safety Engineering Report Revision 3

j The safety case for the LWD Reaction tank and Precipitation Centrifuge System is that no more than a safe mass of uranium is allowed to be held in the system at any one time. The U235 loading of the inputs to the LWD Reaction tank and Precipitation Centdfuge System will be determined by sampling. Similar determinations will be made for the uranium loading of the outnuts comprising reaction tank liquid and citric cake. A uranium mass balance will be maintained for the system. Following the processing of a safe mass of uranium, estimated to occur at approximately six monthly intenrals, the Reaction tank and Precipitation Centrifuge will be intemally inspected to ensure that there is no buildup of precipitate. For the safe mass of uranium to be exceeded in the Reaction tank and Precipitation Centrifuge System, there must be multiple failures of tank management and uranium accountancy procedures in both the input streams and in the Reaction tank and Precipitation Centrifuge System itself. In view of the annual throughput of uranium through this system, these failures would have to occur over several months. The double contingency principle is therefore satisfied in respect of the Reaction tank and Precipitation Centrifuge System. I 4.11.6 Decontamination Effluent Tank And Rinse Water Baths 4.11.6.1 Normal and Abnormal Operations There are two rinse baths of 450 gallons and 100 gallons capacity situated in the Decontamination Facility. Their purpose is to rinse wash components which have been treated in the Citric Acid Baths. Both baths are equipped with ultrasonic agitation, electdc I heaters manually operated spray nozzles and rim vent extraction which exhausts to the GEVS. Recirculating pumps provide mixing for obtaining representative samples. Due to the low uranium loading the rinse water baths are sampled weekly when in operation. The Decontamination Effluen* 'ank is a vertical 650 gallon, cone-shaped bottom tank located in the TSA. The tank has two input streams. It may receive rinse water from the Decontamination Facility rinse baths or it may receive centrifuged solution from the LWD Reaction tank. When full the tank is circulated, sampled and transferred to the Dryer Feed tank. O Louisiana Energy Services 4-41 April 1993 Criticality Safety Engineering Repon Revision 3 I i

t 4.11.6.2 Criticality Safety Discussion i The safety case for the rinse water baths is that no more than a safe mass of uranium is allowed to be held in both baths taken as a pair. The annual loading of the rinse water baths is estimated at 84G gU (Reference 4). Although l the uranium holdop of the rinse water baths is very much smaller than that of the Citric Acid Baths the criticality control procedures for the rinse baths mirror those of the Citric Acid Baths. The same logging procedures, mass limits and sampling methods that applied to the Citric Acid Baths are enforced on the rinse water baths to ensure criticality safety and to f apply the double contingency principle. The rinse water baths are sampled at approximately weekly intervals in view of the small quantities of uranium involved. l The safety case for the Decontamination EfDuent tank is that no more than a safe mass of uranium is allowed to be held in the tank. The uranium loading of the Decontamination Effluent,alk will be low and the input streams are characterized as follows: l A. Rinse water from the rinse baths comprising 3300 gallons per annum containing 840 i gU (Reference 14). B. LWD Reaction tank arisings comprising 3115 gallons per annum containing 0.17o of the input uranium, i.e., f (33.7 + 1.572 +0.501) x 0.001 ~ 36 gU (see Section 4.11.5) Prior to each transfer into the Decontamination Effluent tank from either a rinse bath or from the LWD Reaction tank, the rinse bath or Reaction tank will have been sampled inline in accordance with operating procedures. Therefore to exceed a safe mass of uranium in the Decontamination Effluent tank, there must have been multiple procedural failures in the management of the rinse baths or the Reaction tank and the Decontamination Effluent tank. In view of the uranium throughput of the Decentamination Effluent tank, these failures would have to occur over many years. The double contingency principle is therefore satisfied in respect of the Decontamination Effluent tnk. \\ Louisiana Energy Services 442 April 1993 l Criticality Safety Engineering Report Revision 3

4.11.7 Laundry Tank l 4.11.7.1 Normal and Abnormal Operations I The Laundry tank is a vertical 2000 gallon tank with a cone-shaped bottom located in the l TSA. Liquid waste collected in the Laundry tank comes solely from the Laundry washing machine. The washing machine is used to launder clothes used in cleaning and maintenance in the Radiation Control Zones (RCZs). Typical clothing is listed in Reference 16. Laundry is collected in designated containers which are lined with plastic bags. When a Laundry container is full the plastic bag is scaled and taken to the Laundry Room. There the laundry is sorted and surveyed for radioactive contamination prior to washing. Grossly contaminated clothing is not laundered, but discarded as contaminated solid waste and stored in safe containers prior to disposal. The laundry effluent normally contains no uranium or only trace quantities. 'Ihe Laundry tank is sampled weekly and prior to transfer to the Dryer Feed tank. 4.11.7.2 Criticality Safety Discussion The estimated annual throughput for the Laundry tank is 182,000 gallons containing an estimated 0.1 kgU (Reference 16). The safety case for the Laundry tank is that no more than a safe mass of uranium is allowed in the tank. To exceed the safe mass in this tank at least two independent events are required. Firstly, in spite of all automatic and procedural safeguards a significant release of UF to the environment is required. Secordly, clothing would need to become heavily contaminated and then to be loaded into the washing machine in violation of laundry operating procedures. l The double contingency principle is therefore satisfied in respect of the Laundry tank. b l l l louisiana Energy Services 4 43 April 1993 Criticality Safety Engineering Report Revision 3 l

4.11.8 Laboratory Tanks 4.11.8.1 Normal and Abnormal Operations There are three aqueous liquid waste arisings from the Chemistry Laboratories as determined in Reference 15. This section considers the twin 500 gallon Laboratory tanks. It is anticipated that 280 gallons of uranium bearing aqueous solution will arise per year and that this solution will be conveyed in 5 liter bottles to the LWD Reaction tank. A further 49,000 gallons of non-uranium bearing aqueous solution will arise per year for discharge to the Sewage Treatment System. A further 19,000 gallons of aqueous solution will arise comprising primarily equipment / glassware washings from a Laboratory dishwasher. These arisings are stored in a pair of 500 gallon Laboratory tanks prior to sampling and discharge to the Dryer Feed tank. 4.11.8.2 Criticality Safety Discussion The Laboratory tanks process 19.000 gallons of aqueous laboratory arisings per year (Reference 15). A 500 gallon Laboratory tank will therefore fill on average once every 1% weeks. The quantity of uranium processed within the laboratory in one year has been estimated to be 1570 + 1680 + 200 - 3.5 kgU (Reference 15) This uranium comprises - 200 gU as solid radioactive waste,1680 gU as mixed waste and 1570 gU as uranic aqueous waste. None of these waste streams are routed to the Laboratory tanks. The mixed waste is disposed of to a licensed facility (See ER Section 3.3.2.3.2.9) and the solid radioactive waste is stored and characterized in the Radioactive Waste Storage Area i prior to shipping to a licensed Low Level Radioactive Waste Disposal Facility. The uranic aqueous waste is transferred to the Liquid Waste Disposal System Reaction tank. The main contribution to the Laboratory tanks is frora dishwashing of utensils, glassware and equipment which is assumed to amount to 1% of the uranic aqueous waste throughput,i.e.,15 gU. Louisiana Energy Services 4-44 April 1993 Criticality Safety Engineering Report Revision 3 ~ -

Because of the very low uranium loading the Laboratory tanks will only be sampled prior to discharge to the Dryer Feed tank. ] The safety case for the Laboratory tanks is that no more than a safe mass of uranium is allowed within a tank. From the above figures, it can be seen that there is insufficient uranium processed through the Laboratory in a year to accumulate a safe mass. In order to accumulate 17 kgU in a Laboratory tank, firstly, five times as much uranium as is normally processed in a year would need to be in the Laboratory in aqueous solution in the 1 % week cycle time of a Laboratory tank, in gross breach of Laboratory operating procedures, and 4 secondly - again in breach of procedures - all this uranium would have to be transferred to a j Laboratory tank. The double contingency principle is therefore satisfied in respect of the laboratory tanks. 4.11.9 Dryer Feed Tank 4.11.9.1 Normal and Abnormal Operations The Dryer Feed tank is a 5000 gallon tank whose function is to receive feed solutions of lovi ( uranic concentration and then provide a feed to the Liquid Waste Disposal System dryer. The Dryer Feed tank receives solutions from the UF. Handling Area Effluent Collection tanks, the TSA Effluent Collection tanks, the Laboratory tanks, the Laundry tank, the Decontacination Effluent tank and the Effluent Monitor tank. The conuection from the Effluent Monitor tanks to the Dryer Feed tank is provided to allow reprocessing of solutions should the first pass through the dryer be unsatisfactory. The contribution to the Dryer Feed tank from the Effluent Monitor tanks is not considered in this analysis as recycle cannot increase the net uranium feed to the LWD dryer. The Dryer Feed tank is provided with online pH measurement, chemical metering (for pH stabilization), agitation and closed loop recycle. A constant pressure pH stabilized feed is taken from the recycle loop to feed the LWD dryer. See SAR Figure 6.4-35 for the flow diagram. Louisiana Energy Services 4-45 April 1993 Criticality Safety Engineering Report Revision 3

l In view of the low uranium loading of the Dryer Feed tank the tank will only be sampled O' weekly for criticality purposes, although it may actually be sampled more frequently for process control purposes. 4.11.9.2 Criticality Safety Discussion All Dryer Feed solutions have previously been characterized and are summarized as follows: A. Effluent Collection tanks for all three plant units: 33,000 gallons per annum containing 0.02 kgU (Reference 13). B. TSA Effluent Collection tanks: 27,750 gallons per annum containing 0.07 kgU (Reference 13). C. Decontamination Effluent tank. 3300 + 3115 = 6415 gallons per annum containing 840 + 36 = 876 gU (see Section 4.11.6). l D. Laboratory tanks: 19,000 gallons containing 15 gU (see Section 4.11.8). l E. Laundry tank: 182,000 gallons containing 0.1 kgU (see Section 4.11.7). I The above figures give an estimate for the annual throughput of the Dryer Feed tank of l 278,0(X) gallons containing 1.04 kgU, this uranium loading being dominated by the Decontamination Workshop rinse bath arisings. s The safety case for the Dryer Feed tank is that no more than a safe mass of uranium is allowed within the tank. Frem the above figures. consider the situation where all the listed feed tanks are full and ready to feed into the Dryer Feed tank. The normal uranium loadings will be: 3 x 1000 gallon Units 1,2 & 3 Effluent tanks containing 1.65 gU 1 x 650 gallon TSA Effluent tank containing 0.55 gU 1 x 650 gallon Decontamination Eftfuent tank containing 89 gU l 1 x 500 gallon Laboratory tank containing 0.4 gU 1 x 2000 gallon Laundry tank containing 1.1 gU 6800 gallons 92.7 gU i In the highest uranium loading case there will thus be about 100 gU waiting to feed into the Dryer Feed tank. a l Louisiana Energy Services 4-46 April 1993 l Criticality Safety Engineering Report Revision 3 i

In writing the Tank Management Procedures for the TSA it would typically be specified,in these low uranium loaded tanks, that f the uranium loading in a tank deviates from its normal i ( value by a factor of three, then that c:.ts operation is suspended and an investigation performed. In order 10 achieve 17 kgU in the dryer feed tank: A. Some unspecified abnormal event wouhl have to occur on the plant to produce the possibility of gross quantities of uraniam entering the above liquid streams. B. Tank Management Procedures, as outlined in preceding sections, must have failed allowing the gross uranium to enter either the Effluent tanks, the Decontamination Ef0uent tank, the Laboratory tanks or the Laundry tank. Procedures must also have failed if the gross uranium loading is not subsequently detected. The double contingency principle is thus satisfied in respect of the Dryer Feed tank. 4.11.10 Dryer, Dryer Entrainment Separator, Dryer Condenser, Dryer Distillate Tank 4.11.10.1 Normal and Abnormal Operation The LWD dryer package comprises the dryer itself, the entrainment separator, the condenser and the distillate tank. The flow diagram for this system is given in SAR Figure 6.4-35. The LWD dryer takes liquid feed from the Dryer Feed tank. The operation of the Dryer Feed tank is described in Section 4.11.9. The feed stream is processed, in one step, into a dy powder and into a vapor stream. The vapor stream is condensed into a liquid distillate. The dry powder is collected into 30 gallon drums directly at the LWD dryer. The distillate is collected in a Distillate tank at the LWD dryer and subsequently pumped into one of three Effluent Monitor tanks. The uranium in the feed stream is separated into the dry powder. The distillate, subject to sampling, is designed to be discharged to the Sewage Treatment System via the Effluent Collection tanks. The LWD dryer is subjected to a flush out each day, the flush water going to the TSA Effluent Collection tanks. Louisiana Energy Services 4-47 April 1993 Criticality Safety Engineering Report Revision 3

4.11.10.2 Criticality Safety Discussion d The LWD dryer is connected exclusively to the Dryer Feed tank. i.e., the Dryer Feed tank is l the only sourw of feed material for the LWD dryer. Correct pipework connections will be l confinned prior to plant startup. l The uranium content of the LWD dryer feed is known from the Tank Management Procedures of the Dryer Feed tank itself and of the contributors to the Dryer Feed tank as discussed in Section 4.11.2,4.11.3,4.11.6,4.11.7. 4.11.8 and 4.11.9. As discussed in Section i 4.11.9. the uranium throughput to the LWD dryer is estimated to be of the order of I kgU per annum. As the LWD dryer and its associated equipment are operated a uranium balance will be kept. Input data for the balance will be obtained as noted above, output data will be obtained from the Effluent Monitor tanks (see Section 4.11.11) and from characterization of the dry powder drums. This characterization will be performed in the Radioactive Waste Storage Area prior to shipping the drums to a licensed facility. g According to Reference 13, the LWD dryer processes 275,000 gallons of feed per annum f containing 1820 kg of non-uranic dissolved solids and I kg of dissolved uranium. It operates g 250 days per year and therefore processes 1100 gallons of feed per day containing on average 7 kg of solids and 4 gU. The criticality safety case for the dryer equipment is that no more than a safe mass of uranium may be held in the dryer equipment. There are two ways in which 17 kgU could conceivably be heldup in this equipment. Firstly, there is the situation where the uranium concentration in the feed is so grossly in error that the uranium contained in the instantaneous process inventory of the dryer equipment is 17 kgU. The dryer equipment is fed exclusively from the Dryer Feed tank and the possibility of significant deviations from the normal uranium concentration in the Dryer Feed tank has already been climinated at double contingency criteria (see Section 4.11.9). This latter argument also serves to eliminate the possibility of significant deviations from the nonnal uranium inventory being present in the dryer equipment. The free volume of the dryer equipment will be significantly less than the volume of the Dryer Feed tank (Reference 13). The above argument will apply in the case of the normal process inventory, a few tens of rik Louisiana Energy Services 4-48 April 1993 Criticality Safety Engineering Report Revision 3

l l l gallons of feed solution. It will also apply in the event of a malfunction whereby the entire free volume of the dryer equipment has become filled with feed solution. Secondly, there is the situation where a uranium residue might steadily accumulate within the dryer equipment over the long term. This possibility is climinated by, in the first instance, monitoring the dryer equipment uranium balance. A criterion would be set for stopping operations and stripping the equipment down for internal inspection to look for accumulated uranium. A typical criterion would be when the amount of material unaccounted for amounted to several standard deviations of the accuracy of the accountancy measurements. In the second instance, a completely independent check would be in force. This would be that when 17 kgU in total as measured at the Dryer Feed tank, has been fed to the dryer equipment the system is stripped down and internally j inspected for accumulated uranic residue. For 17 kgU to accumulate in the long term in the dryer equipment, both the above independent checks must fail. Therefore, in both the instantaneous and long term cases the double contingency principle is satisfied in respect of the dryer equipment. 4.11.11 Effluent Monitor Tanks And Demineralizers t 4.11.11.1 Normal and Abnormal Operations Three 1500 gallon capacity Effluent Monitor tanks are provided. The tanks form a parallel bank and are designed to sequentially accept the dryer distillate. Once a tank is full it will be circulated and sampled to determine if it can be discharged to the Sewage Treatment System. The allowed limit for uranium concentration is discussed in SAR Section 6.4.14.1.2. Following a satisfactory analysis result from the sample., the tank contents can be pumped to the Sewage Treatment System. If the uranium concentration does not satisfy the discharge limits two courses of action are available. Either the Effluent Monitor tanks inventory can be pumped back into the Dryer Feed tank or alternatively the demineralizers can be used. Louisiana Energy Services 4-49 April 1993 Criticality Safety Engineering Report Revision 3

In the first course of action the out-of-specification material is reprocessed through the dryer. In the altemative case the Effluent Monitor Tank is put on recirculation through a pair of deminerlizers. The demineralizers are connected in series. A third demineralizer is provided as a spare. The Effluent Monitor tank is left on recirculation through the deminerlizers for a designated time and then resampled to determine if the required uranium concentration for discharge has been achieved. The arrangement of Effluent Monitor tanks and demineralizers can be seen in SAR Figure 6.4-35. 4.11.11.2 Criticality Snfety Discussion There are three 1500 gallon Effluent Monitor tanks. The normal throughput of the LWD dryer is 1100 gallons of feed solution per day as described in Section 4.11.10. The normal operating state of the Effluent Monitor tanks is for them to move sequentially through fill, sample, obtain sample results, discharge. Approximately one tank per day will be filled at the normal flowrates. The Effluent Monitor tanks receive dryer distillate which is designed to be uranium free down to the levels allowed in relevant discharge permits. The safety case for the effluent monitor tanks is that no tank must hold more than a safe mass of uranium. The Effluent Monitor tanks are uniquely connected to the dryer distillate line. Correct pipework connected will be verified prior to plant startup. The LWD dyer is uniquely connected to the Dryer Feed tank as described in Sections 4.11.9 and 4.11.10. In terms of the worst case uranium loading of an Effluent Monitor tank, it is assumed that there can be a dryer fault whereby feed solution runs straight through the LWD dryer into the Effluent Monitor tank. It has been established in Section 4.11.9 that there is a double contingency criterion argument to prevent 17 kgU being held in the Dryer Feed tank. As an Effluent Monitor tank can only receive uraniura from the Dryer Feed tank, the double contingency criterion for the Dryer Feed tank must hold equally for an Effluent Monitor tank. The demineralizers are used to extract uranium from out-of-specification distillate held in an Effluent Monitor tank. This is done by " polishing" the Effluent Monitor tank inventory as it l is recirculated through the demineralizers. O Louisiana Energy Senices 4-50 April 1993 Criticality Safety Engineering Report Revision 3 l

As discussed in preceding sections the uranium feed into the LWD dryer is estimated to be of the order of 1 kgU per annum. This uranium will be extracted into the LWD dryer powder stream leaving the dryer distillate free of uranium down to regulatory limits for discharge. If the distillate in an Effluent Monitor tank is marginally above discharge permit limits it will be polished by recirculation through the demineralizers. If the distillate in an Effluent Monitor tank is significantly out-of-specification, e.g., by a factor greater than two or three times discharge permit limits, then the tank inventory will be pumped back to the Dryer Feed tank for reprocessing through the LWD dryer. The uranium loading of the demineralizers over the lifetime of the resin is not expected to exceed a few grams. The uranium loading of the demineralizers will be determined from the Effluent Monitor tank i sample results before and after each polishing operation. A mnning total will be maintained. The TSA operating procedures will require the demineralizers to be disconnected and removed to the Radioactive Waste Storage Area should any significant uranium buildup (see below) occur. The safety case for the demineralizers is based on the assumption that they will subsequently be stored in an array in the Radioactive Waste Storage Area. The criticality safety case for arrays of solid waste containers is discussed in Section 4.10.2, and is based on limiting the mass of uranium to less than 4.6 kgU per container and limiting the areal density of uranium to less than 4.6 kgU per sq. ft. To accumulate 4.6 kgU in demineralizers: A. Some huge excursion from plant normal operating parameters must have occurred to get anything approaching 4.6 kgU in the demineralizers. B. The procedures to maintain a running total of uranium in the demineralizers must fail. The double contingency criterion is satisfied in respect of the demineralizers. 4.11.12 30 Gallon Dry Powder Drums 4.11.12.1 Normal and Abnormal Operation l 1 The 30-gallon drums are connected directly to the dryer powder outlet and collect the solids arising from evaporation of the feed solution in the LWD dryer. The drums are provided with a balance in order that the state of fill of the drum can be determined and the drums changed as required. Once full the drums are taken to the ) Louisiana Energy Services 4-51 April 1993 l Criticality Safety Engineering Report Revision 3 l

Radioactive Waste Storage Area characterized in tenns of uranic activity, and stored pending shipment to a licensed Low Level Disposal Facility. 4.11.12.2 Criticality Safety Discussion The expected annual solids arising from the operation of the LWD is 1800 kg of non-uranic compounds and I kgU. The expected holdup of a 30-gallon dmm will be approximately 90 kg of this dry powder. Consequently, we expect 20 drums per year to be filled, each containing on average 50 gU. The safety case for a drum is based on the assumption that they will subsequently be stored in an array in the Radioactive Waste Storage Area. The criticality safety case for arrays of solid waste containers is discussed in Section 4.10.2, and is based on limiting the mass of uranium to less than 4.6 kgU per container and limiting the areal density of uranium to less than 4.6 kgU per sq. ft. A drum will typically be connected to the LWD dryer for 2% weeks. The uranium held up in a drum can be estimated from the volume of feed solution processed in the LWD dryer while that drum was connected. and from the average uranium concentration in the Dryer Feed tank O over that period. TSA procedures will call for the disconnection of the drum if 4.6 kgU is approached. To accumulate 4.6 kgU in a 30 gallon dry powder drum: A. Some huge excursion from plant normal operating parameters must have occurred to get anything approaching 4.6 kgU in the drums. B. Uranium balance procedures over the connection time of a drum must fail. Therefore, the double contingency principle is satisfied in respect of the dry powder drums. 1 4.11.13 Dryer Feed Tank Filter 4.11.13.1 Normal and Abnormal Operations The filter is installed in the input line to the Dryer Feed tank and its function is to filter out suspended particulate matter. l Louisiana Energy Services 4-52 April 1993 Criticality Safety Engsneenng Report Revision 3 l i i

4.11.13.2 Criticality Safety Discussion t The filter will be safe by geometry. In any event, the uranium is processed in hexavalent form and practically all the uranium will enter the liquid waste streams as the water soluble oxyfluoride. Consequently, there is no reason to suppose that other than trace amounts of uranic particulate matter would accumulate in the filter. 4.12 ACTIVE VENTILATION SYSTEMS 4.12.1 Normal and Abnormal Operations There are two ventilation systems in the CEC which are designed to control airborne radioactivity, the GEVS and a portion of the TSA HVAC System which controls the environment of potentially contaminated areas. See SAR Sections 6.3.7 and 6.4.1, respectively, for funher details. The GEVS is designed to remove trace quantities of uranium and HF from potentially contaminated process gas streams prior to their release. The system is connected by piping to: A. The feed and tails, product and product blending UF cylinder vent systems. B. All autoclaves. C. All discharge lines from mobile vacuum pump sets. The system is connected via elephant trunks to places where piping is normally disconnected or equipment is opened such as: A. The vacuum pumps in the product and tails UF systems. B. The autoclaves. ] C. The product and tails take-off stations. D. The feed purification station. 1 The system is connected via the hoods in: A. The TSA. l B. The Chemistry Laboratory. I C. The Contaminated Equipment Workshop. Louisiana Energy Services 4-53 April 1993 Criticality Safety Engineering Report Revision 3 i l i

On entering the GEVS gases from the process gas streams pass through a pre-filter which removes dust to protect a following HEPA filter, then through the HEPA filter which removes acrosols (mainly UO,F particles), then through an activated charcoal filter which captures any 2 HF. The remaining cleaned gases are exhaused through an air fan to the vent stack. The system consists of five parallel filter stations. Each is capable of handling 25% of the effluent. Four are online; the fifth is on standby for use during filter changeout, maintenance, etc. The filter stations vent through one or two air fans, each capable of handling 100% of the flow. One is online; the other is on standby. For malfunctions or maintenance the standby systems are activated. The system presents no general safety or criticality concem under abnormal operation in view of the minute amounts l of uranium which can enter it and because it contains no connections to water systems. The TSA HVAC System serves potentially contaminated areas other than those served by the GEVS and consists of three HEI A filter plenums which are exhausted by three exhaust fans to the vent stack. There are r.a abnormal operations which could present a criticality safety concern. 4.12.2 Criticality Safety Discussion A study has been performed (Reference 11) to assess the total amount of all uranium which could accumulate in the HEPA filters ia the GEVS and the hoods in the TSA. This study, performed under very pessimistic assumptions, predicts a maximum annual uranium accumulation rate of 1.4 kgU per annum in the GEVS and 1.6 kgU per annum in the TSA j hoods. Even if all this were product material at 5% enrichment it would take over 13 years to accumulate a safe mass in the GEVS filters, and 10 years to accumulate a safe mass in the TSA hoods. These small uptake rates have been confirmed via the regular samples taken on the equivalent systems at URENCO's Almelo and Gronau plants. The low uranium accumulation rate in the Process Ventilation System coupled with the regular sampling of the filters thus renders a criticality incident there highly unlikely. No equivalent study has been performed on the standby filters of the TSA room Ventilation System as these filters only see uranium in the event of a release into the air in cenain TSA rooms. The potential amounts of uranium which could be released in such events are gram b Louisiana Energy Services 4-54 April 1993 Criticality Safety Engineering Report Revision 3

4 I l quantities and thus these filters represent a criticality hazard much lower than those of the Process Ventilatica System. 4.13 DEGREASER BATIIS 4.13.1 Normal and Abnormal Operations The decontamination area will be provided with a degreaser bath equipped with a vapor recovery unit and a distillation still. The degreaser will receive oily pump components from the ventilated pump stripping booth. As pan of the pump stripping procedure, any gross oil contamination and/or blockages will have been removed with hand tools within the ventilated booth. Oa arrival at the decontamination area, components will be re-inspected to confirm the absence of gross contamination or blockages. The pump components will be degreased by time immersion into the degmaser via a suspended basket. Solvent vapor will act to degrease the components. s The basket will be equipped with a scale which will be used to determine the weight difference occurring before and after immersion. The weight difference will conservatively be assumed to be all uranium although in reality a large pan ofit will be oil. Using these l weight readings a mass balance will be kept of the uranium loading of the degreaser solvent. The mass balance will take into consideration the accuracy of the scale. When the degreaser uranium loading approaches the safe mass the degreaser solvent charge is changed. During the solvent changing operation, care is taken to clear out all the sludge residing in the degreaser bath. The spent batch of solvent, containing no more than one safe mass of uranium, will be processed through the recovery still. The recovered solvent will be available for re-use and after every recovery run the still will be cleaned-out of sludge. 'lle sludge will join the contaminated J wastes stream. I l l O Louisiana Energy Services 4-55 April 1993 Criticality Safety Engineering Report Revision 3 l

C l l 4.13.2 Criticality Safety Discussion i The fundamental criticality safety principle to be observed when dealing with these uranium loading organic solvents is that they will always be handled in batches and each batch will i contain no more than a safe mass of uranium. The safe mass incorporates a safety factor of 2.3. In reality, this will be further mitigated by the fact that feed and tails contaminated components will be degreased along with product contaminated components, thus lowering the average enrichment of uranium mmoved. Following each batch processing equipment will be cleaned-out. l l i i i I i l Louisiana Energy Services 4-56 April 1993 Criticality Safety Engineering Repon Revision 3 4 --.o- ,.-.m .r ...---m---

l i s i I l l l 4 r i i I 1 ( l I APPENDIX J i i l 1 i Louisiana Energy Services March 1992 Criticality Safety Engineering Report Revisior. 2

l REF: CRIT /TR/3 BRITISH NUCLEAR FUELS plc ENRICKEENT DIVISICN l CAPENHURST WORKS HEALTH PHYSICS AND SAFETY DEPARTMENT TECHNICAL REPORT CRITICALITY 5AFETY OF THE UCN MOBILE VACUUM PUMPSET USED IN CENTRITUGE URANIUM ENRICHMENT PLANT 1. INTRODUCTION This criticality safety assessment has been prepared in support of the licence application for the USJV centrifuge uranium enrichment plant. The mobile vacuum pumpsets (MVPs) to be used at the Claiborne Enrichment Centre will be the same, or similar, to the UCN designed MVP as detailed in references 1 and 2. They are multi purpose and consist of a number of traps which remove HF, small quantities of UT6 and pump oil from evacuated gases during pumping operations on piping and equipment that has been exposed to UT - 6 This report presents a criticality safety assessment of the UCN MVP for enrichments up to 5% W/w U235 It considers both the intrinsic safety of l an isolated MVP and its safety under interaction with other plant equipment. j The assessment was carried out against the BNTL criticality standard and s _s l code of practice. 2. CRITICALITY SAFETY ASSESSMENT t The Monte Carlo computer code MONM kB was used to calculate the maximum reactivity of an infinite array of identical MVPs under the worst credible conditions of moderation and reflection. In modelling the MVP, several simplifications were made which are described below. A more detailed assessment is reserved for a later stage in the plant design,

i. The following pumpset components were modelled:

Koelval K300 cold trap surrounded by a dewar vessel. The top and side pipes of the trap were omitted from the model and the dewar vessel was taken to be a cylinder of water surrounding the trap of radius 18.5 cm. Leybold WSU-251 pump. This was assumed to have the same maximum internal free volume, (including gear box and casing), as a Leybold WS-250 pump. This volume has been determined by a j practical test to be 4.6 litres. The pump was modelled as a single 4.6 litre sphere of uranic material. Activated carbon sorption trap. The lower dust filter and connecting pipe were omitted from the model. The upper dust -s, [ filter was modelled by pessimistically extending the height of the trap above that of the dust filter. Page 1 of 4 DAK(93)B3

Aluminium oxide trap. The two Klein flanges were omitted from the model and the domed top of the trap was modelled by pessimistically assigning a constant height to the trap equivalent to the maximum height of the dome. Leybold D40B pump. The maximum internal free volume of this pump, (including gear box and casing), has been determined by a practical test to be 7 litres. The pump was modelled as a single 7 litre sphere of uranic material. oil trap. This was modelled as a cuboid of dimensions no smaller than the actual external dimensions of the trap. The casing of the trap was omitted from the model. i Vertical inter-connecting pipe adjacent to activated carbon sorption trap.

11. All of the above components,(excepting the dewar vessel), were surrounded by a 1" water reflector. Full water reflection of the system was not considered since it is expected that the MVPs will be used in open areas for which complete or even partial flooding is deemed to be incredible. The 1" of water reflection used adequately accounts for the possibility of spurious reflection from personnel,

, walls, etc. iii. Aluminium walls were modelled for the cold trap, activated carbon trap and the aluminium oxide trap. (Wall thicknesses were obtained from references 3, 4 and 5.) No walls or casings were modelled for the WSU-251 or D408 pumps, nor the oil trap and the vertical [-s inter-connecting pipe. \\s. iv. The positions of each of the components on the MVP model were obtained from reference 1. The two pumps were a special case since they were modelled as spheres. These spheres were placed as near to the activated carbon trap as possible, taking account of the actual positions of the two pumps, to optimise interaction with what was expected to be the most reactive trap of the system. The internal free volume of each MVP component,(except the dewar v. i vessel), was modelled to be completely full with the uranic material / water mix, (ies no chemical adsorber materials were modelled). vi. The steel frame of the MVP was omitted from the model, as were all inter-connecting pipework and valves other than those mentioned above, only one MVP was modelled explicitly, resting 29 cm above a 60 cm thick concrete floor. The use of the ALBEDO MONK option allowed an infinite horizontal co-planar array of MVPs to be modelled. The reactivity of this system was then determined at various H/U ratios of the uranic / water mix filling the MVP components. 1 O ( (. Page 2 of 4 D AK( 93 ) B 3

3. RESULTS 235 Enrichment (W/w) Infinite Array of Unspaced MVPs 5% U f H/U Atomic Kaff a-ksff + 30-Run No Ratio 6 0.8577 0.0040 0.8697 usjv m12 8 0.8835 0.0041 0.13 5 % usjv m11 11 0.9130 0.0039 0.9247 usjv m10 14 0.9131 0.0041 0.9254 usjv m06 16 0.9136 0.0039 0.9253 usjv.-93-m09 18 0.9093 0.0039 0.9210 usjv.93-m08 24 0.8798 0.0035 0.8903 u s j v m07 The above results are shown in graphical form in figure 1. It can be observed that the maximum value of keff + 30 is below 0.93, within the Company standard of 0.98 for this type of uranium system (ref 6). +w 4. CONCLUSIONS 1. The criticality safety of an infinite horizontal co-planar array of ss l identical UCN mobile vacuum pumpsets has been demonstrated under conditions of optimum moderation of uranyl fluoride (5% U235 w/w) and credible reflection. ii. Since an unspaced infinite array of KVPs is safe, it follows that: an isolated MVP is safe; interaction of an MVP with any other KVP or plant component of l lower reactivity will be safe. Other plant components of reactivity greater than an MVP will be covered by criticality safety assessment with appropriate spacing restrictions. h A vQ ms n., ML G b,k 7 D Y& %A %% [O Page 3 of 4 i l DAK(93)83 l

REFERENCES 1. UCN Drawing No 062-U-491-260.100-04-07

• Ceneral Arrangement Mebile

^ s Cascade Evacuation Unit

• 2.

Facsimile from Mr Charles Mol, UCN, Almelo, to Mr Chris Andrews, Fluor Daniel Inc, dated 20 September 1990 " Mobile Pumpsets". 3. UCN Drawing No 002-U-493-260001 "Kodial K-300". 4. UCN Drawing No 035 -17423-260013-04-56 "Scrption Trapa 5. URENCO Drawing No JV 022 " Aluminium oxide Trap Type A". 6. Company Health and Safety Manual Volume 3 - Codes of Practice.

• Criticality Control" - Code of Practice No ll.

i i \\ Page 4 og 4 DAK(93)s3 .. ~ -. . -. ~.....

1 i S 8 h ek k V l ._p j

1. N f7E _,

{ +---,m- 'f y =s - k _?' s = w x M s 7' X =d:U 35F).EM' *:

~./

rJ + i / 4 [ WN---'-i, O _==-{---z+ j' N . s---d -h.--.m? QH -,' L ~ i C* r-' e r h.._f __f._. n =-. A y --41 : I .-3 T -- -#b-- yy 2 -r -... -. 3 J[_l lp=~ ~ ~ J~-- M '._.m: + ) O k' g[j-. 1 ,--'--T*- 'E: M.. ^ -_g Eg;H +- -o

J-y he,m.a.

M

x. m

-. - + C ,,t_ ~ + A-- "E h* g-r '.Q. ed % x + i d e i l [~~N s [ L-- 1 O I .o T o y i .o .7 ( ? ? Y Y Y ? o o o a o o a o h*'D N k ,}}