Johnston Island

. - , •\' Report OEHL TR-78-87 CEEOO-TR-78-38 LEVEi? LAND BASED ENVIRONMENTAL MONITORING AT J~HNSTON ISLAND - DISPOSAL OF HERBICIDE CAANGE - 1 DOC FILE_._C,Ql1J PREPARED BY: - ·~o D C T. J. Thomas. D. P. Brown. J. Harrington. T. Stanfor4. L. Taft, and B. W. Vigon BATTELLE Columbus Laboratories 505 King Avenue Columbus, Ohio 43201 A September 1978 Final Report for Period 11 May 1977 - 30 September 1978 ! Approved for public release; distribution unli111ited ! PREPARED FOR: V US AIR FORCE OCCUPATIONAL AND ENVIRONMENTAL HEALTH LABORATORY Brooks AFB, TX 78235 .• / ·~ ARMANENT DEVELOPMENT AND TEST CENTER/ • Detachment 1 ;/ •~- i Tynda 11 AFB• FL 32403 -:f \~ iJ1 t' "'~ I 19 IO 24 v26 /.j;~ ',-~ ' i ~ i ··-:s--•---------.~-~~,j-;:""I!!-. ■.R~M'i-rt'_,,.,,.,.il·"'"'II...-.. ---------· - - - - - - - - - - , - , ' .~ ,. ; ' ~·· ~;( . , 1 ;t:,:- • "''•':~ .,...,.. •,; ..:. ii Ji 3 I» I 4 C a • 2ijlj..'8iiiii' C ♦ W

' NOTICES This ~epQrt has been released to the National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161, for sale to the general public. Qualified requestors may obtafn copfes of thfs report from .Defense Documentation Center (DOC), Cameron Station, Alexandria, Virginia 22314. ... This technical report has been reviewed and is approved for publication. __ ./ ,,. A..,, £ JJ'll'!,.A .. • 4...,...£ WILLIAM E. MABSON, Colonel, USAF, BSC C0111111nder

,, UNCLASSIFIED SIFICATION OF THIS PAGE (ll'hon Doro Enlorod) Land Based Environmental Monitoring at Island - Disposal of Herbicide O~a~ge, _____ __, 78, MBER 9. PERFORMING ORGANIZATION NA,!"E AND ADDRESS PROGRAM EL.EMENT, PROJECT, TASK AREA 6 WORK UNIT NUMBERS Battelle Columbus Laboratories 505 King Avenue Columbus, Ohio 43201 F0430P I 1. CONTROL.L.ING OFFICE NAME AND ADDRESS Arnament Development and Test Center Detachment 1 (!£_~;;-;7~ T ndall AFB FL 324~ 323 4. MONITORING AGENCY NAME 6 ADDRESS(/1 dlllerenr lrom Conlrollln/1 01/iceJ IS. SECURITY CL.ASS. (o/ thlo roporl) U.S. Air Force Occupational and Environmental Health Laboratory 0-·i:}-;,;);J,--7 UNCLASSIFIED Brooks AFB, TX 78235 ~ ~-,.--e._ ISa. DECL.ASSIFICATIONIDOWNGRADING SCHEDUL.E 16. DISTRIBUTION STATEMENT (ol thlo Report) Unlimited 17. DISTRIBUTION STATEMENT (ol tho ol>otrect onterod In Bloc/< 20, It dlflorent from Report) II. SUPPL.EMENTARY NOTES The US Air Force technical manager for this project is Major James W. Tremblay (512) 536-3491. ft. KEY WOADS (Continue on rev•r•• •Id• II nec•••a,y and Identity by bloclr number) Sampling, Analysis, Herbicide Orange, herbicide, water, sewage, air, biota. -: ABSTAACT (Contlnw on r•••r•• aid• II neceaa_,,, Md Identify by block number) z:l As a part of the U.S. Air Force final Environmental Impact Statement of incineration of Herbicide Orange from Johnston Island, the Air Force stated that a monitoring program would be conducted to demonstrate that the land- based transfer operations were carried out in an environmentally safe manner. Battelle, Columbus Laboratories conducted these monitoring programs on Johnston Island. The monitoring of at-sea incineration operations were conducted by TRW / and have been reported elsewhere. ---~.-=-, /~ 00 ,::."'\, 1.,, ""'o;;:;,.:;;-;;;:;:;r r1"' ., ~~ - wc:u.ssmED SECURITY CL.ASSIFICATION OF THIS PAGI! (ll'llon Dato Snlor- V;---' ~ a (& ••• •• 0 •

UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE(ll'IIM Dala lfnlared) 19. -z ---- oncentrations of 2,4-D and 2,4,5-T found in the ambient air and water - ...sa ples were minimal. No TCDD was detected in any air or water samples. No changes that could be attributed to the operations were noted in indigenous / plant or bird populations. Results of quality control sampling revealed that the required level of drum cleaning was achieved. Industrial Hygiene evaluations of the land-based operations revealed that only minor accidents or injuries I- occurred and exposure of workers to airborne 2,4-D and 2,4,5-T were well below permissable levels. __) r\,~No adverse consequences of the minimal release of Herbicide Orange into the Johnston Atoll environment were observed. :::,_ ,,,,,. This report was submitted in fulfillr,e t of Contract No. F08635-76-D-0168 by Battelle, Columbus Laboratories, under ponsorship of the U.S. Air Force. This report covers the period May 1977 to Sept. 1978. UNCLASSIFIED S!CUIUTY CLASSIFICATION 01" THIS PAGE(ll'llan Da1a lfn1ered) & uswae a 0. •

DISCLAIMER NOTICE THIS DOCUMENT IS BEST QUALITY PRACTICABLE. THE COPY FURNISHED TO DTIC CONTAINED A SIGNIFICANT NUMBER OF PAGES WHICH DO NOT REPRODUCE LEGIBLY. () f.<_ 0-Q__,'-Q__, ~ I I _J L CL:1, IL- e~/7 -H,(.~- (;QL~' f~y1 J<.Ln1GVl{,'D ' BEST AVAILABLE COPY

PREFACE Headquarters US Air Force Logistics Conmand, Wright-Patterson AFB, OH, the office of primary responsibility for the project to dispose of Herbicide Orange, designated the US Air Force Occupational and Environ- mental Health Laboratory (USAF OEHL) as the agency responsible for land based environmental monitoring of this project. The Armament Development J and Test Center, Tyndall AFB, FL negotiated and monitored this contract with Battelle Columbus Laboratories, Columbus, OH. Personnel of the USAF OEHL served as Technical Representatives of the Contracting Officer. Accession For Nns GF.MI DDc TAB ~announced Justitic;.it i ::::i --~--------- B7_ __ - -... ----~ ... ' ,:,.;;;..· ·.·:\f-~• ;;.;;~: ·, • l:e:~4;:;->z £~ ~+i-•; ---- st C a :s. a u , .• • 4 -• ------------~-----~-

---------------------------~----- ~ USAF CEEDO TR-78-38 FINAL REPORT on LAND BASED ENVIRONMENTAL MONITORING AT JOHNSTON ATOLL - DISPOSAL OF HERBICIDE ORANGE to U.S. AIR FORCE OCCUPATIONAL AND ENVIRONMENTAL HEALTH LABORATORY September, 1978 PART I by Thomas, T.J., Brown, D.P., Harrington, J., Stanford, T., Taft, L., and Vigon, B.W. U.S. Air Force Project Officer: James W. Tremblay, Maj., USAF l1 1-,- !'i, < :t .('..... , I:-, j: l, •-.·~ ~- BATTELLE Columbus Laboratories i I. t l 505 King Avenue \ lr- Columbus, Ohio 43201 -..,. :~.:¢. ,<;;,,_:.,:;'••i ;. :· ··«<e:r► U -~.,:i,•.j • • 04 C. • 54 C

DETAILED ENVIRONMENTAL ANALYSIS OF PROJECT PACER HO 1. INTRODUCTION MATERIAL This report is Part I of a three-part report on the environmental consequences of a project conducted on Johnston Island, labeled Project Pacer HO, designed to remove and incinerate the stocks of Orange Herbicide stored on Johnston Island since 1972. The three parts to the report are as follows: Part I Executive Summary Part II Detailed Environmental Analysis of Project P:icer HO Part III Supporting Data In April, 1970, the Secretaries of Agriculture, HEW, and Interior jointly announced the suspension of certain uses of 2,4,5-Trichlorophenoxy acetic acid. As a result the Department of Defense suspended the use of Orange Herbicide since this herbicide consists of approximately 50 percent 2,4,5-T and 50 percent of 2,4-Dichlorophenoxy acetic acid. This suspension left the Air Force with 1.5 million gallons of Orange Herbicide (HO) in Vietnam and 0.8 million gallons in Gulfport, Mississippi. In September, 1971, the Department of Defense directed that the Orange Herbicide in Vietnam be returned to the United States and that the entire 2.3 million gallons be disposed of in an ecologically safe and efficient manner. The 1.5 million gallons were moved from Vietnam to Johnston Island for storage in April, 1972. The cost of maintaining the storage areas, and the ever present danger from the stored HO stocks, let the Air Force to conduct a study to develop procedures for the ecologically safe, efficient, and, if possible, , ,... low-cost disposal of the approximately 2.3 million gallons of HO. I As part of their final EIS,* the Air Force stated" a monitoring 1 .. ,;. * The final EIS for incineration of HO at sea, There were public hearings, and an EPA ~cean dumping permit was issued. 1 . , , l 1-. -~-----·--- -------- .-..;,;. • ':y,'w)•' 1.._..:.; : . • //.,1' !'•:~pc •~.:~»- as C a » a :e C C W1 • • 1 -

2 program will be conducted to document herbicide exposures and environmental exposures should they occur. It is anticipated that this program will generate sufficient data to demonstrate that personnel and environmental safety of this operation". This report contains the results of the land- based monitoring program conducted during the HO disposal program on Johnston Island. 2. THE ORANGE HERBICIDE DISPOSAL PROGRAM The Orange Herbicide stored on Johnston Island represented approximately 25,000 drums of 55-gallon capacity. These were stored in rows stacked three high in an area of about 3.5 acres on the northwest corner of the island, where the prevailing winds rapidly removed any atmospheric HO away from Johnston Island and the atoll and dispersed it in the open Pacific. There were no other locations containing HO. Prior to the disposal operation, the sea environment caused drums to corrode and thus leak. The leakers were taken to a dedrumming facility where they were allowed to drain and were redrummed and restacked, while the old drwns were crushed and stacked. The leaked HO caused a persistent and intense odor downwind of the drumyard. For the HO disposal program, the dedrwn facility was modified to allow transfer of the material from drums to bulk carriers for transport to an incinerator ship. The facility and operation basically consisted of a concrete pad and two fabricated metal racks upon which the full drums were placed in four gr0.ups of 12 each. Drums were transported from the drum yard to the racks in sets of four. The drums were then drained into a collection sump and spray rinsed twice with diesel fuel, exceeding the quality EPA requirements of 90 percent confidence of 85 percent residual removal. After drainage, the drums were carried to the crusher, which consisted of a large weight suspended between two I-beams. The drums were compressed along the longitudinal axis. Crushed drums were bundled and placed in storage on the seaward (downwind) side of the dedrum/cr.ushing area. A large plaatic sheet was used to protect the crushed drums from rain • . .:,;,. , ' ~ ,.;·' , • ""~~->4 •en..--·

3 Herbicide was pumped from the collection sump into standard Air Force R-5* refueling trucks via a dry coupler bottom connection. The refuelers transported the HO to the wharf via a road which was set aside for this purpose. Non-project related vehicle traffic was forbidden along this section of roadway. Once the refueler had reached the main wharf, the procedure was essentially reversed. The same type of dry couplings and spill prevention equipment were employed to pump out the tank and bulk transfer the material to the M/V Vulcanus, a ship designed for the incineration of hazardous materials. The area in which the pumps and hoses were located was diked with sand bags and plastic so that potential spillage could be contained. The drum rinsing activities were subjected to constant monitoring to assure compliance with the EPA requirements. The second rinse from every 100th drum was sampled and analyzed for HO. A quality control chart was com- piled from these analyses to assure that EPA requirements were being met on continuous basis. A certified industrial.hygienist was present during the complete operation. In addition to preventing deficiencies in personal hygiene and safety, he was responsible for the siting and operation of personnel samplers. 3. AIR Surface trade winds were essentially constant throughout the study period with winds from the ENE to ESE at 10 to 20 mph on most days. Being remote from other terrestrial environments, the air at Johnston Atoll is clean, with none of the pollutants normally associated with urban areas. Air sampling for 2,4-D and 2,4,5-T was accomplished utilizing Chromosorb 102 as an adsorption mediwn, a granular polymer well suited for collection of chlorinated hydrocarbons. This material was packed in micropipet tubes through which a sample volwne of 150 liters was pulled at the rate of 0.50 liters/minute. * On termination of the project, all equipment was decontaminated with a diesel fuel wash, which was then loaded on the ship.

4 .I i Air sampling for the herbicide contaminant, tetrachlorodibenzo- para dioxin (TCDD), was accomplished utilizing benzene as the absorption medium. The apparatus consisted of a train of four impinger columns, the first two contained benzene, and the final two contained activated carbon to trap evaporating benzene. In order to determine the impact of dedrumming and transfer operations on the air environment, four monitoring areas were chosen for sampling. These were the meteorology building (located 2 miles upwind for use as a background station), the wharf (300 feet downwind of the loading area), the dedrum facility ( to determine occupational exposures), and a point 310 feet downwind of the dedrum facility. The chromosorb samples taken over the duration of dedrumming and loading operations yielded the following observations: • Concentrations in samples taken at the upwind meterology building ranged from levels below detection to trace amounts ( less than 1 microgram per cubic meter). • There was little difference between data recorded at the meterology building and that at the wharf. The impact on air due to the loading procedure at the wharf was negligible. • Total herbicide * concentrations detected 310 feet down- wind of the dedrum site ranged from 3 to 23 micrograms per cubic meter. • Concentrations inside the dedrum facility were only slightly higher, from 7 to 27 micrograms per cubic meter. The OSHA 8-hour time weighted average allowable concentration for either/or 2,4-D and 2,4,5-T acids is 10 milligrams per cubic meter. All of the ambient measurements were neRli~ible in comparison to the OSHA TWA. The analytical results on air samples in the dedrumming facility show that personnel exposures were two to three orders of magnitude below the TLV of 10 mg/cubic meter for either 2,4-D or 2,4,5-T. No injuries or illness that occurred during dedrumming could be attributed to HO exposure. ** * Concentration reported as sum of 2,4-D and 2,4,5-T. ** Two cases occurred when HO was splashed in eyes, The eyes were immediately flushed without consequence. • st C a » ,...w ,. , w

5 Analysis of twenty benzene impinger samples showed all samples to contain less than the mininum detectable limit (MDL) of TCDO. MDL's ranged from 6.6 to 20.3 nanograms P<:." cubic meter. The impact of the disposal operation on the atmospheric environment was thus found to be insignificant. 4. WATER The existing water environment of Johnston Island consists of several components of the hydrologic cycle. The saltwater cycle is comprised of the lagoon circulation and the groundwater underlying the island while the freshwater cycle includes the rainfall and the drinking water and sanitary system. Johnston Island's water system uses both fresh and saltwater. The saltwater around Johnston Island and the freshwater system have been monitored for the presence of 2,4-D and 2,4,5-T since 1973. The maximum concentrations observed in the offshore area near the herbicide storage were on the order of 3 ug (micrograms) 2,4-D/liter and 0.6 ug 2,4,5-T/liter and those near the saltwater intake were 2.3 and 0.7 ug/1, respectively. The other two offshore sites exhibited maximum concentrations below 0.5 ug/1. Sample taken in the distillation plant never showed measurable concentrations, yet one sample from the storage reservoir showed 1.6 ug/1 of 2,4,5-T. By comparison, most stringent standard appears to be the National Interim Primary Drtnking Water Standard at 0.1 mg 2,4-0/1. The sampling program for the water environment during the oper- ation consisted of four offshore sites and two onshore sites. Samples were taken of the water near the main wharf at two points just off of the bow of the ship at 10-11 meters of depth. The saltwater intake for the desal- ination plat. ~as sampled daily at about the same times as for the wharf samples and at a depth of five to six meters (about one meter from the bottom). The third offshore location sampled on a regular basis was the sewage outfall on the south side of the island. The fourth offshore site, sampled four times, was the shallow offshore area near the drum storage yard. i ~ ~t.' • ~. ' .; ..... L a••• ,.,,.., C 0 .. •'

6 The location of one of the onshore samplers was in the fresh- water system equilization tanks immediately downstream from the desal- ination plant and prior to chlorination. The other onshore sampler monitored sewage in a sump near a lift station. The water in the vicinity of the intake for the desalination plant was monitored on a daily basis. The level of herbicid~ ranged from below detection limits (O.lppb} to 3.43 ppb. Over 50 percent of the samples analyzed had concentrations below 0.2 ppb, a factor for 500 less than the drinking water standard. Potable water samples taken before the operation showed trace concentrations of 2,4-D in one sample. During the operation, herbicide concentrations * were found at trace levels (O.l - 0.2 ppb) in 20 percent of the samples, again a factor of 500 below the drinking water standard. Water samples were taken on alternate days in proximity to the sewage outfall, which is approximately 550 feet offshore. Only trace level of either 2,4-D or 2,4,5-T (0.1 - 0.2 ppb) were detected in the samples analyzed. The sewage samples, contaminated from the washing of work clothes showed concentrations of herbicide ** of from 20.7 ppb to 137.8 ppb. An estimated total of 0.94 pounds of herbicide was released into the sewage system, a markedly small figure in coll)parison to the amount hanlled. Water samples were taken offshore and downwind of the dedrum facility four time during the operation. One sample contained trace levels of 2,4,5-T while all other samples analyzed had no detectable levels. Water samples were taken on a daily basis in the vicinity of the wharf, which included special grab samples during the two deballasting periods from the M/V Vulcanus. The water in the immediate vicinity (10 feet) of the deballast discharge contained levels of herbicide that ranged from below detection to 8,117.7 ppb. The concentrations of these chemicals in the composited water samples at the wharf in the days following the deballasting illustrated an effective dilution process. The concentrations of herbicide dropped from 8116.7 to 1.90 to .75 ppb in the 2 days following the second deballast period, Including the deballasting periods, the concentrations of both 2,4-D and 2,4,5-T stayed below 0.2 ppb (trace) in over 50 percent of the samples taken. * Concentration is reported as sum of 2,4-D and 2,4,5-T. ** Concentration is reported as sum of 2,4-D and 2,4,5-T • .. ·····-·-·-·--··-·-- • ttJ• --------------------- • ...,.,,,.,,,. CC .. •

7 The 11 water and sewer samples analyzed to date for TCDD show no measurable concentrations (MDL's ranged from 3.6 to 8.0 nanograms per liter). With the exception of the deballast operation, the effect of the disposal operation on tne aquatic en~ironment was found to be in- significant. The deballast operation produced no signs of biotic impact, and aquatic concentrations decreased rapidly to nearly undetectable levels after deballasting• 5. BIOTA ) The terrestrial environment of Johnston Atoll has been extensively studied. Although large numbers of aquatic, terrestrial, and avian species have been identified at Johnston Atoll, there is a paucity of native species, the atoll being a link in a migratory chain. The large number of birds present on the atoll were nearly exclusively found on the three islands, unaffected by the presence of thl' disposal operation on Johnston Atoll. No signs of aquatic distress or change were noted in any aquatic community during disposal operations. Young, potted tomato plants, Lycopersicon esculentum, 25-38 cm in height were used as biomonitoring organism to detect the presence of Orange Herbicide in the air. Tomato plants were used because of their sensitivity to HO damage in the parts per trillion range. The injury symptom typical of HO damage, know as epinastic growth, is described as a curling and/or twisting of the apical portion of the plant. Fourteen air biomonitoring sites or stations were selected on Johnston Island. Three days of preoperational observations indicated that concentrations of Orange Herbicide sufficient to cause injury to the tomato plants only at two of the 14 stations. These two stations were approximately 500 feet from the dedrumming site and directly downwind. During the operation, these two stations experienced the most frequent and most severe injury. Ocrasional damage was experienced at two peripherally downwind stations. However,during the monitoring program, no significant physical or morphological changes were noted in any indigenous plant species on Johnston Island attributable to Orange Herbicide. 1 :;;.: ~.-.~,.:i~ .,.;'.' • ,,/. a•r4i!E.:>' •~~~- Ca • ..,....,,P""'•-••-•...,••--------~••--:;:;...

8 I 6. QUALITY CONTROL OF DRUM RINSING Statistical sampling was made of drum rinse samples to assure the residual in the drums was less than that which would be left by the EPA triple rinse procedure. The drum rinse procedure was modified several times to improve removal; the drums on the average exceeded the required triple rinse efficiency. 7. SITE RECLAMATION The U.S.A.F. has developed a continuing soil sampling program on Johnston Island, in the area of the drum storage yards. The purpose of the program is to monitor the degradation of HO in the old seepage areas from drum storage, so as to assure that the residual poses no environmental threat. L J a cs 6 5c_: • t

USAF CEEDO TR-78-38 FINAL REPORT on LAND BASED ENVIRONMENTAL MONITORING AT JOHNSTON ATOLL - DISPOSAL OF HERBICIDE ORANGE to U.S. AIR FORCE OCCUPATIONAL AND ENVIRONMENTAL HEALTH LABORATORY September, 1978 PART II by Thomas, T.J., Brown, D.P., Harrington, J., Stanford, T., Taft, L., and Vigon, B.W. U.S. Air Force Project Officer: James W. Tremblay, Maj., USAF BATTELLE Columbus Laboratories 505 King Avenue Columbus, Ohio 43201 , I .. · ~-~·¥, ···tr~·. 1' .,.: ?~~ a a . i(JSJ•eu • • a

TA~LE OF CONTENTS Page 1. INTRODUCTION . . 1 1. 1 Background. 1 1.2 Need for Field Operations 2 1.3 Application of NEPA . . . 2 2. EXISTING ENVIRONMENTAL FEATURES OF JOHNSTON ISLAND 3 2.1 Physical. • . . 3 2. 1. 1 Land . . 3 2.1.1.1 Location. 3 2.1.1.2 Topography. 4 2.1.1.3 Geology 5 2.1.1.4 Soils • . . 7 2.1.2 Air . . . 7 2.1.2.1 Meteorology During the Interval 7 a. Wind Speed and Direction. 8 b. Temperature. 8 c. Precipitation 15 d. Dew Point . . 15 2.1.2.2 Air Quality . . . 15 a. OSHA Standards. 16 b. Existing Pollution Source. 16 c. O~served Ambient Air Pollution. 18 2.1.3 Water Environment . . . . . . . . . . . . 18 2.1.3.1 Hydrology of Johnston Islan.d . . . 19 2.1.3.2 Oceanography-Currents and Tides. 20 2.1.3.3 Water Quality Criteria/Standards. 25 2.1.3.4 Existing Water Quality 39 2.1.4 Groundwater Quality • . . • • . . • • . 42 2.2 Biological Environmental Features of Johnston Atoll 43 2.2.1 Terrestrial Environment . • 43 2.2.1.1 Plants . . • . 43 2.2.1.2 Invertebrates 45 2.2.1.3 Vertebrates. 45 a. Fish . . • 45 b. Reptiles. 45 c. Birds . 45 d. Mammals 53 2.2.2 Marine Environment. 53 2.2.2.1 Plants • • . 53 2.2.2.2 Invertebrates 54 2.2.2.3 Vertebrates • 54 '.... a. Fish •• 54 . b. Mammals 55 i ' . ___ --------------------!!!! I /1

' TABLE OF CONTENTS (Continued) I 2.3 Human Environment . . . . . 55 2.3.1 Economic and Social. 55 3. DESCRIPTION OF ORANGE HERBICIDE DISPOSAL PROGRAM 56 3.1 Purpose . . . . . • . . . • . 56 3.2 Operational Procedures . . . . 56 3.2.1 Physical Manipulations 56 3.2.1.1 Drum Handling-Dedrumming. 56 3.2.1.2 Drum Crushing . • . . . . 65 3.2.1.3 Transport of HO to Disposal Ship. 65 3.2.1.4 Transfer to Disposal Ship 65 3.2.1.5 Cleanup . . . . • . • . • . 66 3.2.Z Descriptions of Project Activities • 66 3.2.3 Physical Monitoring Sampling Protocol. 70 3.2.3.1 Chemical Sampling . . . • . . 70 a. Air . • . . . . . . . . . 70 (1) Equipment and Procedures 70 (2) Air Sampling Sites . . 72 (3) Air Sampling Intervals 77 b. Water . . . . . • 79 3.2.4 Biological Monitoring Protocol 86 3.2.4.1 Bioassay Methodologies. 86 3.2.4.2 Birds . • . . 88 3.2.4.3 Vegetation . . 89 3.2.5 Analytical Procedures. 89 3.2.5.1 Pre-Departure Tasks 89 3.2.5.2 Pacer HO Analvtical L:boi.·: ,- 91 3.2.5.3 PacerHOLaborato,-vM::,,.,gem, ,r. _;ndClperation. 93 3.2.6 Spill Prevention and Cou,1termeas•-1rc, Plan 105 3.2.7 Safety Plan . . . . . . . . . . . . 106 4. ENVIRONMENTAL IMPACTS OF LAND-BASI-.D 01'1:.RATIONS 108 4.1 HO Impacts on Natural Environment of Johnston Island 108 4. 1. 1 Land . 108 4.1.2 Air . . 108 4.1.2.1 Weather Station (CM). 108 4.1.2.2 Wharf . . • • . 110 4.1.2.3 Downwind Station • . . 111 4.1.2.4 Dedrum Facility . . . 115 4.1.2.5 Tomato Plant Bioassay 115 4.1.2"6 TCDD . • • • • • • . • 118 4. 1. 3 Water • . 119 4.1.3.1 Dedrurn Area Offshore (Site WD). 119 4.1.3.2 Wharf (Site WF) . • • . . . • . 125 4.1.3.3 Saltwater Intake (Site WS) • . • 127 4.1.3.4 Potable Water (Site Pl [Composite] or P2 [Grab]) • • • •.•.. 128 4.1.3.5 Sewage System (Site SEl [Composite] or SE2 [Grab]). • • . . . • 129 4.1.3.6 Wastewater Outfall (Site WO). 131 4,1.3. 7 TCDD • • • • , •• 134 ii I \ C a » • e a e W .( •

TABLE OF CONTENTS (Continued) 4.1. 4 Biota . . ...... 134 134 4.1.4.1 Vegetation. 137 4.1.4.2 Birds . . 138 4.2 Human Environment ..... 138 4.2.1 Industrial Hygiene. 141 4.2.1.1 Air Monitoring. 4.2.1.2 Accidents or Injuries. 143 144 4.2.1.3 Miscellaneous . . . . 145 4.3 Non-HO Impacts . . . . . • . . . . • . 145 4.3.1 Temperature and Dissolved Oxygen. 145 4.3.2 Operational Impacts . . . . . . . • 4.3.3 Land-Based Incineration of Wood Dunnage. 148 4.3.4 Social and Economic Impact •.•.. 150 5. RELATIONSHIP BETWEEN LOCAL SHORT-TERM USES OF JOHNSTON 151 ISLAND TO LO~G-TERM IMPACTS. 151 5.1 Air 151 5.2 Water 152 5.3 Land. 152 5.4 Biota 152 5.5 Summary 154 6. MEASURES TO MITIGATE ENVIRONMENTAL IMPACTS 154 6.1 Drum Yard . . . 1-4 6.2 Dedrumming Facility 154 6.3 Change Building 154 6.4 HO Transfer • . . 155 6.5 Cleanup . . . . . • 155 6.6 Site Reclamation. 7. UNAVOIDABLE ADVERSE ENVIRONMENTAL EFFECTS. 156 156 7.1 Air. 156 7.2 Water 156 7.3 Land 8. SUMMARY AND CONCLUSIONS. 157 General Summary . 157 l. 8.1 Weather Observations. 157 1' 8.2 8.3 Ambient Air Observations. . 157 Water Quality Observations. 158 8.4 8.4.1 Saltwater Intake .. 158 158 8.4.2 Potable Water. 8.4.3 Sewage Outfall ... 159 8.4.4 Raw Sewage ... 159 159 8.4.5 Oedrum 8.4.6 Wharf. .. .. 159 i iii r - ,' --:. - ... ~ ............ __ C id $SW$ U

TABLE OF CONTENTS (Continued) 8.5 Biological Observations . . . 160 8.5.1 Tomato Plant Bioassay. 160 8.5.2 Vegetation Survey . . . 160 8.5.3 Bird Survey . . . . . . 161 L.. 8.6 Industrial Hygiene Observations. 161 LITERATURE CITED . . . . . . . . . . . . 162 iv

LIST OF TABLES Page 1. Acute Toxicity Data for 2,4-D, 2,4,5-T and Derivative Acids, Salts, Esters, Amines, and Ethers • • . • • • • • . • • • . • 29 2. Sub-Lethal Effects of 2,4-D Derivatives Upon Aquatic Animals 30 3. Water Quality Criteria/Standards--Drinking Water 36 4. Operation Pacer HO Data SUJlllllary-Water Pre-Operational 41 5, Response Factors and Detection Limits for Methyl and Butyl Esters of 2,4-D and 2,4,5-T . • • • • • • . . • • . . • 95 ) 6. Summary of All Concentrations at Ambient Air Monitoring Sites 109 7. Concentration Averages at Downwind Station 114 8. Concentration Averages at Dedrum Facility. 115 9. Summary of Data Obtained from Tomato Plant Bioassay 117 10. Water Data Summary - Operational ..... 120 11. Water Data Summary - Interim .... 121 12. Water Data Suunnary - Post Operational 122 13. Plant Species Observ~d on Johnston Atoll 136 14. Bird Species Observed on Johnston Atoll .. 139 15. Summary Table of Air Concentrations Personnel Samplers 142 16. Statistical Summary of Non-HO Water Data 146 17. Percent Dissolved Oxygen Saturation for Mean Temperature 147 18. U.S. Air Force Hi-Volume Sampler Results of HO Dunnage Burn, September 10, 1977 • • • • • 149 I i. V .i . " .. ·:¢.:..,..,.~:ti,

--------------------------------------- ---- ~ - - LIST OF FIGURES 1. Map of Johnston Atoll; Adapted from Amerson (1973). 6 2. Weather Observations at NOAA Station 1977 9 3. Yeatures of Johnston Island 17 4. Inferred Lagoon Circulation (Westerly Flow), July-August, 1965, Kopensky and Wennekens, 1966. . • • . . . . • . . . 23 5. Inferred Lagoon Circulation (Easterly Flow) July-August, 1965, Kopensky and Wennekins, 1966. • • . • • • • . • • . . • 24 6. Microscale Current Directions at the Main Wharf, July, 1977 26 7. Water and Sediment Sites •. 40 8. Breeding Cycles of Seabirds at Johnston Atoll. 47 9. Monthly Cumulative Bird Populations, Johnston Atoll, 1963-1969. 48 10. Monthly Mean Shorebird Populations for Johnston Atol.1 , 1963-1969. 50 11. Areas Used by Sooty Terns and Wedge-Tailed Sheandter on Sand Island, Johnston Atoll, 1965. • . • . . . . . . . . . . • . • 51 12. Nesting Areas of Ground Nesting Birds on the Original Portion of Sand Island, Johnston Atoll, 1963. . . . . • • • . • • • • . 52 13. Nesting Areas of Birds Which Nt•rmally Nest in Low Vegetation on the Original Portion of Sand Island, Johnston Atoll, 1963. • 52 14. Statistical Distribution of Second Drum Rinse Mass. . • . • 59 15. Drum Rinse - Quality Control Chart, Running Averages (All Data) 61 16. Drum Rinse - Quality Control Chart, Running Averages. 63 17. Air Sampling Sites •• 73 18. Air Monitoring Sites, Drumyard •• 74 19. Sampling Sites at Dedrum Facility. 75 20. Water and Sediment Sites . • • • • • 80 vi i l

LIST OF FIGURES (Continued) 21. Saltwater Intake Sampling Points . • 81 22. Location of Sewage Outfall Sampling Site •• 82 23. Water Monitoring Site, Drumyard. 83 24. Biota Monitoring Sites. 87 25. Check for Linearity • • • • 94 26. G C Chromatograms of (A)Chromosorb Sample CD 29Y707J and (B) Water Sample WS 30Y700J. • • • • • • • • • • • • 99 27. Concentrations Measured at Downwind Station 112 28. Chromological Profile of Water Concentrations Contrasted with the Dedrwmning Activities. • . . . . . . . . . . . . . . 123 29. Sewage System Discharges of 2,4-D and 2,4,5-T to the Offshore Area South of Johnston Island. • 130 30. The Distribution of Six Seabird Species. 140 vii

LIST OF PLATES Plate 1. Aerial view of Johnston Island, showing inc·omplete reef • • • • • • • • • • • . 165 Plate 2. View of Empty Drumyard showing observation well 165 Plate 3. Two views of drum crusher. Left hand view shows standing water immediately after shower, and stain from residual diesel fuel expelled from drums •• 166 Plate 4. R-5 Refueler. Notice HO identification on tank 167 Plate 5. Diked area. Spills of several hundred gallons could be contained within the plastic lined area 167 Plate 6. Chromosorb tube, aluminum foil wrap, and storage tube. Pencil added for scale. Chromosorb material in left 1/3 of tube • • • • • • • • • • • • • • . ••• 168 Plate 7. Benzene impinger setup. Benzene flasks are wrapped in aluminum foil to prevent photo-decomposition of trapped species • . • • • • • • 169 Plate 8. Water sampling location off ship bow. Another site was located approximately 10 meters off the bow 170 Plate 9. Sediment sampling. Scuba gear required because of depth. Observation of divers led to understanding of subsurface currents 171 Plates 10, 11, & 12. Water and Chromosorb preparation laboratory . • • • • 172 Plate 13. Safety equipment. Taken in dedrum facility, showing respirator, face shield, gloves and apron • • • • • 173 Plate 14. Wharf air sample site. Wind direction, quartering bow, evident from flags 173 Plate 15. Air monitoring site at dedrum. Two views showing equipment on barrel. Lower view shows second site at far corner, on barrel. 174 Plate 16. Normal tomato plant •• 175 Plate 17. Tomato plant with level 2 injury (tip curling) •• 175 Plate 18. Tomato plant with level 3 injury (stem curling) • 176 Plate 19. Tomato plant with level 4 injury (severe curling) 176 vif i J i •

LIST OF PLATES (Continued) Plate 20. Native flora downwind of drumyard. Tomato plant location in background • • • 177 Plate 21. Epinastic growth in native flora (observed prior to Pacer HO Project) 177 Plate 22. Incinerator for wood dunnage 178 Plate 23. Typical operation of incinerator 178 ix . ,. " ~j'~ • • .. , : . !.~

DETAILED ENVIRONMENTAL ANALYSIS OF PROJECT PACER HO 1. INTRODUCTION This report is Part II of a three-part report on the environmental consequences of a project conducted on Johnston Island, labeled Project Pacer HO, designed to remove and incinerate the stocks of Herbicide Orange (HO) stored on Johnston Island since 1972. The three parts to the report are as follows: Part I Executive Summary Part II Detailed Environmental Analysis Part III Supporting Raw Data 1.1 Background In April, 1970, the Secretaries of Agriculture, HEW, and Interior * As a result jointly announced the suspension of certain uses of 2,4,5-T. of this announcement, the Department of Defense suspended the use of Orange Herbicide since this herbicide consists of approximately 50 percent 2,4,5-T and 50 percent 2,4-D.* This suspension left the Air Force with 1.5 million gallons of Orange Herbicide in Vietnam and 0.8 million gallons in Gulfport, Mississippi. In September, 1971, the Department of Defense directed that the Orange Herbicide in Vietnam Le returned to the United States and that the entire 2.3 million gallons be disposed in an ecologically safe and efficient manner. The 1.5 million gallons were moved from Vietnam to Johnston Island for storage in April, 1972. The cost of maintaining the storage areas, and the ever present danger from the stored HO stocks, led the Air Force to conduct a study to develop mechanisms for the ecologically safe, efficient, and, if possible, low cost disposal of the approximately 2.3 million gallons of HO. After several proposals and draft Environmental Impact Statements, the ultimately accepted course of action was disposal by incineration aboard a specially * 2,4,5-T is 2,4-T-trichlorophenoxyacetic acid, while 2,4-D is 2,4-dichloro- phenoxyacetic acid. Both are commercial brand leaf herbicides. 1 l . ,. • ~ / ~ Li D L $ a as .,-,.,,r•-•.,.•-••-.w-~--------..,.,.,--~

designed incinerator vessel in an isolated location of the Pacific Ocean. The proposed incineration site met the criteria proposed in the Air Force 6 document, (l ) "Final Environmental Impact Statement on the disposition of Orange Herbicide by incineration". 1.2 Need for Field Operations As a part of their final EIS, the Air Force stated, "a monitoring program will be conducted to document herbicide exposures and environmental exposures should they occur. It is anticipated that this program will generate sufficient data to demonstrate the personnel and environmental safety of this operation". Air Force policy was that an independent con- tractor would perform the monitoring program. Thus, Battelle was ultimately selected by the Air Force to conduct the monitoring program for activities on Johnston Island. The ship board monitoring was conducted by TRW under contract with the U.S.A.F. 1.3 Application of NEPA The Air Force complied fully with the tenets of the National Environmental Policy Act through their submission of a well considered and complete EIS. It was decided that the monitoring program results would be presented in a format commonly used to prepare EIS's. 2 ... $ * cs ~•lfiWllr"-·,.•---..-,....--------••-'_;;/~

2. EXISTING ENVIRONMENTAL FEATURES OF JOHNSTON ISLAND The physical and biological features of Johnston Atoll and surrounding waters have been well studied and documented. The ecological baseline descriptions presented in this report are based primarily on accounts published by government agencies or by scientists under government contract. The two major sources of information are "Ecological Baseline Survey of Johnston Atoll, Central Pacific Ocean" by A. Binion Amerson, Jr. ( l) and the "Natural History of Johnston Atoll, Central Pacific Ocean" by 2 A Binion Amerson, Jr., and Philip C. Shilton( ). Both of these documents were prepared by the Smithsonian Institution, Washington, D.C. Tabular material and figures included in this section have been taken from the 2 report prepared by Amerson and Shilton( ). An aerial photo of Johnston Island is presented in Plate 1. 2.1 Physical 2.1.1 Land 2.1.1.1 Location Johnston Atoll is located between the latitudes of 16° 40' 26" and 16° 47' 25" North and longitudes of 169° 24' 15" and 169° 33' 58" West. It is one of the most isolated atolls in the Pacific. The nearest land mass to Johnston Atoll is the French Frigate Shoal in the northwestern Hawaiian Islands, approximately 450 nautical miles (nm) to the north-northeast. Honolulu, Hawaii is 717 nm to the northeast, Kingman Reef of the Line . Islands is about 850 nm to tne southeast, Howland Island is 1,050 nm to the south-southwest, and the Marshall Islands lie almost 1,200 nm southwest of Johnston Atoll. 3 :~ ' ,.. ,·; .. •=-c·· ·,.,.-r.· , • • ~.•~~--•Eats • • JS,......,. OU O .. •

2.1.1.2 Topography Johnston Atoll consists of four islands within a shallow lagoon partially enclosed by a semicircular reef to the north and west. Two of the islands are entirely man-made from dredged coral. These are Akau (North) Island at 16° 45' 52" N x 169° 31' 03" Wand Hikina (East) Island at 16° 45' 26" N x 169° 29' 19" W, having land areas of 24 and 17 acres, respectively. The remaining two islands are highly modified natural islands, having been increased significantly from their original sizes. These are Johnston Island at 16° 45' N x 169° 32' Wand Sand Island at 16° 45' N X 169° 30' w. The smaller Sand Island (about 1,900 yards northeast of Johnston Island) was originally 10 acres in size with a maximum elevation of 15 feet above sea level. It has since been modified to include an area of fill of several acres about 500 yards west of the original island, and a cause- way was constructed to join the two. The entire land mass (fill area, causeway, and original island) has been designated "Sand Island". The only structures present on the original portion of Sand Island are the Loran-C transmitter building and the 625-foot transmitter tower. A few concrete foundations from buildings removed in the late 1950's and some gun emplacements still remain. Generally, the surface composition of the original island is a loose coral sand. The largest island of the atoll, Johnston Island, was originally 46 acres with a maximum elevation of 48 feet. Manipulations made in 1939-1942, 1951-1952, and 1963-1964 enlarged the island to 570 acres using dredged coral from the lagoon, and leveled it to an average elevation of about 7 feet. The island is presently rectangular in shape with a 9,000 foot runway running in the southwest-northeast direction, almost along the island's main axis. The surface of Johnston, Akau, and Hikina Islands, and the man- made portion of Sand Island are characterized by buildings, roads, and bunkers. Due to the packed, crushed ccral surface composition of tlw:-w 4 ., • w W 3L C a a u o •

islands, vegetation is sparce. Only a few small lawns, scattered bushes and trees, and thinly scattered weed species exist. Figure 1 presents a schematic of the Islands and R~ef of Johnston Atoll. I I... 2. 1. 1. 3 Geology Johnston Atoll and its islands are situated atop a seamount of the mid-ocean Hawaiian Ridge. The surface lithology has been eradicated for the most part by construction activities on the island. The visible surface of Johnston Island is largely composed of dredged coral from the adjacent lagoon area. There is evidence of sea terraces that exist near the current mess facilities on the island more or less parallel with the main runway. Such terraces, step and grade towards the south tend to indicate that the basement seamount rim has undergone an uplift orogeny. Beachrock remnants are found on the original island's northwest and south central portions. The composition of this beachrock is primarily coral, fine sands and gravels that have been cemented together by calcium carbonates. Pumice rock was found erratically along a small section of the southeast shore of Johnston Island. It has been cited in the literat~r.,( 3 ) that the outer reefs to the south of Johnston Island are submerged as a result of the tilting of the seamount basement structure towards a strike to the southeast. Due to the volcanic origin of the seamount that supports Johnston Atoll and to the evidence of unequal thrusting and settling Johnston Atoll is not considered to be a geologically stable land form. The literature is deficient in describing the form and substance of the supporting seamount; There are apparently no exploratory deep wells on Johnston Island. There is evidence that the outer reef which breaks the surface of the sea only on the northern shore is undergoing differential settling or thrusting. 5

........ 0 .... < z z .::------,._ 0 ~ u, z :I: / 0 -, ,• ... ,,,.,/ :"...- .-· ..• ; i . -· ,,,,, ,' ' ' ' . \ \, •, . ........ .. . '. .' . •••➔ \ .' ' \ 6 • a:+ D I Z O 6 w ' i,iiirGiiiP 0. •

4 Kroenke and Wallord( ) surveyed Johnston Island gravity utilizing the Bouguer Anomaly effects. These studies suggest that the mass densities beneath Johnston Island are intermediate in value as compared with the Hawaiian Islands and Line Islands. These studies found no evidence of dense magmatic structure existing in the upper structure of Johnston Atoll. The physical geology underneath the Orange Herbicide drum storage area contains alternating layers of coral and beach sands which have been artificially deposited and compacted. The permeability rates would be expected to be high in this unconsolidated dredge fill. 2. 1. 1.4 Soils The soils occurring on Johnston, Akau, and Hikina Islands and the man-made portion of Sand Island consist of compacted coral fragments ranging in size from sand to cobble sized coral rock. These fragments were derived from dredging operations in the deepening and lengthening of ship channels and seaplane landing areas. The entire islands of Akau and Hikina and the man-made portions of Johnston and Sand Islands were constructed from this material. The soil occurring on the original portion of Sand Island is deep, loose, coral sand. This surface is quite similar to that of Johnston and Sand Islands prior to their disturbance by military construction. 2.1. 2 Air 2.1.2.1 Meteorology During the Interval Meteorology data were recorded at the NOAA weather station located on the eastern end of the island. An additional anemometer with strip chart recorder was maintained near the drum storage area (for the period July 20 to August 27, 1977) which recorded additional wind data for the western end of the island. 7 f ·, . ... ,~ • 4 • k,~.> -~.flli...••·11.jiii•lilhllct•Ziiiiiiiiii•..••••1:1•2ia11alllli11s•rt»"iii•li-rt•..u..u,,.•""!s""''-••-.wr--------...,..,i.;,,,~/:JJ

The meteorological records for wind speed, direction, temperature, dewpoint, and rainfall are presented in Figure 2, Superimposed on these data, collected by the NOAA station are the wind speed and direction at the west end anemometer for several sample weeks. These data are discussed further below. In these discussions, the recorded values are compared to norms which were assembled from 30+ years of data and presented in Amersonfl) a. Wind Speed and Direction Surface trade winds were essentially constant throughout the period. Winds were from the east-northeast to the east-southeast at from 10 to 20 m.p.h. on most days. The exceptions occurred on August 8 and 9, 1977, and again over the interval August 14 to 16, 1977, when winds were at Oto 10 m.p.h. from the northeast. Only one directional shift of significance occurred during the period. On August 10, winds were out of the south at about 10 m.p.h. Minimum variation from seasonal norms was experienced over the duration. A comparison of the data taken at the two wind recording stations indicated only a negligible difference. Wind directions were very slightly more northernly at the drum storage station. Also, wind speeds were a few m.p.h. less at this station, attributable to the drag effect of the entire length of the island. b. Temperature As a result of air masses ·passing over the atoll having been conditioned by close contact with the ocean for thousands of miles, there is little daily variation in air temperature. Similarly, only very small seasonal differences exist (about 3° F), with August being the warmest month of the year. Throughout the period observed, daily highs ranged from 83° F to 85° F. Lows were usually between_ 77° F and 80° F, with a daily mean of 81° F, which is normal for this time of year. 8

.08 f/l Q) .06 ..c: CJ c:: H .04 .. - - - ---- ....., ~ ....., ·--- - - - - - - - ca ...... .02 c:: •.-l ca ix: 0 20 21 22 23 . §~~·~l g 90 ) c:: •.-l 80 0 p.. i:,.. 70 :3: 0 Q) 60 0 20 21 22 23 . Q) .... 90 t:ti¥1 ::, .... ca i:,.. 80 •.-l .... < Q)0. 0 70 e Q) 60 E--< 20 21 22 23 40 I ., . - .... . -•---- "C .... 30 ·- --•. Q) Q) ::, 0 LA - 0. ::x: 20 ,, - Cf.I -f/l - "C c:: .-4 Q) 10 •.-l •.-l ::;i: :E: 20 21 22 23 Note: The dark line is the NOAA weather station data, while the light line is com- parative information from the downwind weather monitor. One day each week is compared. July J,, FIGURE 2. WEATHER OBSERVATIONS AT NOAA STATION, 1977 9 i st Ca» _.u., •• a

f' I .08 ------ Cl.I .06 ------- QI .c CJ ....a .04 ---- ......... ....i::<IS .02 .... ~ 0 I - - I 24 25 26 27 28 29 30 24 25 26 27 28 29 30 90 .'"' QI Ft5Ffjjjf:SI ::, 80 1-1 <IS"" .... 1-1 < QIp. 0 70 ffi 60 !-a 24 25 26 27 28 29 30 40 ~ "0 1-1 QI :, 30 QI 0 '2.::i:: en._ 20 Cl.I Note, The dark line "0 a..., QI 10 is the NOM weather ........ 3 :c ~ ~ t a t i o n data, while 24 25 26 27 28 29 30 the light line is comparative infor- mation from the down· t.,[ f ~~;b±fd=~:;:.~ach -•k C 300 - r - - - - - r - - - - r - - - - . - - - - - . - - - - - - - - - - w i n d weather monitor. 0 Cl.I "O .... QI C ., QI 200 .... CJ 1-1 is 3 QI 00 100 '"' -"4AQI A 0 24 25 ~~ 27 28 29 30 July FlGUilli 2 . (Continued) 10 ....., .

. 08 ---- '' Cl) , J <lJ .06 I ,.c: u i:: H .04 ..... ...... ,.....C1l .....i:: .02 C1l i:,::: --.---· I 0 31 1 2 3 4 .µ 90 i:: ..... 80 0 P-. ~ 70 ~ 0 <lJ Cl 60 31 1 2 3 4 s 6 ,.. 90 f5Efrtfi?l (lJ ::l 80 .,.,.. ,.. .µ C1l~ 70 < <lJC. 0 s 60 ....<lJ 31 1 2 3 4 s 6 40 Note: The dark line "1:l,.. ;l is the NOAA weather (lJ (JJ 0 30 station data, while 0. ;i:: (/) "1:l ---rn Qj 20 the light line is comparative infor- i:: ...... 10 ~ •H mation from the :3 ,:: downwind weather 31 1 2 3 4 5 6 monitor. One day each week is com- pared. i:: 300 .,.0 rn Qj 200 "1:l .µ i:: u .,. Qj ,.. Qj 100 Oil ::J: ,.. Qj -H Cl 0 0 31 2 3 4 s 6 August FIGURE 2. (Continued)

.08 ·- - ---- I I ti) I 1 <1J .06 .c: u C H ..... .04 ..... ....<IIC -- -- - -~----- •,-f <II pr:: .02 0 ·-, I ---- ----- .. I I I n. . I - 7 ' 8 '9 10 ' ' 11 12 13 .., C .,-f 0 p.. ~ ;3: 0 <1J 0 7 40 . . . - - - - . . . - - - - - - - - - - - - - - ~ - - - - - - - N o t e : The dark line _ _ __,.____ _--1~--+---~---1----~~--~is the NOAA weather station data, while the light line is -+----~--IIL--+----+-----+----~---~comparative infor- mation from the down- ..,__ _--+___.,__ _,...._ _-+-----+---.,__ _.... wind weather monitor. 7 8 9 10 11 12 13 One day each week is compared. ~i l~~~ ffi?f:tt~t:I C 300 ....------r------.------....---..---...---.. . 7 8 9 10 11 12 13 August FIGURE 2. (Continued) 12 ,,,,__~ -;·; {\!, --~-►4 ~.

.08 -~---- •>-· •1-- -·· ------- --- .,_ ___ - - - -- - - --·-- ------- >---- ~+--· rJl ll> ..s::: .06 . - I u i:: ...... ...... ...... .04 =~J •!'----------- I ---- -- " <.... i:: ..... .02 I ~ " 0 l·- I Ii • I I I I I . I I 14 15 16 17 18 19 20 ~rt .... 90 F t' j "F=I .....i:: 80 0 ~ µ., .-\•11111 ;:3 0 70 0 ll> 60 14 15 16 17 I 18 19 20 90 - r - - - - r - - - - . - - - - - - , - - - - - - - - - , - - - - - , - - - - - - - - , :~~I 14 15 16 17 18 19 20 40 ----""""T---......--~---.,..---....----------.Note: The dark line is the NOAA weather + - - - - - + - - - - - - 1 - - - -...... station data, while the light line is #1~~""""~~~~~~~~~1 comparative infor- mation from the down- -+----.---+----+---.....---+------.1----~wind weather monitor. 14 15 16 17 18 19 20 One day each wet,k is compared. 14 15 16 17 18 19 20 August FIGURE 2 . (Continued) 13

-1 .08 (I) ll .06 .c '.., .....C .04 --- ..., ..... · ~ - - -- ---·- --~- - (13 ..... ....C: .02 (13 tz I 0 I II ... ,__. n -- - I .. I 21 22 23 I I I I 24 I I 25 . 26 I I 27 22 23 24 25 26 27 90 ~~tfi4:tt7 r1 (1) I,.. ..., ::, .... <13 I,.. I,.. i;._. <: Q) 0 p.. s(1) E--< 21 22 23 24 25 26 27 40 Note: The dark line "Cl (1) I,.. ::, 30 is the NOAA weather (1) 0 p.. ::c: 20 station data, while <Jl .._ rJ) the light line is "Cl (1) 10 comparative infor- i::: ,-l .... ..-1 mation from the down- ;:!:~ wind weather monitor. 21 22 23 24 25 26 27 One day each week is compared. 21 22 23 ) .;_' ' 25 26 27 Auhust FIGURE 2 . (Continued) 14

A very slight warming trend (1-2° F) was observed from the beginning of operations in late July through the end in late August. This was to be expected because the monthly mean for July is about a degree Fahrenheit less than that for August. c. Precipitation Rain is extremely variable on the island in both frequency and intensity. The accumulated measurable rainfall was 1.3 inches during the 39-day period. In addition to measurable quantities, trace amounts were observed over hourly intervals on 58 occasions. The heaviest rains (more than 0.10 inches/hr) occurred on July 29, August 4, August 5, and August 16. Rain was most frequent over the intervals August 3 through 5 and August 23 through 26. In comparison to the norm for the season, the period was a rather dry one, with rainfall at about 55 percent of the total expected. However, the rainfallwaswell within the observed extremes of 0.4 to 10 inches for the total period. d. Dew Point Dew point temperatures ranged from 70° F to 77° F throughout the period. Highest readings (75° F) were recorded during periods of rain. On no occasion, however, was the dew point ever reached. 2.1.2.2 Air Quality Being remote from other terrestrial environments, the air at Johnston Atoll is clean, with none of the pollutants normally associated with urban areas. The only air contaminants expected at Johnston Island are those introduced at Johnston Island itself. Routine insecticide spraying was suspended during the HO operations on Johnston Island. I I. I 15 ·- .... ., . . .... pe1~'- ►4. •i9'Qa. . j,.,i,.-M ... / ,,/ ~ t ca: a Uil • as s~---•-••--------....16,~-.!:. 1

The fol lowing sect ions discuss the applicable air standards• existing sources of HO atmospheric, and observed atmospheric HO con- centrations prior to the dedrumming operations. a. OSHA Standards Christiansen(S) discusses the toxicity of 2,4-D and 2,4,5-T and its n-butyl esters. No inhalation toxicities are reported for any species. The Occupational Safety and Health Administration has estab- lished 8-hour time weighted average concentration occupational standards for the acids of 2,4,-D and 2,4,5-T. For both chemicals the standard is 3 10 milligrams per cubic meter (10,000 µg/m ). The constituents of Orange Herbicide, however, are then-butyl esters of the acids. There are no OSHA (or any) standards for exposure to the esters. However, the reported animal toxicities in Christiansen(S) for the butyl esters are even lower than for the acids. It is reasonable 3 to assume that 10 mg/m is a realistic human TWA exposure limit for humans. I b~ Existing Pollution Source The herbicide was stored in a drum storage yard at the northwest corner of the island as illustrated in Figure 3. At this location, the prevailing winds rapidly removed any atmospheric HO away from Johnston Island and the atoll and dispersed it in the open Pacific. There were no other locations containing HO. Prior to the disposal operation, the salty environment caused drums to corrode and thus leak. A team of men patrolled the drumyard looking for fresh HO sorbed on the ground, an indication of a leaking drum. While an exact measurement was not made, an estimate of from 20 to 70 wou\d be found leaking each week. ·i The leakers were taken to the dedrumming faciU ty where they were allowed to drain into a covered collection sump over a period of days. On a weekly basis, the collected drainage would be redrummed in new drtims and res tacked, while the old drums would be crushed and stacked. 16

t- • · ~ ; • ,,, . -· T, 'f I . -~,, Q------ _________ _ -- -------- /Iv , --'-=- ....:::::::::=:~- --- --------------- J11191':------ - ' -~=l~H~IF~::: -- -=~:::~-~::_-~~~ ..,. . -- .. - - Intake ~~.:::,~::== ·---- . Temporary Stora-- ~----- ---- ----- . . . . :..• jlJ !Sf~;; r:;.~, j ( ·~1~~: .,~-... ~~;F'..·,, .... ~- ,,, " ,~:,_~~ ,,~ -, ,, . . ,,-"''' $'~,,~-~~,~~-*~~,~~~~l~'t • _, -- -- ____ _ ~--=------~-::::,;~~- >' i>&'\~;:, >'>Y·,is> »)a)' --- o~---- --::::;::- 0 Water Sampling Sites ~ Johnston Island Crew Facilities 0 Sediment Sampling Sites ~ Wharf 11 Analytical Buildings (Bldgs 120/190) ffl Dedrum Area II Red Hat Area FIGURE 3. FEATURES OF JOHNSTON ISLAND -~-·- ~""

There is no measurement of the volume actually leaked. The incineration records show that the average drum contained 53.9 gallons, but it cannot be said that all drums were initially full. c. Observed Ambient Air Pollution While concentration measurements downwind of the site were not made prior to the HO operation, the values for 2,4-D and 2,4,5-T in the pre- 3 operational period averaged 0.49 and 0.08 ~g/m , respectively, at the downwind station. Furthermore, the odor of the trichlorophenols in the HO was intense across the entire downwind boundary of the drumyard. The consistent, strong winds at Johnston Island are helpful in the removal and dispersion of HO from the atoll. It is expected that the atmospheric stability is typically Class B* during the day and Class D at night. With these stabilities, dispersion processes should reduce concentrations by a factor of 10 within 1.6 kilometers downwind (day) and 4.4 kilometers (night). 2.1.3 Water Environment The existing water environment of Johnston Island consists of several components of the hydrologic cycle. Because of the small size of the islnnd, cycling of material between the hydraulic components is expected to be rapid. The hydrologic components described below include the saltwater and freshwau•r portions of the cycle. The saltwater cycle is comprised of the lagoon circulation and the groundwater underlying the island while the freshwater cycle includes the rainfall and the drinking water and sanitary system. * Turner's stability classes. 18 • C .. • C

2.1.3.1 Hydrology of Johnston Island Precipitation in excess of 0.01 inches occurs on the average of 162 days per year. The mean annual rainfall is 26.11 inches, however, variation from year-to-year is considerable. Monthly rainfall variations are small. During the period 1931-1972, July rainfall averaged about 1.6 inches while August rainfall was about 2.2 inches. In the Central Pacific tropical climate, evaporation is much greater than precipitation. This, together with the flat topography and permeability of the soils minimizes sheet runoff. Storm drainage is collected in a system of French drains, inlets, and open ditches which flow into the lagoon. Since most rains are very light, flow in these ditches is intermittent with evaporation being the predominant removal orocess. Tranpiration from plant surfaces is a very minor part of the hydrologic cycle of the island because of sparce vegetation due to the large areas of paved or otherwise impervious surfaces Bind base coral. Th_re are no permanent freshwater bodies on Johnston Island. The lack of surface water is due to the coarse texture and extreme permeability characteristic of the surface coral sands (Thorp( 6 )). Other factors contributing to the lack of significant amounts of fresh ground- water are the small land area, narrowness of the island and the high permeability which allows rapid mixing between the lagoon water and the percolating rainwater. Johnston Island's water system uses both fresh and salt water. Raw sea water is pumped from the lagoon through a traveling screen to the Salt Water Pump House. From there it is pumped to the Distillation Plant and also into the salt water distribution system where it is used for sanitary purposes, fire protection, air conditioning condenser units, and power plant waste heat dissipation. The Distillation Plant houses twelve distillation units and related equipment; the Freshwater Treatment Plant consists of a pump station, soda-ash treatment area, and a chlorination room and storage facilities for approximately 740,000 gallons (Figure 3 ). The freshwater system is designed to support a population of approximately 19 i C • » il/-W 0 .. •

4,500. Its total rated capacity is 318,000 gallons per day (gpd), but, allowing for maintenance and miscellaneous downtime, about 240,000 gpd can be expected at peak production. Johnston Island has insufficient relief to permit use of a gravity sewage collection system; therefore, a forced system employing pumps and lift stations is used. The force main is a series of 3" to 16" cast iron and asbestos cement pipes in parallel runs along the north and south shores with connecting laterals. Raw effluent is discharged on the ocean bottom at a depth of 25.6 feet through a 10 inch pressure outfall pipeline which extends approximately 550 feet out from the southwest peninsula of the Island. 2.1.3.2 Oceanography-Currents and Tides Johnston Island is approximately in the center of the North Equatorial Current which extends to the north and to the south of the island for several hundred miles. The velocity of this current is relatively constant from east to west at about 1/2 knot (0.41-0.63 mph; 0.61-0.82 ft/sec; 0.17-0.25 meters/sec). The underwater platform on which Johnston Island is located is similar to those associated with many Pacific atolls. Like most other low islands in the Pacific, the main outer reef has a typical cross section, which includes surge channels, an algal ridge, and a reef flat, with coral heads rising abruptly in the deeper waters to the south and east of the main reefs. The outer slope is quite steep, between 16 and 100 fathoms, usually less than one-half mile in linear distance, with an average slope of 19°. The platform on which Johnston Atoll rests stops fairly abruptly at about the 16 fathom line at most points around the circumference of the (7,8,9) ato 1 1 as the bottom begins to slope steeply down. The shallow lagoon area and its bordering reefs together form roughly the northwestern quarter of the triangular-shaped platform on which the atoll rests. At the deeper eastern end of the platform the submerged contours suggest the outline of earlier peripheral reefs. The 20 ,I_, • --,- ...... ....._._..._ J • as a@<w • • • .- 2 •'

main difference between Johnston Atoll and other Pacific islands is the lack of continuous reef around the atoll. The main outer reef extends around less than one-fourth of the circumference of the platform. In addition, there is an extensive zone of shallows to the south of the main reef which is also an unusual feature. The tidal range at Johnston Island, in common with other mid- Pacific islands, is relatively small and the effects of the tides upon the atoll are correspondingly minor. The absolute tidal range during the year (the difference between the lowest and highest tides of the year) is only 3.4 feet. The lowest low is minus 0.5 foot in June, while the highest high is plus 2.9 feet in June and July. The mean spring high tides are plus 2.2 feet while the mean spring low tides are minus 0.2 foot. The mean neap tides are plus 1.6 feet, while the mean neap low tides are plus 0.4 foot. ( 6 • 7 • 8) The time of the tidal crests and troughs is only slightly later than those of Honolulu, the nearest point for which a full tide table is available. High tides are 29 minutes later at Johnston Island than at Honolulu, while the low tides are 23 minutes later. The high-water interval from full tide to the change of tide is three hours and 15 minutes. Tide tables for July and August, 1977, are shown in Table III-14 * The maximum high tide during the assessment occurred from July 27 to 29 and measured plus 2.9 feet while the lowest tide was minus 0.1 feet on July 24, 28 and 31. The ocean currents around Johnston Atoll exert a major influence on the localized circulation within the lagoon because of the "open" structure of the marginal reefs. In addition, the tides have a range within the lagoon only slightly less than in the deep water because of this feature. Tidal currents within the lagoon show some variation with the season. During July and i igust, the normally strong westerly flow weakens somewhat. This allows a div~rgent flow field to be generated to the south- west of the atoll platform. This type of flow was characterized by *This notation refers to Table 14 of Level III Report. The notation will be frequently used throughout this report • . • ,_.,.,...,.ta at C a » s04w • 1

a moderate offshore current with a general set toward the west. Local tides induce clockwise rotary to semi-rotary motions in the regional and local circulation patterns. During rising tides, the predominant flow is to the north in the east and west ship channels and to the northwest in the north channel (Figure 4). The normal current speeds are about 1/2 knot. During falling tide, however, the predominant flow was to the south in the east and west channels and to the southeast at about one knot in the north channel (Figure 5). ( 3 , 3 , 9 ) These current moveme~ts are affected by the numerous patch reefs found in many places. Thd natural depths within the lagoon (except for the dredged portions) vary from a few inches to about 40 feet, because of the presence of coral heads and patch reefs. The greatest area lies between 15 and 25 feet underwater at mean sea level. The lagoon inside the main atoll is about 14,000 yards long at its axis, which runs southwest from Small Island through the center of both Sand and Johnston Islands. At its widest point, just east of Sand Island, the lagoon extends about 3,500 yards from northwest to southeast. West of Johnston Island the lagoon narrows to a few hundred yards in width before coming almost to a point at the extreme southwestern corner of the atoll. The total area of the lagoon within the reef is approximately 13 square statute miles. An exact measurement is difficult because of the need to measure the exact line of demarcation between the lagoon proper 6 7 and the extensive coral flats which form the southeastern part of the atoll. ( • ) At the extreme northeastern corner of the lagoon, south of the opening between the main reef and North Island, there is an area of deeper water in which average depths of mor~ than 40 feet have been reported, but the bottom still has many irregularities and numerous coral heads which almost broach the surface. Artificial dredging in the lagoon has left the seaplane landing area with a depth of eight feet cleared of obstructions, while the harbor and the entrance channel were originally dredged to 23 feet and have been swept to 14-1/2 feet. An approximate value for the volume of 11 3 3 water enclosed by the reef is 1.5 x 10 ft (4.3 x 109 m ). As observed 22 .. ,·.. ._ ,-·i . lJ 3 \ .:,: ~. - ..,., '"1: .: U L $ a cs .,-SU ... •

'\ .... ...,':'""~\.~-;>- • • ,,. ... ~ ,I • ~. ~- . ..... ·1-. , ' ,I j ,.. __ ( --- .... c:::> '- ~~ ;,•Y • = ---- ~ ,- __ ,,.. / ~ -' / / ~ / ' 'l -;i.-----r_ I •~,, b~ <~~~ 3 • N ,,,,,,,.- I • .\ ~~< w ~ 1 . .--- t I ( ,. , ,.__ " 't--__j 100 ,at"°"' , - _ , _ _ _ _ _ _ _ ,!J 5 ------ ~ Current Rotates and Sets Westward FIGURE 4. INFERRED LAGOON CIRCULATION (WESTERLY FLOW), JULY- AUGUST, 1965, KOPENSKY AND WENNEKENS, 1966 \ ,.'\ ...

..-.~_-, ,( T, l' ~ '/ '/2;--:;----~ ,.-,. ...... '- ~ ~ ~ ~ ___, 4 ' ~, r.,,."""1 N ·"'" {~f ~~~:~ • I \ lI ___ .)-r:P, "£at.no tnS ~-~----~-1 ,..-~ l ~ ,- -- _.. 71. Current Rotates and Sets Eastward FIGURE 5. INFERRED LAGOON CIRCULATION (EASTERLY FLOW), JULY-AUGUST, 1965, KOPENSKY AND WENNEKENS, 1966 ii~ ~ ·-

3 by Emery, ( ) these coral heads influence the movement of sediments by blocking the current causing sand to be deposited on the upcurrent side and scouring the areas between the reefs. His examination showed these areas to be about half a fathom (0.9 meters) deeper than the surrounding floor and containing coarser sediment than the adjusted areas. Mirco-scale currents at the wharf observed by the divers when taking sediment samples were a light west-to-east deep current and an east-to-west surface current at 20-25 feet (6.1-7.6 meters) seaward from the center of the wharf. Off the west end of the wharf, the deep current direction was south to north (Figure 6). These observations were made at 1100 hours on July 25. (IO) Water depths immediately off the wharf were 35 feet (10.5 meters). A trough of 45-50 foot (13.7-15.2 meters) depth was noted about 25 feet (7.6 meters) from the base of the wharf. (IO) 2.1.3.3 Water Quality Criteria/Standards Limits on aqueous concentrations of 2,4-D and 2,4,5-T are classi- fied as either criteria or standards. The word "criterion" should not be used interchangeably with, or as a synonym for, the word "standard". The word "criterion" represents a constituent concentration or level associated with a degree of environmental effect upon which scientific judgment may be based. As it is currently associated with the water environment it has come to mean a designated concentration of a constituent that when not exceeded, will protect an T organism, an organism community, or a prescribed water use or quality with an adequate degree of safety. On the other hand, a standard connotes a legal entity for a particular reach of waterway or for an effluent. A water quality standard may use a water quality criterion as a basis for regulation or enforcement, but the standard may differ from a criterion because of prevailing local natural conditions, such as naturally occurring organic acids, or because of the importance of a particular waterway, economic considerations, or the degree of safety to a particular ecosystem that may be desired. 25 j

....... 25' 50' Scale: l" = SO' ... I Deep Current Directions - - - - - Surface Current Directions FIGURE 6. MICROSCALE CURRENT DIRECTIONS AT THE MAIN WHARF, JULY, 1977 26 •• . .~ ·-. :;-• . -, . ,J.}j, ...., ,·, ; .,, . . 't,. . ' ~~ • .:•.,.,,., ·-i:•~: .•.''· i . . . -4...-:: lo~ ,.~:>◄.;~,-,-- 3 Sid Et a 2 C a • ,.,......, C O W2

Water quality criteria are not intended to offer the same degree of safety for survival and propagation at all times to all organisms within a given ecosystem. They are intended not only to protect essential and significant life in water, as well as the direct users of water, but also to protect life that is dependent on life in water for its existence, or that may consume intentionally or unintentionally any edible portion of such life. (l)) The criteria levels for domestic water supply incorporate available data for human health protection. Such values are different from the criteria levels necessary for protection of aquatic life. The interim primary drinking water regulations(l 4 ), as required by the Safe Drinking Water Act(lS), incorporate applicable domestic water supply criteria. Where pollutants are identified in both the quality criteria for domestic water supply and the Drinking Water Standards, the concentration levels are identical. Water treatment consisting of flocculation, settling, and softening may not significantly effect the removal of certain pollutants, (such as the components of Orange Herbicide). The ideal data base for aquatic life criteria application regarding Orange Herbicide would be information on a large number of tropical marine species common to the Johnston Atoll area over their entire life span and that of succeeding generations. Unfortunately, thesedaLa do not exist. Most of the available toxicity data on both acute and subacute effects are for freshwater organisms. These were obtained at temperatures below those typical of the Johnston Island environment or represent time frames of less than the organism's entire life span. Furthermore, independent environmental variables other than temperature have been found to be of importance in determining the toxicity of 2,4-D, 2,4,5-T or m_ixtures thereof. The Environmental Health Laboratory at Kelly AFB, TX conducted bioassay tests in which Orange Herbicide was mixed with water at a theoretical concentration of 200 ppm. It was found that most of the herbicide rapidly sank to the bottom of the tank. None of the test organisms showed any adverse effects after two weeks exposure; however, all of the fish died within 24 hours at a concentration of 20 ppm in a similar experiment but with continuous agitation of the water.< 16 ) Subsequent studies indicated that, in order to establish a dose/response relationship for the organism, some circulation of the water was necessary. . 27 ·,,. ; • "~'-►>~.

A second determinant of toxicity is the actual chemical form of the herbicide in water. The derivatives of 2,4-D and 2,4,5-T used in Orange Herbicide hydrolyze to the respective acids at varying rates. fur ocean water, in studies conducted by the Air Force, 90 percent of the esters were hydrolyzed within 7 days. Toxicity of the acids is decidedly lower than the corresponding esters probably because of polarity influences on uptake . ( 16) mechanisms. The many confounding effects make it difficult to apply a rational criterion which would protect all the potentially exposable organisms. Tebts by the EHL Kelly AFB~ TX on artificial sea water systems produced marked differences between the theoretical concentration due to solubility effects. (Thus, static bioassay results found in the literature which are based on theoretical added concentrations of Her- bicide may indicate a low toxicity (high side bias); the actual concen- trations of HO in solution producing acute or subacute effects would be much lower). The effect of temperature on organism response has received limited attention. Only one study was located which even stated the temperature at which the tests were conducted, This showed a strong temperature dependence, although only two temperatures, 17 :>:: '0 C, were evaluated. (1. 6 ) The philosophy of EPA in assigning crit,,ri., ha: ') ,n to employ a safety factor to protect all life stages of th<! t,,-:;t organism in waters of varying quality, as well as to protect H~s:~L1tcd nrganisms within the aquatic environment that have not been tr•s ·- d and that may be more sensitive to the test constituent. Applicat ic.; Lt.:tors have been used to provide the degree of protection require~. Safe levels for certain chlorinated hydrocarbons and Cl'rtain heavy metals were estimated by applying an 0. 01 application factor to the 96 hour Lc value for sensitive aquatic organisms. 50 A listing of available acute and subacute bioassay data is con- tained in Tables 1 and 2, In addition, McKee and Wolf presented the . f o 11 owing d"iscuss i on concerning . 2 , 4 - D• (1.7) 28 ,. •

TABLE 1. ACUTE TOXICITY DATA FOR 2,4-D, 2,4,5-T AND DERIVATIVE ACIDS, SALTS, ESTERS, AMINES, AND ETHERs(a) - ------ --- - ------- ---~----------------•-. ------------- ----------- --·----•·--- ----------~----------------------- -----·--- T_(•h_t __ C,11.•:, ,n111t1 ________ i, ·t o, 1, ,n: ~.m I• ., C 111t!l l ,11,..., 1 i llu .,. (n,J'/O_____ R<•!!{•N1sc _ <:on::1,t•nt!, _ _ _ _ _ FrfC'r-f'r1·c )-',:!h1 ,I 11J,,,,,,w )';~ 96 hr Tl.50 16 2 ,4-1> ( i1'1\} 2 .,.... p (:,·,-., )'.] , t:i11 177 96 hr noo 16 Ii C 191 96 hr TL,O 32 2 .'t-:) \ ,,:~\) Ch.di.I( 1 l,1tf1 ,I, Cb.llllll" J l" ,11 f j -.,ft ]'J (' 12, 96 hr Ti.~o 39 2,4-ll (I?!.\) 3. 4 4R hr LC:,o 14 ppm TCll 15 2 ,4-ll; 2 .,•• -,-r J.':+t lt,·,.d rdn1ww rrt ·.}11..•.1tt• t· l'St'-.'I (i.1i) Frt•<,l,;.,,ll<·r 2.B 48 hr LC)0 1~ 2 ,I,- LJ (s1,1 J J.',itl1l',"tll n,ir11101,,· I rt·~la1,,•,1t1•r 5.0 48 \11· LC'.50 1(, 2 ,/, '~-T (NBF) l-"atl,,•;1d 1·1 I 11n~1w I- I t•;-!il,,!,ll Pf 270 48 hr Leso ]6 2 ,4- ll J.,llht•,ad niin1111._., F1 v•,h·..:,,l1.· r 333 48 fu- l,C50 ]6 2,4,~-T r.,tli,·.1d n,1 nnL•W s.,1 t w.1t<: r 5.6 48 hr LC~o H, 1,li-D (NllF) Shrfni1• Sa 1 t ...1.111•r 3J 48 hr Leso 16 2 ,4 ,s-·1 U,•1-:) Shrimp ft l'~,Ji.1<1t .•r O. I 48 hr Ttso 3) 2 ,4-ll (1·r.11i:) Il.irhni a 11 ..1~•.n,1 S(_•<>d ~hrfr.ip Frc-!,h"-·,J! £•r 0.3 48 hr Tl.50 33 2 ,4-ll (!'<:!</) ScuJ Fr1•.,la ...·.it1•r 2.(, 48 hr TL5n 33 2 ,4-Il (l't,I<!:) Sowtiug Frrsliwat er 2.2 48 J,r Tl.5(1 3l 2,4-D (ICM.) 2.7 4H !tr TL5c., )3 2 ,4-ll (\'Cl\'.) GLh, ?;hr imp frc~.b·,.t:1tet· IU-31 48 hr TI,,, O'!.>t~1!1wr! ft·rm 3 31, 2 .4-1, ( 11,1;) HhH•j.,i 11 nlflOll f ;JCt 1,l ::rs 17 48hr TI,n 3', '!. '1-n (PCJ;!:) Bllwr.Jll 1.11 48 hr 11 111 31, 2 ,4-ll (BUI\·.) Uluq:Lll 0.3 48 hr "i'Lm 3'; 2 ,4-D (1•,.c:·:J Fish S:ilt~,,.att•r 41'.,-8/,Q 48 hr L<:50 :,"'i 2 ,4-ll (M:0) Bluq;f II J(,(,-4',8 48 hr i.c,o 3(, 2,4-D (l>'L\) HJucgj]l 8.8-59.7 l,R hr LC50 36 2.~ fl ( !(;E) Blu"gill c:,,u) JO 96 hr Leso 36 2 ,4-11 Fatlh••.rl 1.1innow s 9(, J,r u·so 36 2 ,4· II (M) Fathl',ld 111inno\,," Fat h1· ilil /t; l ut'F,i l l 4 Mu. LCJf\ 011 solul.lC" 36 7 ,4-n (A'') 2,4-0 ( r, \•.,:) Fatla•,1cl/hl uq~i 11 2 4 Ho. 1.r.l(, 36 Fit t 1u•:td /bl ucp,i 11 2 72 hr LC50 36 ?. ,4-1) (~<:i.1 i llluq:11 I l.~ 4B hr LC·,o 36 2, 4-lJ (Af"I) Rl,wr,i ll J. 3 1,8 hr l,CJo 36 2.4-IJ {r:u•) 1.1 .'.,8 hr 1.r~10 3(· 2 ,4- Li (1P~) LI U''l',i I I 1,8 hr Tl'm 3(, 2 ,Ii-I> (Bt>1'.E) n,;1, S.:lt.1r1,1l<'t' 5 Sa]t\'c1trr 4.5 4R hr TI •.., 35 2,4-D (1•r:1,;·J Fish FJ1<h 100 !'h1r::;lwld rnn,·. 17 2 ,4-1) for MOI t:'11 lty rerr-h 75 11ll'~••hold r01ic. II 2,4-ll for 0>t>rL1l lty 55 Thrs•e;hold cone. 17 },4,S-T P1.·n.h for \\lnt'l;o.lil y (,() 111 rt••1h" Id rf•nc. 17 2 ,t, ,5-T Bh·.1k fell "101t ,11 t Ly - - - . ·-- -·- ---·--- -- --- - . --~--- (a) (DMD)= dimethylamine; (NBE) = normal butyl ester; (PGBE) = propylene glycol butyl ether; (IOE) = isooctylester; (BOEE) = butoxyethylester; (MS)= alkanolamine salt; (AA)= acetamide; (AS)= amire salt; (IPE) = isopropyl ester. (b) See literature cited for references. 29 l • C J Q# ...,..._, C . ..

TABLE 2. SUB-LETHAL EFFECTS OF 2,4-D DERIVATIVES UPON AQUATIC ANIMALS(a) - ~-------- -- ----" Test Compound Test Organism Dose Response Butoxyethanol Oyster 3.75 ppm 50% Decrease in ester (96 hrs) shell growth Butoxyethanol Shrimp 1 ppm No effect L- ester (48 hrs) Butoxyethanol Phytoplankton 1 ppm 16% Decrease in ester COz fixation Dimethylamine Oyster 2 ppm No effect on shell (96 hrs) growth Dimethylamine Shrimp 2 ppm 10% Mortality or (48 hrs) paralysis Dimethylamine Fish (salt water) 15 ppm No effect (48 hrs) Dimethylamine Phytoplankton 1 ppm No effect on COz I ( 4 hrs) fixation Ethylhexyl ester Oyster 5 ppm 38% Decrease in (96 hrs) shell growth Ethylhexyl ester Shrimp 2 ppm 10% Mortality or (48 hrs) paralysis Ethylhexyl ester Fish (salt water) 10 ppm No effect (48 hrs) Ethylhexyl ester Phytoplankton 1 ppm 49% Decrease in ( 4 hrs) COz fixation PGBE ester Oyster 1 ppm 39% Decrease in (96 hrs) shell growth PGBE ester Shrimp 1 ppm No effect 48 hrs) (a) Source: Reference 16. 30 . CID xrz. a ,qua *w#T._._.,. . . •

"In laboratory rests, the lowest concentration of 2 ,4-D to cause mortality of fish was 100 mg/t, the threshold value of toxicity to perch and bleak (Alburnus Lucious) was 75 mg/t. However, certain esters and amines of 2,4-D have been found to be more toxic and, particularly in still, shallow water, may harm fish at dosages used for weed control. Fingerling bluegills suffered losses of up to 40 and 100 percent from concentrations of 1 and 5 mg/1, respectively, of the butyl ester. The isopropyl ester was somewhat less toxic but caused complete mor- tality of bluegills at 10 mg/1, as did the alkalolamine at 40 mg/t. A few fish also died during a 4-day exposure to 4 mg/t of the latter material. The sodium salt was not observed to kill small rainbow trout below a con- centration of 112 mg/i. The Fish and Wildlife Service tested a large number of phenoxyacetic acids and related compounds in rough screen- ing studies in Lake Huron water at 12 C. Trout and blue- gill were killed but sea lamprey were unaffected by 2,4- dichlorophenoxyacetic acid, butyl ester during a 24-hour exposure to 5 mg/t. Fish-food organisms vary in sensitivity to the derivatives of 2,4-D. Tests with the isopropyl ester showed that losses of over 25 percent were sustained by crustaceans at 0.1-0.4 mg/i, insects at 0.4-2.0 mg/i, and snails at 2.4- 3.3 mg/1. These animals were more resistant to poisoning by the mixed propylene glycol and butyl esters of 2,4-D, and certain species of insects and snails were not killed at 6.6 mg/L It was found that the safe concentration to minnows was 1500 mg/1 and for sunfish and catfish 500 mg/1. Some mortality of bream and bass occurred at 100 mg/i and of carp at 65 mg/t. A concentration of the sodium salt of 2,4-D of 260 mg/1 was not toxic to carp. A mixture of neutral aromatic oils (57 percent), 2,4-D (12.5 percent), emulsifiers (8 percent), and water (to 100 percent) was toxic to three-month-old rainbow trout at a concentration of 3.0 mg/1 over a 24-hour period, and at 2.2 mg/1 over a 48-hour period. A commercial weed killer that combines 6.25 percent 2,4-D and 6.25 percent 2,4,5-T with propylene glycol, butyl ether esters, and inert ingredients, in concentrations of 50 mg/t or more caused the test fish to become immediately dis- tressed. In a 72-hour period, a 25-percent kill occurred at 10 mg/t, but no fish died at 5 mg/1. 31

It is clear that few saltwater species have been assayed and perhaps no tropical saltwater species have been tested. For short ter.'11 (shorter than 24 hours) exposure, it can be assumed that less than one-half of an ester form of 2,4-D or 2,4,5-T added to water will be hydrolyzed to the less toxic acid form. Furthermore, the offsetting effect of higher temperatures should more than compensate for the lower toxicity of the hydrolyzed fraction. The 48 hour LC 50 or TLm values for saltwater fish species exposed to 2,4-D ranged from 0.3 mg/l using the PGBE derivative to 5 mg/l using the BOEE derivative. The Air Force's data using actual HO or normal butyl esters, is about the same, although a freshwater test organisms, the fathead minnow, Pimephales promelas, was used. Using the EPA methodology of determining a "safe" con- centration as 1 percent of the 96 hr LC , a value between 0.01 and 3.6 mg/l 50 2,4-D results for a water quality criterion, ignoring the possible inappro- priateness of the test organisms or test conditions. The toxicity of 2,4,5-T to aquatic species has been studied to a much lesser degree than the toxicity of 2,4-D. Comparative studies on 2,4-D and 2,4,5-T toxicity have been conducted by the Air Force on a number of species. Freshwater tests on fathead minnows showed the same trend as for 2,4-D, namely, that ester formulations were much more toxic than the acids. Measured toxicities of 2,4,5-T were 20-50 percent lowe· t'.:T for 2,4-D, however, the TCDD content of the 2,4,5-T tested was not stated. In tests using actual Herbicide Orange, the toxicity was intermediate to the two individual components. Tests on other varieties of fish hav2 been performed that show the opposite trend. Perch exhibited slightly greater toxicity response to 2,4,5-T. Again, the TCDD content was not given. Finally, saltwater shrimp comparison tests showed the normal butyl ester of 2,4,5-T to be significantly less toxic than the NBE ester of 2,4-D. The range of acute toxicities of 2,4,5-T observed in the data is 5.0 to 333 mg/L Using the EPA methodology of determining '·'safe" concentra- tions as one percent of the 96 hr Lc , a value of between 0.05 and 3.3 mg/l 50 results for a water quality criterion. 32 .,.. . a • a» ij_...,_..,.il,__,__"I__P"_ _ _ _ _ _ _ _""i...,,/ _-=

Both the National Institute for Occupational Safety and Health (NIOSH) Registry and the Water Quality Characteristics of Hazardous Materials assign aquatic toxicity range ratings of 1-10 ppm for 2,4-D and 2,4-T (5,43). Concentrations of 2.5 mg/1 for each of the components (5 mg/1 of HO) has been selected as the criterion concentration. According to the li-terature, pure 2,4-D and 2,4,5-T are con- sidered to present a moderate toxicity to humans. An extensive study of the literature on the human health and toxicity of the major and minor constituents of Herbicide Orange has been conducted by the National Academy of Sciences. ( 43 ) Judgements were made on a wide variety of organic substances relative to their carcenogenicity or the available information that would permit estimation of the "no observed adverse effect level". After a substance had been identified as a carcinogen, the risk to man was expressed as the probability that cancer would be produced by continued daily ingestion over a 70 year lifetime of 1 liter of water containing 1 ~g/1 of the substance. Assumptions required in the calcu- lation were the conversion of the standard human dose to the physiologically similar dose in the animal and the application of an exponential risk model relating dose to effect. 2,4-Dichlorophenoxyacetic acid toxicity data for man and other terrestrial species were reviewed to determine permissible intake levels. Observations in man are primarily expost facto judgements of accidental or intentional (suicidal or medical) ingestion. Poisoning and death have been attributed to ingestion of dosages ranging from 67 to 100 mg/kg. Subjects in two other studies took or were exposed to lesser quantities or similar quantities over longer time periods with no harmful effects. Observations in other species supported the moderate toxicity designation. Lo values of 100-541 mg/kg were found for rats, mice, 50 guinea pigs, chicks, and dogs. Salts and esters of 2,4-D showed an even lower degree of acute toxicity than the acid. 33 ,. ,,·· ~-.i.::.-~. ,,. . .. . .:;r.'r ·· • t . _tJ ...,._., ... >st • a ••• ct..,.«....: C ¥ .. • 1

Subchronic and chronic effects have been measured using rats and dogs. Experiments with rats showed no adverse effect levels ranging from 30 to 1,250 mg/kg and those with dogs ranged from 20 to 500 mg/kg. The results of these studies were analyzed to determine the daily no adverse effect doses. These were found to be up to 62.5 mg/kg/day and 10 mg/kg/day in rats and dogs, respectively. Based on these data, the accept- able daily intake for humans was calculated to be 0.0125 mg/kg/day. The NAS report stated that the substantial disagreements in the results of the sub- chronic and chronic toxicity studies were cause for concern and caution and that additional study is warranted. These deficiencies were considered in the determination of the no adverse effect level from drinking water shown in Table 4. Toxicity data on 2,4,5-trichlorophenoxy acetic acid and 2,3,7,8- tetrachloro-p-dibenzodioxin were considered together since most of the 2,4,5-T preparations testQ.d contained TCOO at 1-80 ppm. A few studies have been conducted with TCOO "free" material ( < 0. 05 ppm). Observations of toxicity in man depend on the TCOO content of the test material. Two studies in which 2,4,5-T containing low concentrations of TCOO was used failed to produce toxic effects in the concentration range of 1.6-8.1 mg/day. Another study where contaminated 2,4,5-T was used pro- duced cases of moderate to severe chloracne and several cases of porphyria. Toxicity testing results on other species likewise depend on the TCOO content. Early data on 2,4,5-T show oral Lo 50 v~lues for male rats, male mice, guinea pigs, and chicks were 500, 389, 381, and 310 mg/kg, re- spectively. TCOO contents were unknown. Testing of TCOO alone established its extreme toxicity as shown by Lo values ranging from 0.6 to 115 µg/kg, 50 depending on species. Subchronic and chronic effects of 2,4,5-T and TCOO have been ob- served in relatively short-term studies on rats, mice, dogs, and guinea pigs. Effects most often observed included lesions, bone marrow irregularities, degenerative liver and thymus changes porphyria, serum enzyme changes and weight loss. 2,4,5-T doses eliciting adverse effects ranged from 2 mg/kg/ day for dogs to 100 mg/kg/day for rats. TCOO doses yielding responses were as low as 0.1 µg/kg 5 days a week for 13 weeks. 34 • 0 • •C:iiiJE4ii • •

The conclusions reached by the NAS report were that contamination of 2,4,5-T with TCDD greatly increases the toxicity of the mixture from moderately toxic to very toxic. No adverse effect doses for 2,4,5-T were 10 mg/kg/day for dogs and mice and up to 30 mg/kg/day for rats and for TCDD were 0.01 µg/kg/day in rats. Acceptable daily intakes for humans were calculated as 0.1 mg/kg/day for 2,4,5-T and 10- 4 µg/kg/day for 2,4,5-T and 10-4 µg/kg/day for TCDD. The lack of data on long term toxicity and the substantial differences in toxicity values for 2,4,5-T due to varying de- grees of TCDD contamination were cited as reasons for conservative estimation of permissible drinking water concentrations, shown in Table 3. Maximum contaminant levels as contained in the Drinking Water Standards and in the 1976 Water Quality Criteria are shown for comparison. (l 3 ,lS) Ambient water standards are applied at the point of withdrawal for supply which in this case is the saltwater intake (site WS), while drinking water standards are applicable at the delivery end of the system (site Pl). There are two additional factors which serve to alter the normally encountered conditions in a drinking water supply. First, the production of freshwater is intermittent. Higher than allowable levels at the saltwater intake are not of concern if freshwater is not being produced on a given day. Second, freshwater on Johnston Island is pro- duced by distillation. The boiling points of 2,4-0, and 2,4,5-T acids (18) are related derivatives are all greater than 160 C. Therefore, the fraction of distillable HO at the process temperature is certainly less than-SO percent of the concentration on the saltwater side of the system. (See also Section 4.1.3 for more detailed discussion of these points). Since the water quality criteria represent lifetime consumption levels, the short term exposure levels could conceivably be much greater than the average and still produce no effects if the subsequent exposure is correspondingly lower to offset the initial dose. The tour of duty for most military personnel is one year; however, some of the civilians have been on the island for upwards of 15 years. It is not expected that 35 .,.1 » - u ... 4

TABLE 3. WATER QUALITY CRITERIA/STANDARDS-- DRINKING WATER Uncertainty or Compound Concentration in Water (µg/1) Safety Factor (a) ReferenceCb) 2,4-D 100 ( c) (Approval limit) 500 15 100 ( d) (MCL) 13 ( e) 37 87.5 (No effect level) 1000 4. 4 ( f) (No effect level) 1000 37 (e) 37 2,4,5-T 700 (No effect level) 100 35 ( f) (No effect level) 100 37 -4 (e) 37 TCDD 7 x 10 (No effect level) 100 3. 5 X 10-S (f) (No ~ffect level) 100 37 (a) The uncertainty or saf~ty facto·r is introduced to reflect the amount of information available on a specifi~ contaminant. An uncertainty factor of 100 represents a good set of chronic oral toxicity data available for some animal species while a factor of 1000 was used with limited chronic toxicity data or when the only data available were from inhalation studies. (b) See literature cited for references. (c) Represents lifetime no adverse effects level assuming that 20 percent of the safe intake is from water. Standard man equivalent to 70 kg and 2 liter/day water consumption used, (d) A maximum contaminant level (MCL) means the maximum permissible level of a contaminant in water which is delivered to the tape of the user. (e) No adverse effect level assuming 20 percent of acceptable daily intake is supplied by water. Same standard conditions as in (a). (f) No adverse effect level assuming 1 percent of acceptable daily intake is supplied by water. Same standard conditions as in (a). ' Ii -- 36 cu O.

the lifetime consumption would be approached by anyone on Johnston Island. Therefore, the water quality criteria are probably conservative in esti- mating risk. The most stringent standard appears to be the National Interim Primary Drinking Water Standard at 0.1 mg 2,4-D/t. Other water quality criteria pertain to the organoleptic properties of 2,4-D and its breakdown products, as well as potential non-OH related project effects such as oil and grease, turbidity, and reduced dissolved oxygen concentrations. It has been reported that 2,4-D acid was decomposed in water exposed to the sun into 2,4-dichlorophenol, 4-chlorocatechol, 2-hydroxy-4- chlorophenoxyacetic acid, and 1,2,4-benzenetriol. Taste and odor thresholds for chlorinated aromatic hydrocarbons are very low. McKee and Wolf report the taste threshold for 2,4-D as 0.01 mg/£, whereas the same concentration of dichlorophenol derivatives gives noticeable 7 tastes. (l ) Several investigators have reported the taste or odor threshold concentrations for various chlorinated phenols. For 2,4- dichlorophenol the reported taste values are 0.008 to 0.02 mg/£ and the I at 60a C. (19 ,20) odor values range frcm 0.00065 mg£ at 30 o C to 0.00 6 5 mg£ I Spills from 2,4-D manufacturing operations have reportedly produced unpleasant tastes in drinking water at dilution ratios as high as 10,000,000:1. (l 7 ) Turbidity (suspended solids) influences on fish life are divided into those whose effect occurs in the water column or those whose effect occurs following sedimentation to the bottom of the water body. Five general effects on fish and fish food populations have been noted: • direct effects on swimming fish by k~lling them or impairing physiological functions • preventing the successful development of eggs and larvae • modifying natural movements and migration • reducing the availability of food • blanketing of bottom sediments causing damage to invertebrates and spawning areas and increasing benthic oxygen demand. 37 -,-~. ,; ,, ,~:, • , , "'4'!i@Qlt '> .4 t s a es •U<u • 0 ..

, I Conversely, a partially offsetting benefit of suspended matter in water is the sorption of organics such as herbicides onto particles which leads to more rapid settling. (l 3 ) However, experiments conducted to ascertain the sorption properties of 2,4-D ester and sodium salt showed very low sorption capacity for three clay minerals (bentonite, kaolinite, and illite) and very good sorption for dry coral. Desorption properties of contaminated coral in seawater were not investigated. <21 > The criterion proposed by the EPA relates primarily to freshwater fish and other aquatic life and states that "settleable and suspended solids should not reduce the depth of the compensation point by more than 10 .< percent from the seasonal norm 11 13 ) The compensation point is defined as that depth where the rates of photosynthesis and respiration are equivalent or approximately the depth at which one percent of the incident light remains. The water quality criterion for dissolved oxygen similarly pertains to freshwater aquatic life. A minimum value of 5 mg 0 /1 is 2 given. (l3) Effects of oil and grease on ocean communities range from inhibition of oxygen transfer when heavy concentrations are present on the water surface to acute or sublethal toxicity to specific compounds present in the oil. Because of the range of possible compositions, criteria have been specified with respect to bioassay techniques on important species: For domestic water supply: Virtually free from oil and grease, particularly from the tastes and odors that emanate from petroleum products. For aquatic life: • 0.01 of the lowest continuous flow 96-hour LC to several 50 important freshwater and marine species, each having a demonstrated high susceptibility to oils and petrochemicals. • Levels of oils or petrochemicals in the sediment which cause deleterious effects to the biota should not be allowed. • Surface waters shall be virtually free from floating non- petroleum oils of vegetable or animal origin, as well as petroleum derived oils. 38 - ... _.._.. _... _,. _ -- ............ ----- .,,, • . . ~·,--1-• :, ... > • pc;~~-A ~ .

2.1.3.4 Existing Water Quality , i j 1 The salt waters around Johnston Island and the freshwater system have been monitored for the presence of 2,4-D and 2,4,5-T since 1973. Eight locations, including an offshore control, have each been sampled a number of times. Table 111-13 is a summary of baseline water quality data gathered by the Air Force from 1973 to 1977. The data show occasional instances of HO being detected at most of the locations. Of special significance to the disposal operation are those locations which were also sampled by BCL during Operation Pacer HO, These include the wharf, the south side of the island, the offshore area near the herbicide storage yard, the saltwater intake and the distillation plant. Corresponding site codes used in the Pacer HO operation are WF, WO, WD, WS, and Pl, respectively (Figure 7). The maximum concentrations hist?rically observed by the Air Force in the offshore area near the Herbicide storage were on the order of 3 µg 2,4-0/liter and 0.6 µg 2,4,5-T/liter and those near the saltwater intake were 2.3 and 0.7 µg/£, 're.spectively. The other two offshore sites exhibited maximum concentratiops below 0.5 µg/£. Samples taken in the distillation plant never showed measurable concentrations, yet one sample from the storage reservoir showed 1.6 µg/£ of 2,4,5-T, This number is not only much higher than any of the other concentrations from the reservoirs, but also reverses the trend for the 2,4-0 concen~rations to be greater than those for 2,4,5-T. Data gathered by Battelle during the baseline monitoring period from July 24 to July 27 shows 100 percent of all samples analyzed below the quantitative detection limit of 0.2 µg/£ (ppb) (Table 4). It can therefore be concluded that the water environment at Johnston Island has in the past been affected by the storage of Orange Herbicide, but that, immediately prior to the dedrum/transfer operation, the water showed no serious degradation in quality from the herbicide. 39 •

f, .... .f.' ... ;, I, I I~ ~--::z:::-_ - Q::.:::::::=~=z~ --------- ... --------------- --·---- --- --·--.....--·- ------ __ ~ - --------- -·----- -::::;:::: . .,. .,, -~-::::::::.:.::-------~- __ .,..,.___ _.,.____ .,.......... . ... . .,.,. ------ ----- ... ..,..,. - - -- -. ... - . ;t~ ' f:. .:""t f .,. 0 _______ _..---,;.;. --- ---~~-- -----------~ ----- 0 Water Sampling Sites- (WD,WS,WF,WO, 0 Johnston Island Crew Facilities P,SE) C Sediment Sampling Sites ~ Wharf II Analytical Building CJ Dedrum Area EB Red Hat Area FIGURE 7. WATER AND SEDIMENT SITES ;'\

. . . .~ 1:'; .('-~ . ._........ ... 1 -}•• . TABLE 4. OPERATION PACER HO DATA SUMMARY-WATER PRE-OPERATIONAL f 'I' Maximum Minimum Positive Average Percent Percent Percent No. in PPB in PPB in PPB Positive Trace N.D. Location Samples D T D T D T D T D T D T WS saltwater 4 <.l <.1 <.1 <.1 -- -- 0 0 0 0 100 100 intake WF wharf 4 <.l <,l <,l <,1 -- -- 0 0 0 0 100 100 WO wastewater 3 <.l <.1 <.1 <.1 -- -- 0 0 0 0 100 100 outfall .... ~ WD downwind 1 <.1 T <.1 T -- -- 0 0 0 100 100 0 I dedrum • Pl&P2 potable 3 T T <.1 <.1 -- -- 0 0 33 67 67 33 water SEl&SE2 sewage 1 <,1 <.l <.1 <,l -- -- 0 0 0 0 100 100 RW rainwater 0 \. '-, II' ~-~

Other environmental indicators measured were temperature and dissolved oxygen. The mean monthly water temperature for Johnston Island · 2 6 . 4° C. (ll,l 2 ) Th e water temperatures measure d by f or July an d August 1s BCL during the baseline period were 26.8° Cat the wharf, 26.1° Cat the saltwater intake, and 26.4° Cat the wastewater outfall. Dissolved oxygen concentrations at all the offshore sites were near saturation for an assumed chloride concentration of 15 ppT (parts per thousand). No values below the water quality criterion of 5 mg/R were observed. Dissolved oxygen concentrations were lower in both the potable water and sewage samples as expected. Potable water composite samples showed mean oxygen concentrations of 6.0 ~ 0.3 mg0 /£ or 81 percent of saturation at 32° C. 2 Sewage samples were nearly anaerobic measuring only 1.1 ~ 0.2 mg/£ of oxygen at a temperature of 32.5° C No a~ute adverse environmental effects in existing water quality were noted during the baseline monitoring. 2.1.4 Groundwater Quality On Tuesday, August 25, two days after the dedrumming operation was completed, a 6 in. water sampling well was found in the barrel storage yard (Plate 2). The exact location is shown on the engineering drawings of the island and has since been filled in with coral. The well casing terminated flush with or just below the ground surface possibly permitting surface water to flow into the well. Therefore, it was judged not to be a good site for groundwater sampling. A sample of the water was nonetheless examined by smell and found to have a distinct odor of HO indicative of contamination. The water table was measured at the hole and found to be 9 feet 3 inches below the ground surface. This measurement was taken near a period of low tide. The Air Force has monitored contaminants in test wells, as docu- mented below: 42 ·.-,,,.. ~ ,. • ...,~_JFW-.Lf; Qa-.,

Historical Groundwater Data Summarr TCDD Detection Analyses Location Limit (PPT) Results L- Well hole-center of herbicide area 0.37 ND Well hole-west side of herbicide area 0.24 ND Analises Results 2 ng/i 2 1 4-D 2 1 4 1 5-T Ester Acid Ester Acid Location 200* 100* 50* 20* Well hole-center of herbicide area ND 44,000 ND 1,200 Well hole-west side of herbicide area ND 77,000 ND 3,600 2.2 Biological Environmental Features of Johnston Atoll 2.2.1 Terrestrial Environment The terrestrial environment of Johnston Atoll has been extensively (1,2) studied, As a result, much is known about the plants and animals which inhabit the four islands of the atoll. 2. 2.1.1 Plants To date, 51 families, 109 genera, and 127 species of vascular plants have been identified from the four islands of Johnston Ato1f 1 • 2 ) Table 111-5). This number of plants is remarkable in view of the fact that only three species existed in 1923. These three plant species are believed to have reached the atoll by natural means, either by water currents, air, or birds. The majority of the remaining 124 species have been introduced by man. Undoubtedly, some of these introductions were intentional, others came as stowaways or adventives, * Detection Limits, ng/t. 43 ·, .. • / ~ . . , ~ _ p; ~.,4,_~.,.--~~iiiiirlliil1alliiiiiiiiiiii•..•••1•-=i12•2•■01111111ii••llll'l•~siiiiiiilii'f1W•w,__,_,.•....,..,..,..________.,,.,1-'-~-.~-

Materials from land clearing and dredging operations have increased the size of the two original islands and have made two new islands, thus bettering the opportunity for more plant species to become established. Disturbed soil coupled with the freedom from competition from established flora have created conditions suitable for the establishment of many plant species. Undoubtedly, because of the poor soil and climatic conditions, many of the ornamental species intentionally introduced by man would not survive if not frequently cared for. Of the 38 species of vascular plants found on Akau Island a majority have been transplanted from Johnston Island. This man-made island was completed in 1964, and by September, 1967, 31 species were found there. Fimbristylis cymosa grew over most of the island and was the most predominant species. Other species which were common were Spergularis marina, Sesuvium portulacastrum, Eleusine indica, and Cynodon dactylon. A similar plant distribution was noted in November 1973. To date, only 14 species of plants have been recorded from Hikina Island. The construction of this island was completed in 1964 and by September, 1967, five species of plants were found growing there. Only three of the five species were abundant, Fimbristylis cymosa, Sesuvium portulacastrum and Spergularis marina. Two additional species, Eleusine indica and Lepturus repens, were also present in 1969. The flora was found to be similar in 1973. In 1923, only three plant species were known to be growing on Johnston Island. Early photographs of the island reveal that Lepturus repens was the dominant species. By 1967, 111 plant species were recorded from Johnston Island, many of which were under cultivation by residents. Major species were Pluchea carolenensis, Cenchrus echinatus and Casuarina equisetifolia. There are 54 species of plants which have been recorded from Sand Island. Only three species (Lepturus repens, Boerhavia repens, and Tribulus cistoides) were known to the original portion of Sand Island in 1923. Lepturus repens was the dominant species. By 1967, the number of plant species known to the original portion of Sand Island had increased 44 it • • ....,....,__,_....,____________,,.,;... // ..

to 25. At this time the fivt> most common species wt>re Lepturus .!:_l~~n_s_, , ' Tribulus cistoides, Sesuvium portulacastrum, Boerhavia repcns and Amaranthus viridis. The man-made portion of Sand Island was completed in 1941. By 1967, 50 plant species had been recorded as growing on this portion of the island. The most common were Fimhristylis, Conyze, Sanchus, Cenchrus, Pluchea, Cynodon, Sesuvium, Euphorbia, and Scaevola. A similar distribution was found in November, 1973. 2.2.1.2 Invertebrates The terrestrial invertebrate fauna of Johnston Atoll is not well known. Insects are the only member of the invertebrate fauna which have been studied to any extent. Insects totaling 68 species of 35 families are known from the four islands of Johnston Atoll (Table III-16). 2.2.1.3 Vertebrates a. Fish There are no freshwater fishes which inhabit the islands of .Johnston Atoll. b. Reptiles Four species of reptiles are known from the terrestrial environment of Joh::ston Atoll. These species are Hemidactylus frenatus (house gecko), Hemidactylus garnotti (fox gecko), Lepidodactylus lugubis (mourning gecko) and Ablepharus boutonii poecilopleurus (snake eyed skink). c. Birds There are 56 bird species which are known to the islands of Johnston AtoJl (Table III-17), which constitute a national bird refuge. These species belong to 10 orders, 19 families, and 38 genera. Twenty- 45 a••:~ •• • C

two species are classed as sea birds and 34 species are waterfowl, marsh, and land birds. Of the 22 sea bird species recorded at : ; .. · .. , Ato i. ·, l2 are breeders, 3 are former breeders, and 7 arc visitors r I'ar te 11 '-18). Al 1 of the 12 resident breeding species also nest in th,o Hawaiian Islands and other parts of the tropical Pacific. The three species which formerly bred at Johnston Atoll are Diomedia nigripes (Black-footed Albatross), Diarnedia illDDutabilis (Laysan Albatross), and Sula dactylatra (Blue-faced Booby). The seven sea bird visitors to Johnston Atoll came from the north, south, and east Pacific. The 34 species of waterfowl, marsh, and land birds recorded at Johnston Atoll are divided into five groups: regular migrants consisting of seven species, irregular visitors consisting of six species, stragglers consisting of two species, accidentals consisting of 16 species, and introductions consisting of three species (Table III-18). The annual breeding and bird population cycles vary greatly among the bird species at Johnaton Atoll. The sea birds breed during all seasons of the year (Figure 8). Nine of the 12 presently breeding sea- bird species breed during the spring and summer seasons. Thus, May through September is the peak breeding period for the sea birds of Johnston Atoll. Many of the bird species known to Johnston atoll leave during part of the year while others stay throughout the year. There is however, a population buildup for each species sometime during the year. The breeding population of sea birds of the Atoll consists of 12 species. However, only five species are dominant in terms of total numbers (Figure 9), The Sooty Tern, with a mean population of 300,000 to 310,000 breeding birds dur1nr? March, April, and May, makes up 95 percent or more of the total ,\ l .,,pulation between March and July. Possibly as many as 600,000 Sc·· 1 ;· Terns used Johnston Atoll annually. Red-foot<.· ·.,,.;bies, whose mean population ranges up to 3,750 birds, ranks secor,G , n ~ea bird numbers in winter and spring. Most of these birds are t1 ;::as tents for only a few young are produced each year. 46

-----------------------------~ Bulwer's Petrel ....... ~ Wedge-ta i I ed Shearwater Christmas Shearwater ••• ••••• ~ l 1 1 I D::D:::.. Red-tailed Tropicbird Brown Booby Red-footed Booby Gray-backed Tern ••••••••• Sooty Tern •••• Brown Noddy •,. ,,.&r1M!i8tttt(bl=l44:tltt~ ...... Black NO<ldy ....... ,:ffihthVbkt~................ White Tern J F M A H J J A s 0 N D FIGURE 8. BREEDING CYCLES OF SEABIRDS AT JOHNSTON ATOLL: STIPPLED AREA REPRESENTS EGGS, BARRED AREA YOUNG, AND BLACK DOTS NONBREEDING BIRDS 47 . ,I• • C a D stall«W a ..

i 320,000 270,000 c::::::J ~ooty Tern 220,000 ~ Red-footed Booby 1111 Brown Noddy llfJII Wedge-tailed Shearwater 170,000 - Great Frigatebird Ill Others 120.000 ... VI ~ Ci0 :E: ::, .-.: ... 70,000 - > I- < ..J :::, ::5: :::, u 20,000 ,, ~ ....., ~ ,,~,' ,, 9,000 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1.000 0 ,J F M A -M J J A s 0 N 0 FIGURE 9. MONTHLY CUMULATIVE BIRD POPULATIONS, JOHNSTON ATOLL, 1963-1969 48

The Brown Noddy ranks third in mean population numbers. The Wedge-tailed Shearwater ranks fourth in numbers of adults using the Atoll but is present only from March to early December. The Great Frigatebird, with a main peak of 750 in March and April ranks fifth in population numbers. Mean monthly populations of all other species combined never totals more than 600 nor less than 300 birds. Of the seven regular migrants, only the American Golden Plover, Wandering Tattler, and Ruddy Turnstone are known in all 12 months. Although the Wandering Tattler is present in low numbers throughout the year, American Golden Plovers and Ruddy Turnstones show peak p ,'ations of 120 and 100, respectively, in fall and mid-winter (Figu1c 10). The four islands of Johnston Atoll vary with ref ;ivct to .; ,_ze evaluation, soil, vegetation, and degree of human disturbance. Major differences exist in the ecological distribution of bird ~pecies hetween disturbed and non-disturbed islands. This is particularly true for the bird species which breed on the islands of the Atoll. Fifty-two of the 56 bird species known to the Atoll are known to Sand Island. Of these 52 species, 44 are known from the original portion while 35 are known from the man-made part. Furthermore, 35 species are known from Johnston Island, while eight are from Akau Island and five are recorded from Hikina Island. The bird populations of Akau, Kikina, and Johnston Islands are known to be small in comparison to that of Sand Island. The population cycles shown in Figure 8 are essentially those of the birds on Sand Island. During the spring and summer, Sooty Terns are most predominant species and nest on the bare ground over most of the island (Figure 11). The nesting areas for other species are shown in Figures 12 and 13. Brown Noddies nest on the ground around the perimeter of the island. Red-tailed ;ii Tropicbirds nest under low vegetation about the island. Wedge-tailed Shearwaters nest in burrows over much of the island. Brown Boobies nest on the ground on the southeast hill, the south edge, the northeast penninsula, and the southwest islet. Red-footed Boobies build their nests on the east hill, on the Tournefortia bush northeast of the transmitter buildings. Great Frigatebirds nest along the east hill and the south edge. 49 t a a • ciC S D D W •

130 120 110 100 90 .. 80 •• •• 70 .•· .. .·. •••• •• •• •. •• •. 60 • • •• V, •• a:: ...., m ::E :) 50 • •• • •• z •• 40 \. •• 30 • •• 20 •• •.....__..,. .. •••. 10 0 J F M A M J -- . J _ .,•--·--• ..... _ _.., A s 0 N .... D FIGURE 10. MONTIILY MEAN SHOREBIRD POPULATIONS FOR JOHNSTON ATOLL, 1963-1969; GOLDEN PLOVER (SOLID LINE), RUDDY TURNSTONE (DOTS), WANDERING TATTLER \. (DASHES) 50 a a es .,-Z-W 0.

February April ·::::-:-:-!'::::::::'.-.L:::J- ·-- ··-- "•-===:. ' -,-.--·-.. . -=-~ June c:J August -==.:·~{~ ::.:::::~: ::.-::--.·.t:J - ,:,,:, - ,::.--. ~,,~ 0 toO 200 CJ October December '"' FIGURE 11. AREAS USED BY SOOTY TERNS (STIPPLED) AND WEDGE-TAILED SHEARWATERS (BARR.ED) ON SAND ISLAND, JOHNSTON ATOLL, 1965 51 • .... ...__.. 4. • • f / / Jl,I

• • Brow,1 ,,oddy I!> G,·ay-bad.ed T~rn N ;j l ' o Brown Booby -----;: :·. .. 0 100 200 A I fHt FIGURE 12, NESTING AREAS OF GROUND NESTING BIRDS (EXCEPT SCOTTY TERMS) ON THE ORIGINAL PORTION OF SAND ISLAND, JOHNSTON ATOLL, 1963 * Black Nodrly e Great Frtgateblrd N @ Red-footed Booby f ~:. ________ ~:~b \ -~~,,, t ., •'•' . 0 100 Feet zoo FIGURE 13. NESTING AREAS OF BIRDS WHICH NORMALLY NEST IN LOW VEGETATION ON THE ORIGINAL PORTION OF SAND ISLAND, JCllNSTON ATOLL, 1963 52 u lJ 2 St a, a a•• a Q 0. • = 4 U

Mortality in the bird population of Sand Island has been studied. The main cause of mortality was birds flying into the guywire system of the LORAN-C antenna. This system contained 24 top-loaded guywires which stretched from the top of the 625 foot tower to concrete pillars located in the lagoon in a circle around the island. There also were three sets of guywires stretched from part way up the tower to two sets of concrete anchors located on or near the periphery of the island. d. Mammals There are no manunals native to Johnston Atoll. With the exception of human occupants, five species of mammals are known from the terrestrial and one species from the marine environment of the Atoll (Table III-19). It is likely that the two rodents arrived in ship or plane cargoes, while dogs, cats, and rabbits were purposely introduced by military and civilian personnel. 2.2.2 Marine Environment The marine environment of Johnston Atoll has been studied to a considerable extent. It has been heavily disturbed by man during dredging operations associated with the deepening and lengthing of the ship channel and seaplane landing area. 2.2.2.1 Plants Prior to the dredging operations of 1964 only one marine algal species was known to Johnston Atoll. In 1965, as part of a study of the effects of dredging on the marine environment, 67 species of benthic marine algae were collected from Johnston Atoll. Additional collections in 1966 added 26 more species to the known species list. In all, 93 species of benthic marine algae are known from the waters of Johnston Atoll. Of the 53 . ·~~ .'·~~-!~·~.~~:-, • J2¥tsec;,~z Dt C • • 4 C O W

93 species, 12 were found only from the marginal reef, while 33 were found only in the lagoon waters (Table 111-20). Of these 33 lagoon species, 11 occurred only in open water, 11 were found only in the inshore area of Johnston Island, and 2 were taken from the inshore area of Sand Island. 2.2.2.2 Invertebrates The invertebrate fauna of Johnston Atoll is not well known. Several scattered collections ~ave been made but no extensive systematic sampling programs have been conducted. There are 18 species belonging to 11 genera of Cnidaria (hydras, jellyfish, sea anemones, and corals) which are known to Johnston Atoll (Table 111-21). Fifty-eight species of Mollusca have been collected from t~e Atoll (Table 111-22). To date, only 12 species of Annelida belonging to 8 families are known from the lagoon waters. These are listed in Table III-23. A total of 75 species belonging to 20 families of Crustacea have been recorded from the lagoon waters at Johnston Atoll (Table 111-24). 2.2.2.3 Vertebrates The marine vertebrates of Johnston Atoll are well known. Fish species have been studied most extensively and are separated into two categories: pelagic fishes and inshore fishes. a. Fish Numerous large pelacic fishes have been recorded around Johnston Atoll. Although no extensive species list exists for this area, various species of tuna, sharks, and barracuda are known to occur in the waters around the Atoll. To date, a total of 194 species of inshore fishes have been recorded from the waters of Johnston Atoll (Table 111-25). A majority of these species have also ?een found in the fish fauna of the Hawaiian 54

Islands. Only two of the 197 species have not been recorded elsewhere. These are Centropyge nigriocellus and Centropyge flammeus, neither of which is abundant at Johnston Atoll. b. Mammals The Hawaiian Monk Seal is the only mammal recorded from the marine environment of Johnston Atoll. These are known to have arrived from the resident population of the northwestern Hawaiian Islands. It is also likely that porpoises visit the Atoll's lagoon waters, although an official record has not been made. 2.3 Human Environment 2.3.1 Economic and Social There is no indigenous population on Johnston Atoll. Rather, the population is transient representing 4 agencies; namely (1) The Air Force, who administer the island. (2) The Army, whose "Red Hats" guard and maintain a munitions storage area. (3) The Coast Guard, who maintains the LORAN equip- ment on Sand Island. (4) Holmes and Narver, Inc., staff, who perform island maintenance, food, laundry, medical etc. services. There is not a local economy, all goods and services being provided by these agencies. The island personnel live in a cooperative atmosphere with very little violence or crime·. People who do not abide by the established standards of behavior are rapidly and permanently transferred from the island. 55 •

3. DESCRIPTION OF ORANGE HERBICIDE DISPOSAL PROGRAM 3.1 Purpose Following the decision by the Secretaries of HEW, Agriculture, and Interior in 1970 to suspend some uses of 2,4,5-T, the Air Force conducted an environmental impact study to determine the most ecologically sound method to dispose of the 2.4 million gallons of Orange Herbicide stored on Johnston Island and at the Naval Construction Batallion Center, Gulfport, Mississippi. The approved alternative for accomplishing this objective was the dedrumming of the herbicide at Gulfport and on Johnston Island and the transferal of the TCDO-contaminated material to the Dutch-owned incinerator ship, M/V Vulcanus, for thetmal destruction. The operational plan and subsequent activities discussed in this report represent the Air Force's efforts to implement the recommendations contained in the final environmental statement, and to comply 6 with the provisions of EPA permits. (l ) 3.2 Operational Procedures 3.2.1 Physical Manipulations Physical manipulations as discussed include only those portions of the overall operation plan which specifically had implications for causing environmental degradation of the island or its immediate offshore area. 3.2.1.1 Drum Handling-Dedrununing The 1.5 million gallons of Orange Herbicide stored on Johnston Island represented approximately 25,000 drums of 55-gallon capacity. These were stored in rows stacked three high in an area of about 3.5 acres on the northwest corner 56 • •

of the island. A dedrum facility was modified* to allow transfer of the material from drums to bulk carriers for transport t0 the ship. The facility and operation basically consisted of a covered concrete pad and two fabricated metal racks upon which full drums were placed in four groups for 12 each. Drums were transported from the drum yard in sets of four using fork lifts equipped with specially designed clamps. Each set of 12 drums was handled independently by the dedrumming crew. Once the drums were on the rack and the fork lift had withdrawn, a crew member would punch one hole near the top of each drum to allow the crew's supervisory personnel to check the contents of the drum for Orange Herbicide.** Any sus- picious looking drums were removed from the line and held for further testing prior to loading. Three closely spaced holes were then punched in the bottom of each drum and the fluid allowed to drain. A set drain period of S minutes was determined in prior testing to give the most rapid throughput of drums and still achieve good drainage. Following the 5-minute drain, the inside of each of the drums was rinsed with 1 gallon of diesel fuel using a spray wand. Operators were instructed on the proper technique to cover the entire drum interior. After draining for 2 minutes, a second one-gallon spray rinse was initiated and 2 minutes allowed for draining herbicide and rinse drained into a trough which flowed into a sump equipped with pumps to transfer the material to a tank truck. Quality control procedures were carried on through the entire operation. In addition to the testing of contents mentioned previously, samples of the second rinseate were obtained from about every hundredth drum. A total of 219 such samples were taken. A target value of the sum of the concentrations of 2,4-D and 2,4,5-T was derived from test rinses con- ducted by the Air Force at the Naval Construction Battalion Center, Gulfport, * The facility had originally been installed for redrUJ1D11ing of leaking drums. ** Drums containing material other than HO were taken off the rack and sealed for future disposal action. Only HO was allowed to drain. As the EIS and permits were only for the destruction of HO, other chemicals could not be allowed to mix with the HO in the sump. Each barrel was examined by pipettinR a sample prior to drainage. Visual and olfactory examinations were used to verify contents as being HO • • .. i

Mississippi. The level of rinse achieved was to be equivalent to the Environmental Protection Agency triple rinse procedure.(16) Recommendations on the Johnston Island drum rinse procedure were made by Battelle-Columbus Laboratories based on the results of these studies: • Five spray rinse studies showed that the ~irst rinse efficiency averaged 68 percent removal (range from 64 to 74 percent) while second rinses averaged 69 percent removal (range from 62 to 79 percent). As an approximation, the first and second rinses yielded the same efficiency of 68 percent removal. • Thirty-five drainage studies showed that, on the average, total mass of 2,4-D and 2,4,5-T remaining in a drum after being allowed to drain for 5 minutes is 261.29 grams with a standard deviation of 139.73 grams. • The herbicide mass removed in the second rinse was shown to be proportional to the first rinse residual. Increased draintime decreases residual and, hence, second rinse herbicide mass. Increased wash efficiencies on the first rinse also cause a decrease in the second rinse mass. • Using 68 percent rinse efficiency, and the distribution of residuals from the drainage studies, it can be shown that 50.6 grams of herbicide in the second rinse represents 85 percent removal with 99 percent confidence bounds. Likewise, 46.1 grams represents 90 percent removal. • Assuming the rinse volumes are exactly 1 gallon (3.785 liters), the sum of the concentrations of 2,4-D and 2,4,5-T for 85 and 90 percent removals (99 percent confidence) are, respectively, 13.36 mg/ml and 12.18 mg/ml in the second rinse. Nominal values will be at 56.4 grams or 14.9 mg/ml (for population mean, nominal 90 percent removal). Because of the overlap, a 68 percent confidence bound was suggested, Accordingly, the 85 percent removal for these upper and lower bounds requires maximum second rinse concentration of 15,30 mg/ml (Figure 14 ). 58 ,, . "' f D LC c•a c $ ••• z....-..,,. .__....,...,.....--------aa-....,,/ 1-

Overall Removal Efficiency .85 By analysis, all second rinse concen- trations should lie to the right of th .90 70 line. However, the original residual is not known. results in this triangle yield 60 "out of control" Control Line ,__...,________ .95 signals By NCBC tests, all second rinse concentrations should 50 be below the control to be"in control" 40 ~ CfJ .99 CfJ 30 ~ a) CfJ c:: "M er: "Cl 20 c:: 0 u a) rJl 200 300 1r1-.ln.1i R,•'41dual "ldss, Grams • IS .. -- ...

• Because of randomness in the original residual mass, the proposed quality control line is only one-half standard deviation from the expected value for the residual mass of any given drum. Thus, 34 percent of the individual sample results will appear to be out-of-control if plotted. Accord- ingly, a more accurate trend line can be constructed if only the average concentration of every five samples and the total running averages are plotted. Figure 15 shows the results of the drum rinsing for all data obtained. Occasionally a series of samples would show a very high average and move the running average up toward the control line. This problem was encountered early in the program and again during the second loading operation. The operation was analyzed following the first loading to determine why the quality control program showed this behavior. During the first half of Operation Pacer HO, 121 drum rinse samples were analyzed. The overall average concentration for these samples was 17.33 mg/ml of second rinse or.65.5 g/gal. To have achieved the required control level, the concentrations should not have exceeded 14.90 mg/ml. It was noted during.Battelle's observation of t.1 ::> dedrum operation that the pipet used to obtain drum rinse samples w.1.s of~.-.,-, placed in close proximity to the pipet used to check the drums for suspir:ious material, inviting a mix-up. The effect that this would have on the rinse quality control would be to have one sample be very high and successive samples be diluted in proportion to the original ~ontamination and the actual rinse efficience. Other possible reasons for the extremely high values observed, none of which have any bearing on the actual rinse efficiency achieved, are an unrepresentative sample of drum rinse or an accidental first rinse sample. The first is caused by a delay in taking the sample and results in a sample which has separated into its component phases, Since the HO is much more dense than either water or diesel fuel, a sample obtained from the bottom of the container would have exhibited a much higher concentration of herbicide than a well mixed sample. The second, although not directly observed, could easily have occurred during an operation of this nature. 60 ? 11...•___□_:c-._....,.____.._-__•__•____._,____2______•__._-_ _._7C_.,, __-__::::::::::·•-t.·-:.::~!j. . . __-_._-_~

~ 1"·"':1;;.•=:·· , ~ . 30 I I l I I I First Ship Load I --,I~I I I I Second Ship Load 25 I I --.. 20 ! /- ~ • ..... ~I '- e CD e ~ C: 0 ..; .u 15 - ,..<11 :/ i Control • ,...°' u .u C: Q) u C: 0 10 - I, .I Line I I I 5 . I I I I 0 I 0 25 so 75 100 125 150 175 200 225 250 Sample Number FIGURE 15. DRUM RINSE--QUALITY CONTROL CHART, RUNNING AVERAGES (ALL DATA)* * Assuming N.D. (<l mg/ml) 1.0 mg/ml '

Since it could not be determined which of the samples was affected, a statistical review of the rinse procedure was used to determine an upper bound for contaminated or otherwise biased samples. At a nominal residual of 261 grams and a 69 percent efficiency for the first and second rinses, the concentration in the first rinse would be 47.57 mg/ml and in the second 14.77 mg/ml. Furthermore, for the standard deviation of 139 grams, one percent of the drums would be expected to show as much as 678 grams of residual. With "worst case" assumptions of a 64 percent first rinse and 79 percent second rinse, the expected second rinse concentration for one percent of the population is 50.94 mg/ml. It was expected that, on the basis of the statistics, one percent of the drums sampled would have shown a true second rinse concentration of greater than 60 mg/ml. These could not legitimately be rejected as outliers. At the same time, the nominal first rinse concentration was about 48 mg/ml. If an accidental first rinse sample were included, its concentration would have been about the same as the "worst case" residual described above. A first rinse sample should be rejected. A compromise between the errors involved in including a first rinse sample as an estimator of second rinse efficiency and of rejecting a true second rinse which falls on the "tail" of the sampling distribution was needed. It seemed reasonable, therefore, to reject as outliers all samples showing second rinse concentrations in excess of 47.0 mg/ml. A total of nine samples were rejected during the first loading period and 14 during the second loading period. The resulting running averages are shown in Figure 16 and are seen to comply with control cond1tons. Suggested improvements to the drum rinse quality control program were as follows: • Control of the drum rinse sampling pipet should be by the person who counts drums. He should also be responsible for selecting the drum to be sampled so as to assure that one station is not biasing the sample. • As the drum is sampled, he or another man should make sure that a second rinse sample is being taken and not a first rinse. It may be that in the confusion of the operation mistakes are being made. i i' I 62 : f.. =-~,. ·:. ~"J· . •·:· ' . . . • a s W d W' I •

-- r -- :~·.•~·t • •f; r, 25 - I I First Ship Load I I ~,.:I 1 I I Second Ship Load I - I 20 r ~b QI Ill C: -.-4 I iii:: I "Cl I 5 15 CJ ' (!) I ~ Control °' w ..... - ti) - I -- Line C: -.-4 ~ .-I ~ 10 - .~ I ClO a I I I • 5 I r I I 0 ! I 0 25 50 75 100 125 150 175 200 225 250 j Sample Number I 1 FIGURE 16. DRUM RINSE--QUALITY CONTROL CHART, RUNNING AVERAGES (47 mg/ml cutoff on high end= 179 g/gal) .. , -~·-··

• The sample container should be stirred with the pipet before sampling to obtain a more homogeneous sample. • The pipet should be specially marked with paint or other easily seen and indelible marker. • The location of the sampling pipet should also be marked to avoid cross contamination with drum-test pipets. These should not be kept near the sample pipet. • The sample should be drawn using the index finger rather than the thumb. This will maintain better control and permit faster sampling, thus, minimizing the possibility of in- homogeneity. • The drum selected for the rinse sample should have only one drain hole punched in it. This facilitates the capture of the rinse in the gallon can. • The drum counter is also responsible for assuring that the drum drains for exactly 5 minutes prior to the first rinse. A second category of special drum rinse samples consisting of a set of four run in duplicate was used to verify that the concentrations of HO in 3,300 previously emptied drums was below the control line using only a single rinse. The material in these drums had been subjected to weathering for a period of from two weeks to over six years. The mean concentration using one gallon rinse was 3.56 mg/ml (13,5 g/gal) with a standard deviation of 3.21, Thus, there is negligible probability that the observed values do not meet the EPA triple rinse criterion. The Air Force's decision was to forego further quality control testing on the remaining emptied and weathered drums and to provide a single one gallon rinse to these drums. After the second rinse had been allowed to drain for two minutes, the drums were removed from the racks by rolling them the northwest corner of the dedrum facility. Fork lifts with a ramp attached to the forks were used to transport the empty drums to the crusher. 64 • --,.,. a 2 a a uw •

3.2.1.2 Drum Crushing The fork lifts transported six drums per trip to the crusher feed ramp. Drums were fed to the crusher one at a time. The crusher consisted of a large weight suspended between two I-beams. The drums were compressed along the longitudinal axis. There were no spray shields around the crusher to trap the mist of oils and residual HO which was released on impact (Plate 3). Several times the crushing operation fell behind the dedruuuning operation and the empty drums were stacked up on the ground around the crusher. Crushed drums were bundled and placed in storage on the seaward (downwind) side of the dedrum/crushing area. A large plastic sheet was used to protect the crushed drums from rain. 3.2.1.3 Transport of HO to Disposal Ship Herbicide was pumped from the collection sump into standard Air Force R-5* refueling trucks (Plate 4) via a dry coupler bottom connection. Because of the difference in density between the HO and JP-4, the R-5 1 s were only filled with 3,000 gallons of HO versus a 5,000 gallon capacity. During the filling operation, a drip pan under the coupler was used to prevent any herbicide from contaminating the loading pad. When disconnection took place, a few drops at most were observed to be discharged into the pan. The refuelers transported the HO to the wharf via a road which was set aside for this purpose. Non-project related vehicles were forbidden traffic along this section of roadway. 3.2.1.4 Transfer to Disposal Ship Once the refueler had reached the main wharf, the procedure was essentially reversed. The same type of dry couplings and spill prevention equipment were employed to pump out the tank and bulk transfer the material * The pumps on the R-5 were bypassed to prevent their contamination and i seal destruction by HO. I• 6S - -_-t-,-:·-~--.-•-..--•·· --,..ti ,;=-,--.-_ •• a C .... • •

to the ship. The area in which the pumps and hoses were located was diked with sand bags dnd plastic so that as much as a full truck load of spilled material could be contained (Plate 5). All hose-to-hose couplings were similarly wrapped in plastic to catch any herbicide. Under normal conditions, an R-5 could be emptied in about 20 minutes with another arriving to replace it just about the time it became empty. The only problems noted in this operation were the clogging of I screens used to trap sludge particles, and the formation of a flow retarding vortex in the R-5's. I ~ I 3. 2 .1. 5 Cleanup ~ I I After the last HO had been transferred, all of the equipment, trucks, etc., were rinsed and decontaminated with diesel fuel which in turn was transferred to the ship. 3.2.2 Descriptions of Project Activities This section provides, in outline form, all environmentally relevant project related activities contained in the official memos for the record or in BCL project records. • July 23-- All personnel involved in the project were briefed by the Project Director on matters of spill prevention, countermeasures in case of ~pills and personal safety. Contingency equipment was inspected and positioned. • July 23-24--BCL task leaders held discussions with corres- ponding Air Force officers regarding placement and start-up of land-based environmental monitoring (see Section 3.2.3). • July 24--First day of baseline environmental monitoring. • July 25--M/V Vulcanus arrived at approximately 1500 hours. • July 26--Training operations for dedrum crew began at 1300 hours. Three drums were taken through procedure on day shift and three on night shift. • July 27--Full-scale loading operations commenced at approximately 1500 hours. Several small leaks in R-5 were noted and corrected. I. 66 ;llir-.~- t~ ~r • a t aa u 167•- ••

One badly leaking drum was located and removed to the dedrum rack. Clean-up was instituted. An estimated 25-30 gallons were spilled onto the coral storage area. • July 28--A very small (<l gallon) spill on the wharf was noted. No water contamination was observed and spill clean-up was accomplished. • July 30-- During deballasting, an orange colored plume was observed on the port side of the M/V Vulcanus from 1100 hours to 1800 hours. Black oily trailings were visible in several places. Samples were taken at 1100 hours near the discharge of the deballast pump at a depth of 1 meter below the surface. • July 31--Air Force was informed of preliminary air and water sampling results. • August 1--The Air Force was advised on the trend of the drum rinse quality control results up toward the control line. Results of previous day's deballast water sample submitted to Air Force. • August 2--EPA decision to require one.tank filled with pure herbicide will result in 600-650 empty drums that have not been rinsed being temporarily stored near the dedrum facility. It was recommended that plastic be spread on the ground to prevent any spillage. Dedrumming resumed at 1900 hours after 24-hour hiatus. • August 4--Drum rinse sampling procedure changed to obtain samples from all stations uniformly. Personal samples from pump operator inside dedrum facility eliminated because of low concentrations measured. • August 5--Dedtumming completed 2100 hours. Land-based monitoring schedule for interim period submitted to TRCO, Improved procedures for sampling of drum rinse were suggested by BCL and accepted by the Air Force. 1 t .. i I !. i .. 67 - -:. a • -----------------------------. • 4d • .

• August 6--M/V Vulcanus departed 0830 hours. Dedrum crew began rinsing and crushing 648 drums from temporary storage. • August 11--All drums have been rinsed and crushed. Lab work load adjustments discussed with TRCO to permit analysis of wipe samples from ship at close of program. • August 15--Results of first load lab analyses submitted by BCL to Air Force. Drum rinse quality control program improvements were brought up again. In the course of conducting tomato plant bio-assay studies, it was found that the plants uniformly were wilting due to the extreme evapotranspiration. The problem occurred because the pots, as provided, were too small and the peat potting medium lacked the necessary water holding capacity. • August 16--Suitable volcanic mineral soil was added to the potting medium. The surface of the soil was covered with aluminum sheets to reduce evaporation. The plant wilting was eliminated. The previously damaged plants were replaced. • August 17--All air, water, and biological observation schedules were reinstated. Drum rinse sampling monitoring was initiated preparatory to the second burn. Tomato plants downwind of the dedrumming facility continued to be affected by the herbicide. It was suspected that the vaporization of the HO from rows of crushed drums compounded chis phenomenon. The bed of the truck used to haul tomato plants and equipment was found to be contaminated with HO. The bed was replaced immediately with clean materials. The loading of M/V Vulcanus began at 1300 hours. Continual spill reconnaissance was initiated. • August 18--The industrial hygiene consultant notified the Air Force that some civilians were smoking adjacent to loaded R-5 refuelers. The operations officers were notified that no smoking materials or food should be taken into the dedrumming facility. Appropriate actions were taken to prevent future occurrences. I ' J_, 68 l- / - .,., ___ ....... , .... ___ .J. _ , , ...,.... ~ · - ·- - - - - • -- ---- -- J ..,..., C •• C •

• August 19--A brownish plume was observed and photographed, as the M/V Vulcanus was pumping ballast while berthed at the wharf. A grab water sample was taken near the stern of the ship. Dedrumming and ship loading was suspended at 0600 hours. No marine ecology impacts were observed. Fish were noted swimming in the area of the deballast plume. The potable water intake was closed during the deballasting operations. • August 20--Slight water discolaration still existed between the M/V Vulcanus and the wharf. • August 21--Battell's analytical laboratory reported that the deballast sample results had several peaks and without further dilution studies, they reported that they could not state the levels of 2,4,5-T or 2,4-D in the grab sample. Dilution studies and a rerun of the sample was requested. • August 22--Deballast results were submitted to the Air Force. • August 24--A ground water sample taken from a bore hole in the barrel yard storage area smelled strongly of herbicide orange. It was highly probable, due to the lack of a berm, that the surface contamination entered the bore hole or observation well. The post-operational monitoring program was begun. • August 27--Numerous bird species were observed and surveyed on Akau, Hikina, and Sand Islands of the Johnston Atoll.(All were in apparent good health except birds with broken wings that had flown into antenna guywires.) Abundance and type of fish species were noted in the wharf area. No marine ecological stress was evident. • August 28--Plant species on Johnston Island were surveyed. There was no evidence of native plants being affected by the Orange Herbicide disposal operations. This was the last day of post-operations! monitoring. 69 - i - ............. -- a C a au C C W2 -

3.2.3 Physical Monitoring Sampling Protocol 3.2.3.1 Chemical Samplin§ a. Air (1) Equipment and Procedures, In order to assess the impact on the air environment due to the possible presence of the N-butyl esters of 2,4-D and 2,4,5-T and the dioxin, TCDD, two methods were employed. Air sampling for 2,4-D and 2,4.5-T was accomplished utilizing Chromosorb 102 as an adsorption medium, a granular polymer well suited for collection of chlorinated hydrocarbons. This material was packed in micropipet tubes which were then wrapped in aluminum foil and stored in rubber stoppered test tubes (Plate 6). In order to sample a volume of air of about 150 liters, a flow rate of 0.50 liters/minute for a period of about five hours was required. A good adsorption efficiency could be obtained at this flow rate. A five hour sampling time was adopted which corresponded to the. length of one-half shift. This sampling procedure for the operations area avoided interruptions when the shifts were breaking for meals. The sampling apparatus consisted of an MSA Model G Personnel Sampling Pump mounted on top of an upright clean 55 gallon barrel for all ambient stations. The chromosorb tubes were connected to the pumps with Tygon tubing or, for the samplers worn by workmen where greater flexibility was desirable, latex rubber. In order to minimize the likelihood of rainwater contamination, the tubes were attached so that the opening to the tube would face downward. The pumps at the ambient stations were maintained on constant "high" recharge throughout the period, regardless of whether or not the pump was in use. The pumps worn by workmen were battery powered for the five hours. These pumps were then recharged in one of the sample-preparation rooms in Building 190 during the next half-shift. 70

Flow rates were checked at hourly intervals with a rotameter and adjusted to ensure that the 0.50 liter/minute flow was being maintained. In only a few instances did the pumps fail to maintain the desired flow. Air sampling for TCDD was accomplished utilizing benzene as the absorption medium. The apparatus consisted of a train of four impinger columns, the first two filled with 350 and 250 ml of benzene, respectively, and the final two with activated carbon (Plate 7). Activated carbon was used to adsorb the vaporized benzene from air flow through the first two columns. The benzene columns were wrapped with aluminum foil to avoid photo-decomposition of the TCDD in the sample. Following the carbon columns, a paper filter was attached with Tygon tubing to prevent any carbon from entering the pump. The pumps were operated directly off the 110-volt AC lines located at the sampling stations. The entire impinger train with pump was mounted on the same barrels as the MSA pumps at each station. As with the chromosorb apparatus, the flow rate through the impinger was periodi- cally checked using a rotameter and adjusted as necessary at a bleeder valve. A rate of 1.0 liter/minute was chosen; however, this rate may have been in error by as much as 20 percent, as variability in the pumps' speed and the effect of increasing amounts of saturated carbon caused fluctuations in flow. The established running time of five hours was about the maximum duration for maintaining flow without saturating both columns of carbon, which would result in a benzene breakthrough. About halfway through the study, it was found that the columns were becoming saturated after about 4 hours. As a result, the procedure was modified such that the last column filled with saturated carbon was removed and replaced with a column filled with fresh carbon during the sampling period. This enabled the entire half-shift to be represented as well as to provide a larger sample volume. Reasons for the more rapid adsorption rate are speculative, but it is j .• believed that the carbon used in the second half of the study was of lesser quality 71 • a D . , . , . .. . •

Samples were removed from the sites with the entire impinger trains intact within wooden holders. The benz~ne was drained into brown glass jars in one of the sample preparation rooms of Building 190 (Figure 3). The glassware was then rinsed once with benzene into the sample containers to collect any portions that my have adhered to the impinger walls. The samples were stored in a dark, cool room in Building 190 before being packed for shipment to the Occupational and Environmental Health Laboratory at Kelley Air Force Base for later TCDD analysis. Prior to reuse in the field, the impinger glassware had three acetone rinses followed by one rinse with benzene. (2) Air Sampling Sites. Four areas were sampled for the N-Butyl esters of 2,4-D and 2,4,5-T and TCDD. These were: (a) the dedrumming facility, (b) a position 310 feet west of this facility, (c) the wharf where the M/V Vulcanus was docked, and (d) the weather station. Figures 17, 18 and 19 show the locations of the air sampling sites. The remaining three areas (b), (c), and (d) were ambient 9ites. Each station was characterized by an impinger and chromosorb apparatus placed upon clean, 55 gallon drums. Site (a) Air inside th~ dedrumming facility was sampled to allow for a comprehensive ,industrial hygiene report. An impinger was located on a clean barrel at the southwest corner of the shelter for TCDD detection. In order to obtain workmen's exposure to 2,4-D and 2, 4, 5--T, persons working inside che facility in close contact with the herbicide were required to wea.'."' an MSA pump around the waist with a chromosorb tube attached near the breathing zone. When a workman wearing a sampler would leave the area to take a break, the samplers were turned off preventing such potential contaminants as cigarette smoke from being drawn into the sample. This procedure assured the detected concentrations to be representative of that inside the facility. As a further precaution, most of. the chromosorb tube was left wrapped in aluminum foil to minimize contact of the outer portion of the tube with the herbicide, a possible route to 72 Ls ffi a • .. •

I I, I, 15 i } :I !"' Cll .~ a::: "'l!Ij lzl ;:: f-< :::; H Cll ."' : ~ c., z i I .• LI "' ~ ' C □ H ...:I ~ Cll c:,: H < ,.._. ~ ; c., H r,.. ! i I g 5 ! l, ! I I m ~ 6a !11 I 73 ,: ! 1 .. . :..,,~ .... .;_·......~- :~:,-,.-1,-~· a ED 2 ..,-WU $ C C O • a -1 a

·,,. i~ ~i ~~--~ - . I ty ~/ l {t ---::::::::.,__~. ,,i ~!..- ~_:::::---~ D-1 -~~ ~~ ... . ~ ~~~ ·g ~~~~.I~ ~ ~~~~.~~ ~~~~~~--~ ~--~ *~:.~ S;~~~~~~~ c,..... - . ~ ... . ~ ~- ~~~~'! ~ • ~, ~~~/ -6 ~~ - I +o-s ._. I'..~/ -t- U-11 I!] AIR MONITORING SITES EXPANDED IN FIGURE 19 * TOMATO MONITORING SITES + SUPPLEMENTAL TOMATO SITES (POSTOP) FIGURE 18. AIR MONITORING SITES, DRUMYARD ~

Pump OperatorJ 0 Refueler RS pp r----- N Cl) 9,.. § Cl) 0 FIGURE 19. SAMPLING SITES AT DEDRUM FACILITY -- 75 l ~J L J&tsaz,cuovs • f

contamination of the sample. Once the sampling duration was complete, the chromosorb tube was rewrapped in clean foil and sumitted to the lab. The tubes were then cut inside the laboratory and the lower contaminated portion of the tube discarded before removal of the Chromosorb 102 granules. Because the pumps were turned off during breaks and some time was required for the crews to dress and undress during each half shift, the five-hour sampling time could not be achieved. In most instances, however, a sample volume of at least 100 liters was obtained at the 0.50 liter/minute flow rate. In addition to the impinger and personnel samples, chromosorb samples were taken on occasion at two western (downwind) corners and at the center of the eastern wall of the dedrumming facility. Most of these were taken during inoperative periods, when crews were not dedrumming the herbicide. Site (b) Located 310 feet west of the dedrum site, the downwind site was chosen to assess the affects of the barrel storage area, dedrumming the herbicide, and other operations on the air environment of this area. A comparison of the ambient levels at this station with observed tomato plant damage was possible due to the proximity of the plants with respect to the site. The downwind station was located near the crushed drum storage area (to the south), the contaminated wood stockpile (to the southwest), and the wind recording station with anemometer. The effects of the crushed drum storage agea and the wood stockpile on detected concentrations at the downwind station was minimal due to the constance of the wind from perpendicular to opposing directions. The proximity of the anemometer with the station allowed a close correlation with immediate wind directions and speeds. 76 . • - · ... -- .... __......... - -------------------- '· '.:~ ,.;~,..;,~': a au a e 0. a

I :I Site (c) A third air sampling station was established on the wharf CI at the western most light pole, approximately 300 feet \ ,j I from the truck-to-ship pumping station. Although winds were usually slightly out to sea with respect to this area, the position of the station does allow for an assessment of the ship's presence and pumping operations on the ambient air levels of the land adjacent to the wharf. Site (d) The fourth site, located at the weather station, was utilized for measuring the air background levels and was far upwind of all operational areas. ) (3) Air Sampling Intervals. (a) Preoperational. Air sampling was cnnducted for a three-day period (July 24 to July 26) before dedrumming ope, ;ions commenced for the purpose of ~stablishing baseline for the stud .• Benzene and chromosorb samples were run daily at the weather station, wharf, and downwind sites. Additionally, three benzene samples (one/day) and three chromosorb samples (all on July 26) were run inside the dedrumming facility. These samples were representative of the late-morning, early-afternoon hours. (b) Operational. Air sampling during dedrumming and associated operations commenced on July 27 and lasted through August 5 for the first loading of the M/V Vulcanus. The second loading took place over the interval August 17 through August 23. Generally, sampling during operations was limited to the five-hour half-shifts of the morning and evening. From the study performed at Gulfport, it was learned that the time of day had little effect on concentrations detected in the field. Nearly constant climatic conditions suport this idea for Johnston Island. A total of 120 valid chromosorb samples were taken at the four areas of study during the two operational intervals. Their distribution is shown below. • Weather station - 22 • Wharf 18 • Downwind station - 26 • Personnel samples - 43 • Comers of dedrum - 11