- TSP 3Eette I I I THIS DOCUMENT MAY NOT BE REPRODUCED WITHOUT THE CONSENT OF THE ORIG I NATOR HIS SUCCESSOR OR HIGHER AUTHORITY CLA551Fl D BY EXEMPT FRoMGW CATf GORY 5 BJ 2 - 'L' ' E S2 E EMPTION AUTOMATIC ALL y DECLASSIFIED ON Date mpoSBiole to Determine WARNING NOTICE SENSITIVE INTELLIGENCE SOURCES AND METHODS INVOLVED THIS REPORT HAS BEEN PREPARED BY THE SPECIAL PROJECTS DIVISION LAWRENCE LIVERMORE LABORATORY LIVERMORE FOR THE DEPARTMENT OF ENERGY THE VIEWS CALIFORNIA EXPRESSED HEREIN ARE THOSE OF THE PROJECT PERSONNEL AND NOT NECESSARILY THOSE OF DOE TOP SECRET '1 c i TOP SECRET I I • Iii LAWRENCE LIVERMORE LABORATORY -•· 'bSla University of California Livermore Californi¥94550 SPECIAL PROJECTS I - OCi 1 2 '97 TCS-326 019ns ---- - -- concp yIS- A PROLIFERATION GROUP QUARTERLY REPORT JANUARY -MARCH 1978 UNE 1978 1 ·-- ATOMIC ENERGY ACT 1954 a DG Bte 007792 0L T-OP SECRET oGA 1 SPECIAL TCS-326 019 78 PROJECTS Page i-ii This document consists of 80 pages __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _N_o__ o f copies Series _A __• PROLIFERATION GROUP QUARTERLY REPORT January-March 1978 I _I 1 • I June 1978 'E bu TCS-326 019 78 Page iii-iv PREFACE This is one of a series of regular quarterly reports that r provides substantive discussion of activities of the Proliferation Group LLL Special Projects Division This work uses the capabilities of the Special Projects Division and of LLL generally in a broad range of country studies and technical projects relevant to the problem of nuclear proliferation This program whlch supports DOE and the intelligence community has been in existence since late 1974 i buJ oGA I br I lOP SECRET - OG Ai - -- - ' l Cfl T C I U T l I b t - TCS-326 019 78 Page v-vi CONTENTS Page 1 Do£ TOP ------ SECRET b t TCS-326 019 78 Page 1 PROLIFERATION GROUP QUARTERLY REPORT January-March 1978 INTRODUCTION AND SUMMARY • i l This report describes work done by the Proliferation Group LLL Special Projects Division for the period 1 January to 31 March 1978 We present an extract from a Special study done on the nuclear program of Pakistan an article on the geology and history of the Kalahari drill site in South Africa and two articles from our ongoing high explosives program Most of the work was supported by the Office of International Security Affairs of the Department of Energy but support for the high explosives work was also received from the Office of Research and Development of the Central Intelligence Agency The material we present on the Pakistani nuclear program was taken from our report An Evaluation of Pakistan's Capability to Acquire Fissile Material for a Nuclear Explosive 11 which was published separately during February That work was in response to a request from the Department of State for an analysis of possible sources of fissile material for a Pakistani weapons program The use of spent fuel from the I ANUPP power reactor and the possible construction of a production reactor were considered as primary plutonium sources and the question of fuel reprocessing was addressed from two points of view the indigenous construction of a scaled-up version of the PINSTECH hot laboratory and the completion by Pakistan of the French reprocessing plant in the event that the French government decides to cancel the agreement to supply that plant The study showed that the data available at that time did not permit a very detailed analysis The conclusion was reached however that the Pakistanis could succeed at some of these ventures given appropriate national prior'Ix bc1J GAb l D 6 b _J ' ' TOP SECRET - --- -- I • I I TCS-326 019 78 Page 2 ities Some suggestions we e given for intelligence collection relevant to these questions The study of the geology and drilling history of the Kalahari drill site attempts to bring out finer detail than has previously been made available on this subject Open and classified literature and overhead photography were examined and the results provide some additional insights into possible intended uses of the site Some comments are provided on hole depths and corresponding capabilities with respect to nuclear test yields were calculated Geologic data based on nearby boreholes have allowed us to draw stratigraphic cross sections that indicate a 60- to 85-m-thick sand layer overlying a granitelik e rock layer These data are consistent with the 75-m thickness estimated from hole casing and from analysis of other drilling apparatus After the casing of the hole to 75 m there was further drilling into the hard rock and estimates of the extent of this additional drilling are made based on the usage of bit cutters known from collateral data and on estimates of the number of days spent in drilling The result based on number of bit cutters used 94 to 152 m is considered most reliable This gives a total hole depth between 169 and 227 m The high explosives work that is described here is a part of an ongoing program directed toward identifying signatures of ooe nuclear-explosives-related HE testing 6 1 previous quarterly reprogram ports The article on energy partitioning in HE detonations deals with a question that is fundamental to an understanding of observables such as total light outpl t J ireball growth streamer Ot kl A oG-A 6 I Ot U J - TOP SECR T - - - TCS-326 019 78 Page 3 phenome nology and so on The reported results are based on theoretical calculations using LLL weapon-design codes At late times it is clear that most of the available energy from an event has been transferred to heating and to movement of surrounding air Roughly 10% of the energy has gone into dispersal of pit material and about 10% remains as internal energy of the HE products ' The final article describes and extensive series of tests fired to determine the utility of seismology for measuring HE yields We have argued that whereas seismic signals will not identify an implosion test per se the ability to measure HE yield even fairly crudely could be very important in foreign test-site monitoring An indication that large charges SO kg or greater are being fired together with other indicators site configuration etc might identify nuclear-related testing we fired 46 special shots over a several-week period The explosives were for the most part simple configurations of bulk HE and seismic measurements were made using a net of broadband velocity meters We examined the following questions • Does the seismic-signal amplitude depend on height of burst Series of 23-kg and 45-kg charges were fired at heights from zero to 2 1 m above the ground no burstheight dependence was noted • Would seismic coupling depend on device configuration i e hemisphere pointing downward vs hemisphere pointing sideways Tests simulating extreme cases showed at most a small perturbation on signal amplitudes • What is the yield scaling relationship for near-surface bursts of HE in the yield range relevant to nuclear HE work A series of explosive masses ranging from 11 to 234 kg was fired with fixed firing geometry and a simple scaling relationship was obtained t b 1 TOP SECRET OG A b ct TCS-326 019 78 Page 4 • How important is loc l geology Shots were fired at different locations and several seismic stations were used It is clear that travel-path variations will frustrate attempts at absolute yield measurements by seismic means A strong acoustically coupled signal that arrives after the direct seismic earth-coupled signal may hold more promise for absolute calibration OGA _ ti MiAb • TCS-326 019 78 Page 5 PAKISTAN COUNTRY STUDY In response to a request by the U S Department of State we undertook an analysis of Pakistan's capability to acquire fissile material for a nuclear explosive Our report on the study titled An Evaluation of Pakistan's Ca pability to Acquire Fissile ' I • i Material for a Nuclear Explosive was published as TCS326 009 78 and was distributed by DOE OISA We outline the report here including its executive summary and a list of questions pertinent to future collecting on the subject As Pakistan's possible sources for fissile material three scenarios were considered the use of a production reactor stealing fuel rods already in the cooling ponds of the Karachi Nuclear Power Plant KANUPP power reactor and running unsafeguarded fuel elements through KANUPP The success of any of these scenarios for the acquisition of weapons-usable material requires that Pakistan have facilities for reprocessing For this two alternatives were considered First we assessed Pakistan's ability to construct a scaled-up version of the Pakistan Institute of Science and Technology PINSTECH hot laboratory A second evaluation centered on the possibility that Pakistan may try to complete the French-designed Kundian Nuclear Center KNC reprocessing plant in the event that the French Government decides to cancel or alter its agreement to supply that plant To limit the scope of this study the Department of State suggested that the following general assumptions be made · • Pakistan is attempting to acquire fissile material sufficient for one to three devices from 10-35 kg of plutonium • The effort will be given priority and financial resources will be redirected from other projects if necessary of C 1 t oor uU TOP SECREl oG- A bc1 TCS-326 019 78 Page 6 • i I I The constraints of health and safety standards will not apply It should be noted that this study was meant to address the issue of Pakistan's technical capabilities no effort was made to analyze or evaluate the political and economic ramifications that might affect Pakistan's decision to follow one or another of the scenarios It was our opinion when this study was undertaken that no very definite conclusions were likely As we implied in one of our earlier reports on Pakistan's nuclear program the Pakistanis have only a marginal capability to succeed at any complex large-scale technical enterprise They clearly have many of the required skills but success in acquiring fissile material for a nuclear explosive device depends on factors other than pure scientific expertise Pakistan must also possess management skills ample capital industrial capacity and a host of lesser supporting capabilities This study shows that the data available at this time do not permit a very detailed analysis The conclusion is reached however that the Pakistanis could succeed at some of the ventures outlined above given appropriate national priorities Whereas this study does not present definitive conclusions hopefully it will serve at least to clarify some of the relevant questions The more important questions raised by this study are listed below Executive Swnmary This study addressed the questions of whether Pakistan can acquire and chemically separate fissile material for one to three nuclear explosives Unfortunately lack of information has precluded our giving definitive answers Rather we have culled as much relevant data as possible and have drawn very general conclusions OGA h 1 J J LJ C 1 J -- TOP SECRET oG A 1 - ______ 1 1 TCS-326 019 78 Page 7 • • • I • • l I t L Pakistan would be unlikely to build a pr duction reactor - due to the cost time and requirements of pursuing what would be for Pakistan a new line of technology Stolen KANUPP fuel elements could be reprocessed at the cost of breaking safeguards and endangering future technology transfers Repercussions of reprocessing KANUPP fuel could be avoided by running unsafeguarded fuel through the reactor and later extracting the plutonium under the guise of peaceful nuclear purposes Using KANUPP as a production reactor would necessitate having both unsafeguarded uranium and access to fuel fab- TOP SECRET TCS-326 019 78 Page 8 - ---- - - - - T 11 1 To give more definitive answers to the questions posed about Pakistan's capabilities more detailed information must be provided Below are some questions that arose in the course of this study they may be helpful in formulating priorities for future collection Suggestions for Future Reporting A number of questions need to be answered before a more definitive evaluation of Pakistan's capabilities can be made We give here a set of collection suggestions some of which we consider to have very high priority in the context of the present study Much of this material could be collected by us Government personnel from open sources within Pakistan TOP SECREl - I - -- 0 • - f - --SECRET 6 TCS-326 019 78 Page 9 l I t• - I I i _ '' oG A b' t TCS-326 019 78 Page 10 J oF • t U iOP SECRET i ' TCS-326 019 78 Page 11 ' GEOLOGY AND HISTORY OF THE KALAHARI DRILL SITE Introduction This article attempts to bring out finer analytic detail from open and classified literature and imagery of the arill-site portion of the Kalahari Probable Nuclear Test Site than has been previously available Detailed knowledge of the subsurface stratigraphy is vital in any functional scenario that one may be considering for the Kalahari site Likewise a detailed day-by-day observation of changes and procedures at the drill site the one location we have the best history on might give a few extra clues as to the purpose of the Kalahari site Applying the geology of the site to what we see and know of the drill site a minimwn-maximum depth for the drilled hole has been estimated and corresponding capabilities in nuclear test yields calculated Geology of the Kalahari Site - I Stratigraphy 112 f h 1 h ' d escriptions ' ' Previous geo 1 ogic o t e a a ari site were generalized ones coming from available regional open-literature geologic descriptions P J Srnit's description of the Karoo system 3 gives detailed stratigraphic information of nearby wells Niete Min 7 3 km Nl0°W from the drill site--Figs 1 and 2--and Hop Hop 13 3 km N25°W from the drill site as well as the stratigraphy of more distant boreholes this information allows us to determine regional trends in the di ection of the drill site From the extrapolated cross sections Fig 3 we can determine that the Karoo system the primary sedimentary bedrock underlying much of the Kalahari region has cv ly pinched out i e is ' - BOTSWANA 27• ' fn w b j S -'-'U U 'G ·-' - TOP SECRE o' TCS-326 019 78 Page 13 L£GENDE Topogrophicol hl' _ m ret111iva lo sea levtl Topogrotie e hao m relol i• f tot nevlall • • 10 Ho ghr m of base ct Koroo System Hoogre m van t c si• on Sislaem k0r011 · ffil C ' 80re -cle of which profile is given Boorgol waorvon profiel ge tt woril I Direction of movement of ict Rioting van beweging van ya L Strike and direction of dip StreHing en rioting van helling _f_ Strike and dip infuredl Slrekking en helling lotg• l1i • ' I Oolerile • sheet outcrop Dole rietplaat Idogsoom Dolerite aheetlinr • rredl Oolerietplaotlofgel• il l Dolerite dykt outcrop Doler i• tgono dogsooml I Dolerile dyke infetred l 'olerietgong ofgelei l 7 7 Sandstone ord hole oulcropl LLl Sondsleen en 11kolie Idogooom Sandstone r I $gndsteen D Tillite and mudston• conglomorote outcrop T lliet • n modderstee konglomerool dogsoom t oroo Syetam Siateem Koroo r o• o• l owat Owyko Siege £toge Onder· Owyko Owyko Serlea Serie Dwyko Pr• ·Koroo outcrop Voor - l aroo dggaoom L5 LJ Pre- Koroo inferrt d - Vo0r-1 0ro0fofgel11i •R Drill she •T Tower sl1a A---A B- ----8' C-··-C' l Profil• la ations sea F ig 3 0-0' S Figure 1 Distribution of the Karoo system in the Kalahari Cape Province k6 ' '- I J X 't b l j TO·P SECRET TCS-326 019 78 Page 14 ' I 1 I l TSR Figure 2 The Kalahari Site showing nearby wells QOc D fOP SEC TCS-326 019 78 Page 15 nonexistent at the tower and drill site s The rock type 1mmeal- -- ately underlying the Kalahari sand in Smit's report is designated by a symbol normally denoting granite on South African geologic maps We therefore assume it is granite or a similar hard gran- itelike rock below the thick surface sands This information substantiates the need for hardrock cutters for drill bits as indicated in collateral reports Figure 3 four stratigraphic cross sections produced from d ata from boreholes shown on Fig 1 s how regional trends as follows 1 The portion of the Kalahari we are interested in lies on the eastern slope of a troughlike de pre ssion of sedimentary facies which deepe ns to the northwest from the Kalahari site and thins to the west and south of the site The Karoo sediments thin drastically to the east apparently pinching out just a few kilometres north and west of the drill site Fig 3 The Karoo beds the aquifer system of the r e gion pinch out between the Niete Min well which is located within the security perimeter of the Kalahari site and the drill site This is corroborated on KH imagery by the fact that the nearest wa ter well to the tower and I i t drill sites is 4 5 km north-northwest of the drill site probably at a point where the Karoo system beds are still present Water was piped from this well to the artifact water reservoir near the tower site and to the reserve water reservoir at the drill site Fig 2 2 The Kalahari beds comprising a e olian windblown sands thicken from west to east Fig 3 cross sections A-A' and D-D' J but appear rather uniform in thickness from north to south cross I I I section B-B' through the drill site At the drill site the sand thickness appears to be from 60 m Fig 3 cross section CC' to 85 m cross section B-B' From the casing seen at the drill site and subsequently installed in the hole and from analysis of the observed drill and drilling-assembly string seen l t - I ' 1 A A' West East Niete Ver Potsc Pan Orieling Kruispan 0' e· B North Niete Ver Surface 'd Pl South Niete Min '- Drill site Surface Q D I-' O'I - 100 m - 100m -200 m -200 m -JOO m -300 m i ·• · · · ••11 ·r• · r-·••• · - · - • -· l # 8 ' Ul I w N O'I 0 I-' Pinchout _ - • • • · - • ' - · ' · · · · · · · - ·t' ' • · • - · · · • · _ - • · · • - - 00 -400 m ti 1 J c· C D South Springputs Hop Hop Drill Site X Surface ----- -- ------ e- --- --- -- --t-- -- -- ---r- -r---- - ' ' ' T' North D' West Hop Hop _ East Potse Pan Drieling '- N1ete Min Surface--1-- -- - -- ------- --1-- -- I• -- ' -100 m -IOOm -2DOm -200m -300 ' -JOOm 100 1 Vertical exaggeration --400 ' 0 10 20km Horizontal scale -500 m --- _ I· J Kalahari Sand e Undifferentiated Kalahari sand Karoo W Karoo System Undifferentiated Karoo Pre•Karoo I Figure 3 Generalized stratigraphic cross sections A-A'--D-D' Pre-Karoo i - D -- In S r' fOP SECRET I ' TCS-326 019 78 Page 17 by the rig the depth to the base of the sand at tfie - drills 1te is 75 m well within he range indicated by the cross sections Variations in sand thickness can be accounted for by the uneven uncomformable contact of the granite like rock and the sands and by elevational variations in the surface sand dunes 3 The hard gra nitelike rock approaches the surface to the east and south of the drill site apparently lying directly below the Kalahari sand at the drill site No sediments other than the Kalahari sand and the Karoo are present in this portion of the desert It may be noted that according to our profile s the dashed line approximating the easternmost limit of the lower Dwyka stage of the Karoo system on Fig 1 falls more than 10 km • l I east of where our analysis indicates it should be The Fig 3 pro files a re generalized and do not show the surface topography From profiles we have constructed but do not show here we can outline the regional trends of the surface as follows e ' I I t From east to west the regional surface elevation rises about 8 5 m km rising from 865 mat Potsepan to 995 mat Niete Ver • From north to south the regional surface elevation re- mains fairly uniform with only local variations Figure 4 shows what we believe is the most likely geology at the drill site Seismicity of the Kalahari Site iI From seismicity maps of South Africa 4 Fig 5 the Kalahari site can be seen to be in an aseisrnic area where the nearest earthquake between 1950 and 1975 was 90 km to the east The site is several hundred kilometres away from any earthquake epicenters having a body wave magnitude of 4 0 or more for the same time period Therefore any seismi c signal emanating from the -1 oE b cJ TOP SEGRE I l 1 boe b g CG TOP SECRET I TCS-326 019 78 Page 18 ---- - - - - - · -· ------ 7 Apparent water content Lithology ·-·· Kalahari beds aeolian sands Granite or hard granitelike rock Figure 4 Mostly dry 200 200 300 300 Apparently dry enough not to require casing Stratigraphic cross section at the drill site I Dof b OJ TOP SECR El TCS-326 019 78 Page 19 1 'I I - ' I • • J I I t' t' ' I • • • • • • • _ • ' - -i I • • --r • I • • • • 10 • • I -- J •• • • •• • 15 •' '_ • • • - • • - _J · • r 1 I •• • r 0 0 • • • t • • •• • • •• • • •• • • 0 • • • I'- • • • • · · ' ••• ' D ' _ • I SITE _ r- • I 30 '°7 __ • • I I I I - ' r •• •• r r•• • 9 • I - ' I • • • •• • 25 _ • • e I _ - _______ l - - - - -• I • • ' 35 -z 15 • • 20 25 30 35 40 $ Figure 5 -r SECR T TCS-326 019 78 Page 20 general location of the Kalahari site should be looked at as a probable explosion Ground Water From the procedures used in the drilling and casing operations there is no evidence of any saturated rock being encoun- 1· ' ' tered Though we are reasonably sure the surface sands are dry the water content of the hard granitelike rock is not clear 1 Because of the apparent lack of casing used in the hard granitelike rock portion of the drill hole see Drilling Rates and Procedures it is unlikely that any saturated rock was encountered If this rock were truly competent granite with few fractures one normally would not expect to encounter the water table within it • Chronology of Activities at the Drill site July-December 1977 I ' i site Layout The drill site when first imaged on 4 July 1977 comprised the following a A Wirth German-manufacture trailer-mounted A-frame drill rig either L-10 or L-15 5 having a height of 19 7 m and a nearby support shed b A drilling-fluid supply system comprising four pits numbers referenced in Fig 6 No 1 Mud pit into which the shale shaker discharges No 2 Small possible mud mixing pit No 3 Settling pit containing a baffle No connecting line to the rig is visible No 4 Reserve water supply pit connected by line to Pit No 3 j 1 oGA i tJ - Q I A TCS-326 019 78 Page 21 a i l D bLtJ fOP ECftET ' I'' • TCS-326 019 78 Page 22 c A d cuttings Two turbine generators with four fuel storage tanks e f An open dri1ling-equiprnent yard A metal gable-roofed shed possibly for equipment stor- g h probable shale shaker to receive the drilling mud on its return to the surface and to screen out coarser age Six locations of concrete pads mo st of which have small sheds both open and c losed Their exact purpose is unidentified but the y apparently support the drilling operation because they we re removed after drilling s topped An open-canopied structure that a pparently has camouflage-type garnishment on top The garnishment and or the materials underneath frequently change in outline This may be simply a location for open storage of Jerry gas cans or an attempt to camouflage such things as drill pipe that could be counted to determine the hole depth If it is for drill-pipe concealment the pipe would have to come in short sections It seems unlikely that drill pipe would have been stored here due to the difficulty in handling and moving it even in i short sections Miscellaneous support trailers and tents Drilling Method The Kalahari drilling oper ation appears to utilize direct circulation i e downward flow of drilling fluid through the drill pipe and upward in the annulus around the drill pipe Fig 7 The fluid or mud is discharged at the surface over a probable shale shaker into a series of three earthen mud or settling pits The drilling techniq ue observed is less sophisti- b i f TOP SECKEt --- br -J Doe bu TOP SECREl EC E'f TCS-326 019 78 Page 23 • Figure 7 I Circulation method believed to have been used at the Kalahari drill site Arrows show path of drilling fluid from slush pumps A to swivel B down through kelly C through the drill pipe D to bit E Fluid washes cuttings from hole bottom carrying them to surface through annulus F through shale shaker G to mud pit No 1 H on to pit No 3 not shown from which e - _ cle begins - J OE' b6 felt EC Eif ECl't T TCS-326 019 78 Page 24 cated than those currently in use at NTS where concentric dual - -7 drill strings are presently used with fluid circulation down the inner string and air-assisted circulation return up the annulus between the inner and outer drill strings Reconstructed History of the Drill Site Figure 8 is a chronology of activities at the drill site from 4 July when i t was first seen through 26 December when the rig had been removed and other equipment was being removed Our reconstructed history shows the various activities we believe to have taken place at the drill site during that period The gaps between activities notably between completion of drilling a 48- or 52-in hole and the beginning of 1-m-diam casing installation and between the 1-m casing installati m and the resumption of drilling suggest that there was no high priority in comple- tion of this hole The crews could even have been given time off between the casing operation and resumption of drilling Drilling Rates and Procedures An estimate of drilling rates must take into consideration the working depth hole size and lithology Typical large- diameter hole drilling operations are done in stages Normal dimensions such as hole bit sizes that are standard throughout the world are given in inches all other dimensions are given in SI units Hole sizes have been estimated based on what would be required to accommmodate the mensurated 1-m-diam casing · '' ' t I r Activity First phase l I I·· I Oates of clear KH imagery I l Rig removed n I I l I I I I Hole completed 0 ts- __ Drilhng 36-in hole lnst1111ing 1-m-diam casing Drilling 48- or 52-in hole Installing surface casing --------- Rig set up MAY JUNE Second phase ••••••• -- --- r------- - Est depth 9 m JUI V I 1-· • I I I I I I I m 'I ----- I -----1 I ----1 I I I I I I I I I I I I I Imagery gap l I I I I I II I I Third phase Depth 169-227 Depth 75 m AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER l 1977 Tl l -0 Li Figure 8 Reconstructed history of the drill site ttl Ill i Q D 'U1 1-3 en I w N O'I ' ' 0 I-' 0 - J 00 • lOP SEER1Wr C A l r I SCR T TCS-326 019 78 Page 26 1 I The rig spuds in drilling a large-diameter hole to a shallow depth This hole is normally cased with surface casing and the annulus between the hole and casing is usually cemented This procedure minimizes the hazard of having •t he hole wash out or slough out from the weight of the drill rig 2 Cleanout of the hole and continued drilling with a smaller-diameter bit to fit inside the casing in this case to the bottom of the Kalahari sand and installation of 1-m-diam casing 3 Hole cleanout into the hard granitelike rock below the sand to a desired depth Casing of an estimated 26-in diameter could have been installed particularly if the granite were saturated and a dry hole were desired Few nuclear devices are tolerant of submergence under any significant head of fluid whether it be water or mud Because of lack of imagery during this critical third phase of drilling we cannot determine if the hole was pumped down or dewatered We believe that casing is unlikely due to lack of evidence of casing being delivered to the site For the first or second emplacement of a nuclear device the constraints of a hole diameter smaller than 26 in seem excessive The drilling rates and procedures in the Kalahari sand and the hard granitelike rock will be discussed separately Sand Drilling Rate and Procedures - A hole 52-in or more in diameter probably 64 in was started and drilled to a depth of about 9 m sometime between 4 July and 12 August 1977 Surface or conductor casing was probably installed Based on the drilling rate seen later in the site history we estimate this early drilling time was about 2 days using 4-s· m day as an average penetration rate J ltl b Q l TOP SECRE' f TCS-326 019 78 Page 27 The vidence for this first having been cleaned out at l e ast 4 July 1977 the pit was clean in certainly never cleaned out No ·' I ' ' phase consists of mudpit No 1 once by 12 August 1977 On outline apparently unused and surface casing was visible on 4 July or 12 August imagery A 5-m-long cylinder large enough to be surface casing appe ared in the storage yard at the drill site between 18 and 21 September 1977 The second phase of drilling drilling to the bottom of the sand and possibly about a metre into the hard granitelike rock took place from about 22 August through 1 September 1977 The new drilling assembly- consisting of several pieces of 9-5 8- or 13-3 8-in drill pipe drill collars we ights and bit--was present on the ground on 20 August but gone on 22 August The hole was deepened to 75 m during this time Because the hole was smaller in diameter e stimated to be 52 or 48 in a drilling rate of about 5 5 to 6 m day was achie ve d during this 11- to 12day period In sand this rate still is not good and the slow rate suggests crew inexperience The hole was apparently fairly stable because the installation of casing did not begin until about 17 September 1977 This time gap between completing the 48- or 52-in hole and when the installation of casing was started is most unusual Casing installation is normally started I I as soon as possible to minimize the danger of the hole caving in Even holes in more competent rock tend to slough or deteriorate with time Thirty sections of c asing totaling 75 min length were installed between 14-17 September and 1 October Figure 9 shows the apparent rate of installation with annotations speculating on what happened during ca sing installation No evidence of cementing the casing was ever seen although cementing could have been carried out during a one-week gap in imagery 30 Ssptember-7 October 1977 An explanation for this 48 in if surface casing of 52 -t used b J 1 I l oE bu TOP SEGRE TCS-326 019 78 Page 28 1 piece left over I Installation could have been finished 10 October I· i • 20 - C - i i n C · Ill CJ s - 0 II c E z o- Installation started between 14 and 17 September 0- 13 14 15 16 1 17 1B 19 20 21 22 23 24 25 2a l 27 l 2e 29 September 1977 D Figure 9 Uninstalled casing D Installed casing 30 1 2 3 4 5 6 7 October 1977 TSR Seen later in the day History of drill site casing installation Do€ 16 1 I - b tj TOP SECRET TCS-326 019 78 Page 29 might be that the casing was run open-ended- al ffie oottom - a metre or so into the hard granitelike rock and tacked by pouring in cement completely filling the bottom of the hole for a few metres Then the remainder of the hole outside the casing was either filled with sand or gravel or left open The cement rhis method of cementing also tacking cementing technique to be used the load capacity of the rig must be enough to suspend 75 m of 1-m-diam casing weighing about 29 000 kg 32 tons This load is far short of the hook capacity of 90 000 kg 100 tons for the small 6 L-10 rig Further corroborating the open-end tacked-casing premise is the apparent difficulty or delay in installing the 1-m casing to the bottom of the drilled hole If the casing had been floated in the bottom of the casing would have been closed with a hemispherical head allowing additional fluid to have been pumped into the casing to give it added weight and facilitate getting it past any tight spot to the bottom Hard-Rock Drilling Rates and Procedures - The third phase of drilling to a planned depth in the hard granitelike rock began about 25 October 1977 and was completed by 1 December 1977 The Keyhole imagery for this time period is unfortunately scarce there was no imagery between 5 November and l December and therefore we do not know with any accuracy the number of days used for drilling We can calculate the hole depth two ways from the usage of 5 the bit cutters known from collateral data and from the number of days drilling was in progress Do C bO i TOP SECRET I TCS-326 019 78 Page 30 I I It is known that six sets of bit cutters were ordered and 5 delivered One set was returned unused Therefore it is likely that somewhat more than four but no more than five sets of cutters were used to drill in the hard granitelike rock At the rate of 22 9 to 30 S m set minimum and maximum cutting capability per set in hard rock 5 the depth drilled in the hard granitelike rock ranges between 94 and 152 rn Total hole depth would range between 169 and 227 m As stated above the number of days spent drilling in the third phase of drilling is not known We estimate it to range between 20 days to a definite maximum of 35 days Given the daily minimum and maximum capabilities of the cutters in hard rock 3 85 to 7 7 ml 5 if the rig were operational 70% of the time the estimated depth in the granitelike rock ranges from 77 to 270 m Total hole depth would range from 190 to 345 m Additional imagery during the 5 November 1977 through 26 December 1977 period might have allowed us to determine 1 Whether casing or a liner was installed in the hard 2 3 rock The amount of drill pipe and a more accurate calculation of hole depth Whether or not the hole was dewatered If a hole is not to be used shortly it will keep better with fluid in 4 it Whether anything was put in the hole before it was cov- 5 6 ered Whether the hole was stemmed or not The number of days of drilling The amount of actual time the drill bit was on bottom in a drilling mode was probably considerably less At the Nevada Test Site drill bits are working on bottom about 35-45% of the time J TOP SEER iEettET TCS-326 019 78 Page 31 Best Est1mate of Tota Hole Depth We believe the minimum-maximum depth cf hole estimates based on cutters alone is better than the estimates calculated from the number of days drilling due to the large range in the number of days The best overall estimate of the depth in the hard granitelike rock is 94 to 152 m and the best total-hole-depth estimate is 169 to 227 m Based on a drilling rate in granite of between 3 85 and 5 0 m d ay we do not believe a faster drilling rate was achieved since the average daily rate in the s a nd was only 5 5-6 m the time s pent drilling in granite was 19 to 39 days Drilling should have been comple t e d by the end of the first eek in December The cancellation of the order for additional cutters on 1 Dece mber 1977 indicates that drilling was completed on or about that date Nuclear Yields Possible at the Drill Site Given a depth range of 169 to 227 m what nuclear yields might one expect if the drill hole we re to be used for a nuclear DL bv OP SECRE 1C 4 ' dI 6 1 TCS-326 019 78 Page 32 _ I t fl L TOP TCS-326 019 78 Page 33 ENERGY PARTITION IN HE DETONATIONS The energy from an HE detonation divides itself among the following • Internal and kinetic energy of any associated metal parts Internal and kinetic energy of the detonation products Internal and kinetic energy of the surrounding air For a typical 90-kg 200-lb detonation the partitioning process • • i I i s complete after 100 µs By this time the radius of the detonat ion products is only three or four times the original radius of the HE We have done calculations on four spherical implosion test devices that we fired last summer The four devices which were described in detail in our July-September 1977 Quarterly Report were I ' e K2 • K3 o K4 • KS J We have calculated and plotted in Fig 10 the kinetic energy per· unit volume of the pit material for the K2 and K3 tests The time of this calculation is 75 µs after detonation by this time the pit has undergone compression and has started to expand In Fig 10 and Figs 11-13 as well though we give the energy at 75 s afteF detonation for cla rity of presentation we describe the location of the pit material in terms of its position at zero time 11 A similar plot for the K4 a nd KS tests is shown in Fig Note that kinetic energy is proportional to the square of the velocity At the time of these c a lculations 75 s all pit material is moving outward I' OP SEGRE 1 J ' b1 D01 hu laOP SECRET • SliiCPl T Rr IS FT IID tU T O TCS-326 019 78 Page 34 OOEE C3 Fi gure 10 Kinetic energy per unit volume of pit material for devices K2 and K3 calculated for 75 µs after detonation plotted as a function of the zero-time position of the pit material • D 'f b6 1 oG' A I - 'I TCS-326 019 78 Page 35 Figure 11 Kinetic ene rgy per unit volume of pit material for devices K4 and KS calculated for 75 µs after detonation plotted as a function of the zero-time position of the pit material D tu TOP SECA T TCS-326 019 78 Page 36 Plotted in Fig 12 is the internal energy per unit volume of the pit material for the K2 and K3 tests By the time of these calculations the tions Dot rs Plotted in Fig 13 is the internal energy per unit volume of the pit material for the K4 and KS tests ' J l TOP ECKEl TCS-326 019 78 Page 37 -- ---------l' s Doe bC3 Figure 12 Internal energy per unit volume of pit material for devices K2 and K3 calculated for 75 µs after detonation plotted as a function of the zero-time position of the pit material lot btJ TOP ECK l Ji«1J O -A 'Pa£ 1- TOP SEER T oc I TCS-326 019 78 Page 38 Figure 13 Internal energy per unit volume of pit material for devices K4 and KS calculated for 75 s after detonation plotted as a function of the zero-time position of the pit material J oe bl TOP SECR-T TCS-326 019 78 Page 39 -- · - - - - - - - --- -- - - - - 1 I I I 4 iI I I 'I l I II ' Figure 14 ' Fraction of energy in the pit material as a function of time PO b l TOP SECAE J'°1 OGA __ _bc 1 s eREl' TCS-326 019 78 Page 40 j Therefore the calculations of the energy transfer into the air were made for an implosion device with no pit material We think it entirely proper to simply subtract the energy carried away by the pit from the total en ergy of the HE before the transfer to air begins In Fig 15 we show as a function of time the total partition of energy of an HE detonation The calculation was made specifically on the KS test however this will serve as a close approximation for all the events that we have considered At late times it is clear that most of the energy has been transferred to heating and movement of the surrounding air Roughly 10% of the total energy has gone into the pit material and about 10% remains as internal energy of the HE products This latter component is the source of energy output in the optical spectrum Note that most of the energy that is imparted at early times into motion of the HE products has been transferred to the air It is this transfer that defines fireball size the maximum occurs when hot HE products have expanded into pressure equilibrium with the surrounding air It is important to emphasize that in these calculations we have not considered afterburn phenomena in which the turbulent mixing of air with unburned HE products can contribute considerable additional latetime energy This phenomenon is difficult to calculate varies with different HE compositions and presently frustrates attempts at rigorous calculation of fireball-growth dynamics cG br TOP s B4 · TCS-326 019 78 Page 41 Figure 15 Total partition of energy as a function of time for event KS This result will closely approximate the situation on the other events we have examined Q 4J TOP SECR T TCS-326 019 78 Page 42 EVALUATION OF HE CHARGE SIZES FROM SEISMIC AMPLITUDES Introduction We have discussed in an earlier report 8 the possibility that an HE-yield identification could be established based ·1 ·1 ii II follow-on experiments were planned In this article we describe those experiments and present the first portion of our analyses of the experimental results The work we will describe involved 46 special HE detonations directed specifically toward the seismology question They were for the most part firings of simple configurations of bulk HE The seismic measurements were made by seismologists associated with tne LLL Earth Sciences Division The present analysis should be viewed as very preliminary and more detailed analyses will be contained in the final report to be written by our seismology team The present report is based on only the vertical components of the direct P-wave signal A stronger signal arriving at acoustic-signal transit times is not discussed at all here it is being examined by the seismologists and may in fact hold more promise than the other signal components as a means of arriving at an absolute calibration for an amplitude-vs-yield relationship We first present site details covering the locations of the various firings and the positions of the seismometers Following that we discuss the specific types of experiments made and analyses to date considering ---- --TCS-326 019 78 Page 43 • • • • • eight-of-burst dependence How does the seismic amplitude vary with the height of burst We cover the full range of heights expected in nuclear hydrodynamics testing Geometric perturbation effects How do seismic amplitudes vary with different device types and different device orientations We consider some extreme examples Yield dependence How does the seismic amplitude vary with the explosive yield of a shot We cover the full range of explosive yields expected in nuclear hydrodynamics testing In this preliminary article we deal only with the question of relative rather than absolute yields High-explosive-type dependence How does the seismic amplitude vary with different HE compositions we cover more than the full range of types of HE expected in nuclear hydrodynamic tests Shot-site variation effects How do seismic amplitudes vary for different shot sites Site Details All the HE firings were at LLL's Site 300 HE test area Fig 16 Most of the firings were on a level area near Bldg 802 The area was covered with about 150 to 300 mm of pea gravel for the firings and this was smoothed after each firing To evaluate local variations with geology one set of shots was fired in a cross-shaped distribution with about 75-m arms centered on the normal firing point one shot was fired at each of Bldgs 801 804 845 and 851 to evaluate the effects of major changes of location For most of the shots three-component broadband Geo-Tech with useful frequency seismometers that measured ground j c f _ b v TOP SECRET DoE i o · j FOP SEERE TCS-326 019 78 Page 44 Firing site • Seismometer site A ·····•···· 1Disposa jsite j 1 I I I Sc ale km I 1i IU Figure 16 Seismometer e mplace ments at LLL's Site 300 HE test area J £ loCV OP S_ECREl Dof k D TOP SECRET - - · - - - TCS-326 019 78 Page 45 response from esentially 1 6 to 60 Hz were emplaced at Bldg 845 Bldg 858 and a Linac Road site as noted on Fig 17 After the main shot series was completed the seismometer at Bldg 858 was transferred to the position noted on Fig 17 as the disposal site This simulated field emplacement at an uncalibrated site We show in Fig 17 the position of a long-range seismic station fielded on several of the firings The seismometer was a Hall Sears 10 1 Hz that measured vertical velocity only Height of Burst Dependence If the seismic signal is strongly affected by height-ofburst variations typical of those expected in hydrotesting seismic evaluation of Nth country HE testing would have little or no practical value Plausibility arguments as follows suggested that height-of-burst effects would be minimal over a reasonable range of height values from detonation on the pad to detonation 2 m above the pad half the explosive energy will be directed downward with negligible attenuation e xpected The region of seismic coupling will have a characteristic dimension about equal to the wavelength of interest A where A is greater than c v c being the velocity of sound in air and v being a frequency characteristic of the seismometers Details of the coupling mechanism need not be known precisely the real velocity to use in our estimate cannot be less than c so we obtain an estimated lower bound on A For v 10 Hz A 35 m The solid angle subtended at the shot point by the intersection of the ground plane and a sphere of radius r centered at a height H above the ground is 2 1 - H r We haver A 35 m and H 2 m so the bracketed term is no smaller than about 0 95 Thus in the worst-case situation H 2 m we expect that about 94% of the energy that can be coupled to the ground will be coupled to the ground We b u TOP 5ECRi I 1 -'Y P Lr d °' u' l 1l t-3 n I l J ' ' °' ' ' 0 ID 0 l• I ···· ······802 Bldg Seismic stationl-_ _11 5 k m Site 300 UI Figure 17 Position of the long range seismic station The logo shows the location of the Lawrence Livermore Laboratory main facility - --- 0 -- -p I ·'-- --- ---- 5EGREJ TCS-326 019 78 Page 47 thus expected on thi basis that the seismic signaT sn6u1a·- be only weakly dependent on height of burst Several firings were devoted to experimental verification of '· our conclusion A series of 23-kg 50-lb charges of the explo- sive C-4 was fired at heights from Oto 2 m above ground and a similar series was run using 45-kg 100-lb charges of LX-04-1 Table 2 lists the 23-kg series along with relevant firing details Note that the temporal order of the shots was randomized to eliminate possible systematic-error effects The seismometers were emplaced at Bldg 845 Bldg 858 and the Linac Road sites for these shots The data were stored on magnetic tape for sub- sequent playback and analysis We show amplitude results in Figs 18-20 In these and subsequent figures the A-amplitude I is the amplitude of the zero-to-peak initial vertical velocity P-wave and the B-arnplitude is the peak-to-trough amplitude l8 Consider for example the A-amplitude data portion of Fig We have plotted the values of the vertical velocity ampli- tudes recorded at the Bldg 845 seismic station vs the base heights of the corresponding 23-kg 50-lb Comp C-4 HE charges J ' That there is no trend with HE base height value is obvious The mean amplitude of the seven experiments was 49 6 units the units are arbitrary--actually seismometer output in millivolts with a standard deviation of 8 5 units The two horizontal lines brack- eting most of the data depict the ±1 standard deviation limits The same type of information is shown for the B-amplitude data Figures 19 and 20 present the same type of data and analyses ·- for the stations at Bldg 848 and Linac Road respectively None of the results in any instance suggests that height of burst influences the seismic signal Pertinent details for the 45-kg KSA series are listed in Table 3 The A- and B-amplitudes obtained at the vertical motion text continues on page 51 0€ I I b 1 TOP SECR li 1f I lb h 1 fOf' SEC ET TCS-326 019 78 Page 48 Table 2 I· 1 -- · - - - - Firing details of the 23-kg SO-lb charges detonated in the evaluation of possible height-of-burst effects on seismic signals Shot designation a and location I ·-- HE HOB rn b Local date and approx detonation time Detonation time GMT KSO-A 802FP 23 kg 50 lb Comp C-4 0 3 1 30 78 3 28 p m 30 3438 10 014 KSO-B 802'FP 23 kg 50 lb Comp C-4 0 3 1 30 78 4 00 p m 31 0000 9 972 KSC-1 802FP 23 kg 5 0 lb Comp C-4 0 1 31 78 3 03 p m 31 2303 10 000 KSC-2 802FP 23 kg lSO lb Comp C-4 0 6 1 31 78 3 38 p rn 31 2338 10 083 KSC-3 802FP 23 kg 50 lb Comp c-4 0 9 1 31 78 1 48 p m 31 2148 9 973 KSC-4 802FP 23 kg 50 lb Comp C-4 1 5 1 31 78 11 49 a m 31 1949 10 006 KSC-5 802FP 23 kg 50 lb Comp C-4 2 1 1 31 78 2 28 p m 31 2228 10 001 I l -ve· designate the leveled grounded firing pad near Bldg 802 where most of the shots were fired as 802FP bHOB is the height of burst measured from the firing-pad surface to the base of the HE Typically the center of gravity of the HE was 100 to 150 mm higher than the base CV - • I Dot ku lOP SE6AET i TCS-326 019 78 Page 49 280 T-Jjf 240 i 200 B-amplitude mean E I 160 147 3 J • • • Cl std dev 14 9 iJ 1i 120 E • BO A-amplitude mean 49 6 std dev 8 5 8 0 40 0 0 '-- 1 - i - J --'----J ----'C --J -- ____ 0 0 2 0 4 0 6 0 8 1 0 1 2 HE base height - m Figure 18 1 4 1 6 _ __ __ _ ____ 1 8 2 0 2 2 IUI Amplitudes of the P-wave vertical velocity signals see inset recorded at the Bldg 845 seismic station vs the base heights of the 23-kg Comp C-4 charges detonated in the KSO and RSC series of experiments The horizontal lines are at ±1 std dev from the mean amplitude Pt fc 1 Cl lOP SECRit J 'OG - kl TOP sS CRE TCS-326 019 78 Page SO 40 30 · I I I I ' I ' I I Ttf' -- - B-amplitude mean 20 8 std dev 1 8 E I a 'C • I ' • I 20 ___ - t • - - ii E A-amplitude mean 8 90 std dev 0 89 0 8 0 0 0 0 I I I I I I I I r I 0 2 o 4 o 6 o a 1 0 1 2 1 4 1 a 2 0 HE base height - m Figure 19 2 2 U Amplitudes of the P-wave vertical velocity signals see inset recorded at the Bldg 858 seismic station vs the base heights of the 23-kg Comp C-4 charges detonated in the KSO and KSC series of experiments The horizontal lines are at ±1 std dev from the mean amplitude J - TOP SEER 'I' - - TCS-326 019 78 Page 51 - 50 40 l l 30 28 6 std - • dev 1 2 • - • • A-amplitude mean 10 11 std 0 - dev 0 64 --- n D 10 n -v I I I I I I I I I I 0 2 0 4 0 6 0 8 1 0 t 2 1 4 t 6 1 8 2 0 HE base height - m Figure 20 - l • 0 I I 1f 11 0 I I I I B-amplitude mean E I I 2 2 U Amplitudes of the P-wave vertical velocity signals see inset recorded at the Linac Road seismic station vs the base heights of the 23-kg Comp C-4 charges detonated in the KSO and KSC series of experiments The horizontal lines are at ±1 std dev from the mean amplitude b 1 r' _ I TCS-326 019 78 Page 52 I seismometers at Bldg 845 Bldg 858 and the Linac Road stations are shown in Figs 21 22 and 23 respectively These data again support the conclusion that there is no correlation between height of burst and seismic signal amplitude Table 3 Firing deta ils of the 45-kg 100-lb charges detonated in the evaluation of possible height-of-burst effects on seismic signals Shot designation and locationa HE and C-4 equivalent HOB b m 45 kg 100 lh LX-04-1 42 kg 93 lb C-4 0 902FP JCSA-i 45 kg 100 lh LX-04-1 0 6 JCSA-1 Local date and approx detonation time 2 2 78 10 34 a m 2 2 78 Detonation time GMT 33 1833 59 983 33 1909 00 082 802FP 42 kg 93 lb C-4 11 08 a m l SA-3 802 FP 45 kg 100 lb LX-04-1 42 kg 93 lb C-4 0 9 2 2 78 9 58 a m 33 1757 59 09 KSA-4 802FP 45 kg 100 lb LX-D4-1 42 kg 83 lb C-4 1 5 2 2 78 11 40 a rt 33 1939 59 953 J SA-5 802FP 45 kg 100 lb LX-04-1 42 kg 93 lb C-4 2 1 1 31 78 10 45 a m 31 1844 09 938 awe designate the leveled graded firing pad near Bldg 802 where ost of the shots were fired as 802FP bHOB is the height of burst from the pad surface to the base of the HE In Figs 24 and 25 we show the firing geometries of the two extremes of the KSA series For KSA-1 the charge was placed directly except for a thin layer of padding on the firing pad At the other extreme for KSA-5 the charge was supported by a 2 1-m-high firing table The HE charges however were identical The firing geometries of the KSC series were the same as for the KSA series Note that four shots from the KSD series are included in Figs 21-23 The KSD series will be discussed in more detail in text continues on page 58 CiGA br J er A r TCS-326 019 78 Page 53 400 I I I I l I l l I l i •• • • 300 - • • B-amplitude • I ara E 0 200 E In QJ - - - It ' C Cl ii E 100 0 n ' -·• 0 0 A-amplitude I I I I I I I I t 0 2 o 4 o 6 o a 1 0 1 2 1 4 1 a HE base height - m Figure 21 -- I 2 0 2 2 IU Amplitudes of the P-wave zero-peak A-amplitude and peak to-trough CB-amplitude vertical velocity sig- · nals recorded at the Bldg 845 seismic station vs the base heights of the 45-kg LX-04-1 charges detonated in the KSA series of experiments The results of four KSD series shots are included for ·c onvenience in visual comparison As discussed in the text we would expect these values to fall about 8% higher than the KSA-series values The ±1-std-dev lines shown refer only to the KSA-series data J · CJ t TCS-326 019 78 Page 54 -- -- - - - 50 r-'-- --- -- _- ---- --- ----------- 40 30 B-amplitude -a ' E ' · Q II ti ' - ii C E • • • • 20 • • 0 A-amplitude o 0 ___ __ ___ __ _ __ __ __ __ _ __ _ _ ___ _ ___ _ __ _ 0 0 2 0 4 0 6 o s 1 0 1 2 1 4 HE base height - m Figure 22 1 6 1 8 2 0 2 2 IU Amplitudes of the P-wave zero-peak A-amplitude and peak-to-trough B-amplitude vertical velocity signals recorded at the Bldg 858 seismic station vs the base heights of the 45-kg LX-04-1 charges detonated in the KSA series of experiments The results of three KSD series shots are included for convenience in visual comparison As discussed in the text we would expect these values to fall about 8% higher than the KSA-series values The ±1-std-dev lines shown refer only to the KSA-series data cGAhr J c t ' _I Lr I TCS-326 019 78 Page 55 J ·1 I I I I I I I ' I - - 60 I • '' i i ' I 50 'I • • • '-- • I • ' I I j J i E I 40 - '-- Cl CQ O a O ' · a - 0 E x • 8-amplitude - 30 ' - - C Cl J i ' ' 20 1 -· V - V V V -- I A-amplitude 10 - --- I l 0 ' 0 I I I I I I I r I I 0 2 0 4 0 6 0 8 o 2 1 4 1 6 1 8 2 0 HE base height - m Figure 23 2 2 U Amplitudes of the P-wave zero-peak A-amplitude and peak-to-trough B-arnplitude vertical velocity signals recorded at Linac Road seismic station vs the base heights of the 45-kg LX-04-1 charges detonated in the KSA series of experiments The results of four KSD series shots are included for convenience in visual comparison As discussed in the text we would expect these values to fall about 8% higher than the KSA-series values The ±1-std-dev lines shown refer only to the KSA-series data J 'V bl0 TQP SECR'-T 6G brr TCS-326 019 78 Page 56 I ' I U Figure 24 One of the simplest shots to field KSA-1 was the firing of 45 kg of LX-04-1 high explosive directly on the gravel firing pad Note the detonation taped to top of HE charge ClGA ' l l I r TCS-326 019 7E Page 57 r '' I • U Figure 25 At the other extreme from KSA-1 shown in Fig 24 the KSA-5 charge of 45 kg of LX-04-1 was detonated atop a crude but very sturdy wooden table 2 1 m high t t 'f b 0 j e-- bo T-OP SECltE'f I I TCS-326 019 78 Page 58 - ----------------------I the next section here we note that they were 45-kg charges of c- 4 rather than LX-04-1 and that their energy heat of detonation was about 8% greater than that of the RSA-series shots Thus we would expect signals to be about 8% larger than the KSA signals There is apparent agreement with these expectations Swnmarizing to this point we have shown that there is no correlation between height of burst and seismic signal amplitude over the complete range of heights of burst that can reasonably be expected in a nuclear-weapons-related hydrodynamics test program Geometric Perturbation Effects The KSD series of shots was designed to estimate the size of the perturbations likely to result from the differing firing geometries expected in nuclear-weapon hydrodynamics testing For e xample there could be a test involving half a device so that internal motion during the implosion phase could be studied Conservation of momentum and energy indicate that most of the such a device would directed For KSD-2 the HE was in an open ste el frame that supported such a slab of i r o n the HE and so it simulated a device fired with fhe open face downward For KSD-3 the slab was to the side of the HE simulating a device with the open face to the side For KSD-4 no slab was involved The same type and weight HE charge was used in each of the four firings This information and other pertinent firing details are summarized in Table 4 WJ TOP SECReT TCS-326 019 78 Page 59 Table 4 I Firing details of the 45-kg 100-lb charges detonated in the evaiuation of possible seismic signal variations resulting from variations in the geometrical configurations of the firings Shot designation and location a 1 I •' r I f i l J ' GMT 2 2 78 3 03 p rn 33 2303 00 002 2 3 78 10 58 a m 34 1857 59 931 HE None used 2 3 78 34 1947 00 0ll Beneath the KSD-2 802FP 45 kg 100 lb over the HE HE Comp C-4 KSD-3 802FP KSD-4 802FP ' Detonation time 33 2208 00 030 45 kg 100 lb Comp c-4 I Local date and approx detonation time 2 2 78 2 08 p m KSD-1 B02FP I HE Steel slab position 45 kg 100 lb Comp c-4 45 kg 100 lb Beside the 11 47 a m Comp C-4 awe designate the leveled graded firing pad near Bldg 802 here most of the shots were fired as 802FP In Fig 26 we show a photograph of the setup for event KSD2 The steel slab is r e sting at a slight angle from the horizontal on the steel frame housing the HE charge The angle was deliberate--we wanted to be reasonably sure which way the slab would go when the charge was detonated on detonation the slab was thrown nearly vertically upward landing about 60 m away forming an impact crater comparable in size to the one directly made by the explosion It is interesting that this impact does not seem to have been picked up by the seismometers This would seem to indicate that energy expended in crater formation is not a major contributor to the overall energy coupling that creates the seismic signature from an HE airburst A calculation of the energy required for crater excavation is in fact consistent with this observation • oc A brq TCS-316 019 78 Page 60 U Figure 26 Firing arrangement for KSD-2 Inside the steel framework is 45 kg of Comp C-46 in two boxes The 600-mm-square 100-mm-thick steel slab was hurled nearly vertically upward The slight angle at which it was mounted caused the slab to land 60 m away in a safe direction -SECREI TCS-326 019 78 Page 61 '' I I I From conservation of energy and momentum constd·e ratlons we would expect shot KSD-2 with the steel plate over the HE to impact the ground with the greatest energy This seems to have happened for the KSD-2 amplitude is larger at each station than that of any of the other three shots in the series It is the uppermost point in each of the KSD-series results plotted in Figs 21-23 The ratio of the KSD-2 amplitude to the mean amplitude of the other three shots of the KSD series averaged 1 14 for the six sets 3 seismometers x 2 types of amplitudes with a standard deviation of 0 07 If this difference is real the geometrical effect approximating the worst-case situation is thus about 15% Statistically the signals and amplitudes from KSD-1 -3 and ·-4 and perhaps even KSD-2 were the same In any event we may conclude that differing geometrical configurations in nuclear weapons hydrodynamics HE testing produce only small variations in the seismic signals--probably less than about 15% Relationship Between HE Yield and Se ismic Signal Amplitude In this section we address the question To what extent can we estimate the relative size of an HE charge from the amplitude of the seismic signal produced by its detonation The KSG series of tests involved the firing of a range of HE masses with _ _ _and other parameters fixed _· actual weight of the charges their equivalent weights in Comp c-4 high explosive and other firing details are listed in Table 5 The total weight of the charges required was 705 kg 1555 lb and to expedite the program and to economize we obtained scrap HE from DOE's Pantex plant J o£ b6J For our analyses we converted LIOc b 1_ fOti ·r eRel TCS-326 019 78 Page 62 I from actual weight to equivalent weight of Comp C-4 on the basis of experimental heats of detonation Each of the shots was fired directly on the firing pad The actual firing points may have varied by as much as 6 m from one another The KSA and KSD series of shots discussed earlier indicated this much variation in position should not perturb results Indeed if it did there would be no practical point in continuing the experimental program The seismometers were emplaced in the same positions they occupied for all the firings discussed so far Table 5 __ -- - details of the HE charges detonated in the evalof the relationship between yield and seismic amplitude -- ___ _ TCS-326 019 78 Page 63 In Figs 27-31 we have plotted P-wave A- and B-amplitudes vs the c-4 equivalent weights of the charges We consider the results from Bldg 845 last because we found it necessary to reduce the gain of that seismometer part way through the series when it became obvious that the larger shots would overload the instrument The complications of this gain change will be discussed later in this section The A-amplitude calibration points for the vertical-motion seismometer at Bldg 858 are shown in Fig 27 The point for the 98-kg 215-lb charge is missing by accident it was not recorded The data points suggest fitting a straight line on a full logarithmic plot In other words we seek a best fit to the equation I amplitude a equivalent weight b where a and bare constants selected to give the best fit in the least-squares sense 9 Performing this operation we obtain the regression line A-amplitude 4 92 wt 0 • 816 wt in lb with coefficient of determination r 2 0 952 0 Since variation r 2 explained total variation 1' the value of r 2 we have obtained indicates a very good fit indeed as we see in Fig 27 The least-squares fit or regression line is plotted in the figure text continues on page 68 DOE 1lO leP SEC E j I• l ir TCS-326 019 78 Page 64 Comp C-4 equivalent weight - kg 102 10 E A-amplitude 4 92 wt0 •816 r2 0 952 10 Comp C-4 equivalent weight - lb Figure 27 UJ Least-squares fit to t he A-amplitude data points obtained at the Bldg 858 station The 80% confidence interval on the predicted A-amplitude is also shown - -- DS A ·- ' be I I TCS-326 019 78 Page 65 Comp C-4 equivalent weight - kg 103 10 1 1 0 • - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - J I e B-amplitude 21 32 wt0 ·699 r2 0 970 I - e 1i Ji dJ confidence interval 10 _____ __ __--- L L ___ 10 __ __ _ ____ _ _ __ Comp C-4 equivalent weight - lb Figure 28 __ _ _ i 102 U Least-squares fit to the B-amplitude data points obtained at the Bldg B5B station The BO% confidence interval on the predicted B-arnplitude is also shown • • l OG bt JOG b6J fOP SECRET TCS-326 019 70 Page 66 Comp C-4 equivalent weight - kg 02 10 A-amplitude 6 653 wt0 •727 r2 0 986 102 confidence interval 1 10 Comp C-4 equivalent weight - lb Figure 29 o2 U Least-squares fit to the A-amplitude data points obtained at the Linac Road station The 80% confidence interval on the predicted A-amplitude is also shown cx A £r l oE sb U TGP SECReT SECRET TCS-326 019 78 Page 67 Comp C-4 equivalent weight - kg 1 10 103 104 r---r------ -- - --- --- 'T --r--r--r- ----r-- --- ----- oJ E I confidence interval B-amplitude 25 47 wt0 -703 r 2 0 984 10 L---'----'------''-- 1--'- L - -- L --'---'----'----''--- U 1 10 102 Comp C-4 equivalent weight - lb Figure 30 CUI Least-squares fit to the B-arnplitude data points obtained at the Linac Road station The 80% confidence interval on the predicted B-arnplitude is also shown l L rJ T C5-326 019 78 Page 68 Comp C-4 equivalent weight - kg 10 1o2 Least-squares fits to merged data - - - Least-squares fits to solid points Least-squares fits to open points o • A-amplitudes a• B-amplitudes a C - a E RI 1 10 C-4 equivalent weight - lb Figure 31 Least-squares fit to Bldg 845 data UI The solid points are scaled by a constant multiplier to optimize the fit with the other points see discussion in text D o · 1ft lOP SECR l cc L ' ' TCS-326 019 78 Page 69 We followed the procedures given in Gilbert 9 to obtain the I limits within which ' the amplitude will fall with given probability for a given charge weight These limits for an 80% confi- dence interval are also shown in Fig 27 For example for a 36- kg BO-lb charge from the figure or from the equation we see that a value of 175 mV is our best estimate for the A-amplitude We also find we are 80% confident that the measured amplitude j J will be between 128 and 240 mV Figure 28 is the same as Fig 27 except that it shows the results of the analysis of the B-amplitude data obtained at the Bldg 858 seismic station Figures 29 and 30 show the results obtained at the Linac Road station The Linac Road results were the best we obtained the coefficients of determination were greater than 0 98 for both the A-type and the B-type signals Estimates of HE weights to a factor of 1 5 appear possible at about the 95% confidence level for the Linac Road station What we might have obtained at Bldg 845 we can only surmise from the results shown in Fig 31 As already pointed out it was neces- sary to reduce the system gain after the smallest four shots Unfortunately the gain is frequency-dependent and the change in gain cannot be accounted for by simply multiplying the signals by a constant factor For illustrative purposes only we have done this in Fig 31 but will draw no conclusions from the merged data The Bldg 845 data is thus treated in further analysis as two separate sets of data for both the A- and the B-amplitudes The values of the exponent or slope parameter b in the fitted equations of the form amplitude a equivalent weight in lb b and the corresponding standard deviations of bare collected in Table 6 The overall best estimate of bis 0 724 with standard deviation Sb of 0 018 of Table 6 suggests that dif- i· TCS-326 019 78 Page 70 Table 6 The slope parameter from a regression analysis of least-squares fit of the KSG-series results to the e9uation amplitude a wt b Weighted averages are listed for the A- and B-amplitude results separately and in combination Shots included Slope oarameter b 845 A KSG-1 through KSG-4 0 756 0 069 845 A KSG-5 through KSG-9 0 723 0 0139 845 B KSG-1 through KSG-4 0 689 0 0134 845 B KSG-5 through KSG-9 0 550 0 087 858 A All KS but KSG-7 0 816 0 075 858 B All KSG but KSG-7 0 699 0 050 Linac Road A Linac Road B All KSG series 0 763 0 034 All KSG series 0 703 0 034 Seismic station and signal type All results weighted average Standard deviation of the slop1 51 6 0 724 S5 0 018 A-type results only weighted average B-type results only weighted average bA 0 765 S5 bB 0 027 A 0 687 S5 oA - oB B 0 026 0 078 s E' -E' A B 0 037 ferent values of b may be associated with the two signal types We have included in the table estimates of n and Sb for each of the signal types separately and have also included estimates 10 of the difference of the slope parameters b - oB and the standard deviation of the difference S bA - -bB A It seems there has been little if any previous work applicable to the problem of determining HE charge size for seismic TCS-326 019 78 Page 71 amplituqe in the charge-size range and explosion environmen o f interest to the proliferation problem There have been efforts to evaluate the size of seismic signal from underground nuclear 11-13 In this case however not only is the exploexplosions sion environment different but the energy regime is tremendously greater Even so there appears to be reasonable agreement in the slope parameters obtained Slope parameters quoted for the underground nuclear explosions range from about 0 6 to 1 with most lying between 0 7 and 0 8 All in all we have no direct comparisons with other work I l' · in particular with seismic signals from chemical explosions inference i t appears our results are reasonable and that we By might find different slope parameters for different geologic media It should not be difficult however to determine empirically the appropriate slope parameter for an Nth country geology with a few e xperiments in an appropriately matching geology The results of such experiments should be directly applicable to determining relative and probably absolute HE charge sizes in an Nth country proliferation assessment without in situ calibration Dependence on Type of High Explosive Most of our experiments were made using high explosives having heats of detonation nearly equal to their heats of combustion Thus there was no afterburn in atmospheric oxygen We tried a few charges of TNT and one charge of an aluminumcontaining blasting agent Each of these explosives has a heat of combustion appreciably larger than its heat of detonation When analyzed these e xperiments should provide information useful in determining equivalent weights of high explosive insofar as seismic signals are concerned Poe- l OE' Te f' --' Ee T TCS-326 019 78 Page 72 Shot7Site Variation Effects We fired several shots individually in a cross-shaped pattern centered on the Bldg 802 firing pad to provide inform tion on how seismic signals will vary for moderate changes of firing position We also fixed charges at four other firing bunkers at distances from the reference pad comparable to the seismometer distances The results have not yet been analyzed but it appears that subs tantial changes in seismic signal amplitude can result from changes in local ge qlogy One shot fired on a sandstone outcropping roughly 100 m from the main firing point gave markedly different signal amplitudes and signal shapes Distant Station Data Data from the distant lll 5-km station are not yet analyze d in detail The P-wave signals were apparently lost in the rather considerable environmental noise and only the acoustically coupled signal was detected Summary and Conclusions In a series of experiments we have measured the vertical velocity amplitudes of seismic disturbances at distances of 1 to 3 km from HE charges in a size range covering those expecte d in nuclear weapons hydrodynamics testing We have found to date 1 2 Over a reasonable range of burst heights there is no variation in seismic signal amplitude for a fixed HE mass Over a range of device geometry differences simulating those expected in nuclear weapons hydrodynamics testing the seismic less than about 15% TCS-326 019 78 Page 73 3 Over a range of HE charge sizes expected in nuclear weapons hy rodynamics testing the relationship amplitude a equivalent charge weight b holds a is a system calibration constant and bis the slope parameter a constant equal to about 0 72 but possibly depending on the shot seismometer geology i I t ·' « I br Jx £ L Q TOP SECl'ET TCS-326 019 78 Page 74 REFERENCES _ P J Smit The Karoo System in the Kalahari of the Northern Cape Province Ann Geol Surv s Afr 2_ 79 1971-72 J D Fairhead and N B Henderson The seismicity of Southern Afric a and Incipie nt Rifting Tectonophysics i _ Tl9 1977 LLL DISA Proliferation Stud Quar terl u y-Septern er 977 Law ence Livermore 1093 1977 title U report SRD i ''· QL- See any text on statistics for example N Gilbert Statis W B Saunders Co Philadelphia 1976 • • Y Beers Introductio n to the Theory of Error 2nd ed Addison-Wesley Publishing Co Reading Mass 1957 D L Springer and W J Hannon Bull Seismal Soc Am § l 477 1973 and J A Lahoud Bull Seismal Soc Am 59 • ' ep •see ftl T SiiMM ffitl arepi ESRliT