vvEPA
United States
Environmental Protection
Agency
Office of
Radiation Programs
.Washington DC 20460
EPA 520/1-82-009
May 1982
Radiation
of a
the
Site
-------
EPA REVIEW NOTICE
This report has been reviewed by the Office of Radiation Programs,
U.S. Environmental Protection Agency (EPA) and approved for publication.
Approval does not signify that the contents necessarily reflect the
views and policies of the EPA. Neither the United States Government
nor the EPA makes any warranty, expressed or implied, or assumes any
legal liability or responsibility for any information, apparatus,
product or process disclosed, or represents that its use would not
infringe on privately owned rights.
-------
EPA 520/1-82-009
BNL 51102
ANALYSIS AND EVALUATION OF A RADIOACTIVE WASTE PACKAGE
RETRIEVED FROM THE ATLANTIC 2800 METER DISPOSAL SITE
BY
P, COLOMBO, R,M, NEILSON, JR, AND M,W, KENDIG
NUCLEAR WASTE MANAGEMENT RESEARCH GROUP
DEPARTMENT OF NUCLEAR ENERGY
BROOKHAVEN NATIONAL LABORATORY
ASSOCIATED UNIVERSITIES, INC,
UPTON, NEW YORK 11973
SEPTEMBER 1979
REVISED MAY 1982
THIS REPORT WAS PREPARED AS AN ACCOUNT OF WORK SPONSORED
BY THE UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
UNDER INTERAGENCY AGREEMENT No, EPA-IAG-D6-0166
PROJECT OFFICER
ROBERT S, DYER
OFFICE OF RADIATION PROGRAMS
U,S, ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D,C, 20460
-------
Table of Contents
Page
Foreword x
Summary *i
i
I. Introduction 1
II. Retrieval, Storage, and Transport of the 80-Gallon Radioactive
Waste Package 2
A. Retrieval Operation 2
B. Storage and Transfer 3
III. Analysis of the Waste Form 4
A. Description of the Retrieved Waste Package 4
B. Radiography 5
C. Concrete Coring 6
D. Radiochemical Analysis 9
E. Concrete Integrity 13
IV. Corrosion Analysis of the Waste Container 16
A. Visual Inspection 16
B. Corrosion Rates as a Function of Position 18
C. Microscopic Examination of Local Effects 20
D. Chemical Analysis 24
V. Conclusions 25
References 27
List of Tables
1. Major U.S. Radioactive Waste Disposal Sites'1' 29
2. Analysis of Liquid Found in the Inner Steel Container and
Bottom Seawater 30
3.' Core Specific Activity - Curies/Gram ± a (%} 31
4. Cement Specific Activity - Curies/Gram ± a (%} 34
-m-
-------
List of Tables (Cont'd)
Page
5. Cesium-137 Content in Annular Volumes of the Concrete Waste
Form as a Function of Depth 37
6. Concrete Core Compressive Strength 38
7. Mild Steel Corrosion Rates in Ocean Environments 39
8. X-Ray Diffraction Identification of Surface Scrapings 40
9. Trace Element Analyses of Container Materials 41
List of Figures
l(a). Approximate Location of the Atlantic 2800 m Radioactive
Waste Disposal Site.C2) 42
l(b). Location of the Retrieved 80-Gallon Radioactive Waste Package. 43
2. Schematic Diagram of the Hoist System Used for the Retrieval
of the Waste Package from the Ocean Floor at a Depth of
2783 Meters.(2) 44
3(a). Open End of Radioactive Waste Package Immediately after
Surfacing from the Atlantic 2800 m Disposal Site. 45
3(b). Side View of Radioactive Waste Package Prior to Being Brought
Aboard. 46
4(a). A Modified H47 Jet Engine Container Used for the Encapsulation
and Shipment of the Retrieved Radioactive Waste Package. 47
4(b). An Open H47 Jet Engine Container Showing Rubber Faced Interior
Clamp Rings Provided to Hold the Radioactive Waste Package
Against Shock and Vibration. 47
5. Surface Markings on the Exposed Concrete Face of the Waste
Package. 48
6. Orientation System Used to Describe the Waste Package. 49
7. Montage of Waste Package Radiographs. The Top of the Figure Shows
the Flanged End of the Internal Container Which Was Located
Approximately 5.5 Inches from the Open End of the Container
(on the right edge in this figure) Which Resulted from Implo-
sion During Descent in Sea Disposal. Film was Positioned.
Along 270° Axis. 50
-iv-
-------
List of Figures (Cont'd)
Page
8. Radiograph of the Flanged End of the Internal Container. 51
9. View of the Concrete Waste Form with the Steel Drum Removed
(00 axis). 52
10. View of the Concrete Waste Form with the Steel Drum Removed
(90° axis). 53
11. View of the Concrete Waste Form with the Steel Drum Removed
(1800 axis). 54
12. View of the Concrete Waste Form with the Steel Drum Removed
(270° axis). 55
13. Core Drilling of the Concrete Waste Form. The Worker on the
Right is Moving the Drill Bit into the Waste Form While the
Worker on the Left Holds the Pneumatic Chisel Used to
Remove the Steel Drum. 56
14. Letter Designation of Core Locations as Shown Along the 0°
and 200 Longitudinal Axis. 57
15. Letter Designation of Core Locations Along the 90° Longitudinal
Axi s . 58
16. Letter Designation of Core Locations Along the 180° Longitudinal
Axis. 59
17. Letter Designation of Core Locations Along the 270° Longitudinal
Axi s . 60
18. Schematic Showing the Letter Designation of Core Holes as
Related to (e,x,r) Location Coordinates. 61
19. Closed End of the Waste Form after Removal of the Five Inch
Thick Initial Concrete Pouring. 62
20. Waste Form after Removal of the Upper Eight Inches of Concrete
from the Open End, Exposing the Flanged End of the Inner
Container. 63
21. View of the Exposed Flange End of the Inner Container. Note the
Gap Between the Container and the Concrete Formed by Implosion
of the Inner Container Wall During Descent. 64
-v-
-------
List of Figures (Cont'd)
Page
22. Concrete Waste Form after Removal of the Inner Container. 65
23. Inner Steel Vessel after Removal from the Waste Form. Note
Implosion of the Container Walls Along the 0° Longitudinal
Axi s. 66
24. Inner Steel Vessel Viewed Along the 180° Longitudinal Axis. 67
25. Inner Steel Vessel with the Cover Removed, Showing the Enclosed
Wound Filter Assemblies. 68
26. Close up View of the Inner Container Cover, Flange and Enclosed
Filter Assemblies. 69
27. Cesium-137 Content in Concrete Cores Along the 0° Longitudinal
Axis. 70
28. Cesium-137 Content in Concrete Cores Along the 90° Longitudinal
Axis. 71
29. Cesium-137 Content in Concrete Cores Along the 180° Longitudinal
Axis. 72
30. Cesium-137 Content in Concrete Cores Along the 270° Longitudinal
Axis. 73
31. Cesium-134 Content in Concrete Cores Along the 0° Longitudinal
Axis. 74
32. Cesium-134 Content in Concrete Cores Along the 270° Longitudinal
Axis. 75
33. Cobalt-60 Content in Concrete Cores Along the 0 Longitudinal
Axis. 76
34. Cobalt-60 Content in Concrete Cores Along the 90 Longitudinal
Axis. 77
35. Cobalt-60 Content in Concrete Cores Along the 180° Longitudinal
Axis. 78
36. Cobalt-60 Content in Concrete Cores Along the 270° Longitudinal
Axis. 79
-VI-
-------
List of Figures (Cont'd)
Page
37. A Map of the Surface of the Container. 80
38. The Entire Length of the Container. 81
39. A Schematic Illustrating Specific Features of the Container
as Seen in Figure 38. 82
40. An Exterior View of the Upper Portion of the Container - Sediment
Facing Side. 83
41. An Exterior View of the Upper Portion of the Container - Sea
Facing Side. 84
42, An Exterior View of the Mid-Section of the Container. 85
43. An Exterior View of General Attack Adjacent to a Chime within
the Mid-Section. 86
44. Sediment Side Perforation Adjacent to a Chime as Viewed from the
Inside of the Carbon Steel Sheath. 87
45. Sea Side Perforation Adjacent to a Chime as Viewed from the Inside
of the Carbon Steel Sheath. 88
46. The Interior Surface of the Carbon Steel Sheath. 89
47. The Concrete Waste Form. The Upper Portion of the Form is to
the Right in the Photograph. 90
48. Attack Adjacent to the Weld in the Upper Container. 91
49. Attack Adjacent to the Weld in the Lower Container. 92
50. Macroscopic Pits Covering the Carbon Steel Surface at the Closed
End of the Container. 93
51. The metal end of the Container. 94
52. An Example of Filiform Corrosion. 95
53. The Sheath Thickness vs. Container Position. 96
54. A Typical Metallographic Cross Section of the Upper Container. 97
55. A Typical Metallographic Cross Section of the Lower Container. 98
-V11-
-------
List of Figures (Cont'd)
Page
56. Attack within a Rim Fold. 99
57. Attack at a Chime. 100
58. The Coated Interface. 101
59. A Pit Formed within the Coated Region of the Mid-Section of
the Container. 102
60. Scanning Electron Micrograph of the Disbonding of the Inter-
facial Oxide. 103
61. Micrograph of the Attack upon the Upper Container. 104
62. A Perforation Formed at the Upper Container Sheathing. 105
-V111-
-------
Foreword
The Environmental Protection Agency was given a Congressional mandate
to develop criteria and regulations governing the ocean disposal of all forms
of wastes pursuant to Public Law 92-532, the Marine Protection, Research and
Sanctuaries Act, Within this Congressional mandate, EPA has initiated a
specific program to develop these regulations and criteria to control the
ocean disposal of radioactive wastes.
EPA has taken an active role both domestically and within the inter-
national nuclear regulatory arenas to develop the effective controls necessary
to protect the health and safety of man and the marine environment. The EPA
Office of Radiation Programs first initiated feasibility studies to determine
whether current technologies could be applied toward determining the fate of
radioactive wastes dumped in the past. After successfully locating actual
radioactive waste disposal containers in the disused dumpsites, the Office
of Radiation Programs has developed an intensive program of site-characteri-
zation studies to look at the biological, chemical and physical characteristics
and the presence and distribution of radionuclides within the sites, and is
conducting a performance evaluation of past packaging techniques and materials.
During the 1976 Atlantic 2800 meter radioactive waste disposal site survey,
the first recovery of a radioactive waste package from a radioactive waste
disposal site was performed by the EPA Office of Radiation Programs. Under
Interagency Agreement Number EPA-180-D6-Q166, Brookhaven National Laboratory
has performed container corrosion and matrix leach rate and degradation
studies on the recovered radioactive waste container. This report presents
the results of laboratory analyses performed.
These studies have helped to focus attention on the problems associated
with past and present nuclear waste disposal activities concomitant with the
growing national and international concern for the long-term effects of this
low-level waste disposal option.
-IX-
-------
Acknowledgment
The authors wish to thank R.S. Dyer of the U.S. Environmental Protection
Agency for the many helpful suggestions and discussions during the course of
this work.
The assistance of W.M. Becker, W. Vogel, and A.E. Lukas in performing
the numerous tasks associated with the ocean retrieval and subsequent core
sampling of the radioactive drum is acknowledged with thanks.
The authors also wish to acknowledge the assistance of A.J. Weiss,
L. Milian, and S. Garber in the preparation and performance of chemical and
radiochemical tests on the concrete core specimens.
Also contributing was J.R, Weeks, who furnished guidance on corrosion
studies for the metal container.
-x-
-------
Summary
On July 31, 1976, an 80-gallon radioactive waste package was retrieved
from the Atlantic Ocean 2800 meter depth disposal site. This site which is
centered at coordinates 38°30'N, 72°06'W is located approximately 120 miles
(190 km) east of the Maryland-Delaware coast. The radioactive waste package
was transported to Brookhaven National Laboratory where container corrosion
and matrix leach rate and degradation studies were conducted.
The retrieved waste package comprised an eighty gallon mild steel drum
with no lid which contained a concrete waste form. Markings on the con-
crete surface indicated that it had been disposed in 1961. Within the con-
crete matrix, a sealed steel vessel was found which contained some liquid
and three wound filter assemblies. Although this vessel had a major inden-
tation running along its length resulting from the pressure differential during
or after descent to the sea floor, it had not leaked. The integrity of the
concrete matrix had not degraded appreciably during fifteen years in the dis-
posal environment as evidenced by visual observation, weight loss, and com-
pression strength measurements. A conservative estimate indicates that it
would require a minimum of 300 years in this ocean environment before the
waste form would lose its integrity and provide no barrier to activity release
due to cement phase dissolution. Radiochemical analysis indicated the pre-
sence of cesium-137, cesium-134 and cobalt-60 in both the concrete matrix and
the inner vessel. Based on the measured cesium-137 distribution in concrete
core samples, an average cesium-137 release rate of 3.7% per year was calculated,
Corrosion rates for general attack on the upper portion of the 80-gallon
mild steel drum were 0.0013-0.0019 in/yr (0.032-0.049 mm/yr) assuming a con-
stant rate with no induction period. A lower limit for the rate of local
pitting corrosion of 0.0026 in/yr (0.067 mm/yr) was determined. Based on
observations after 15 years in ocean disposal, general thinning attack appears
to be the most important process. Using the range of general attack rates, an
18 gauge (nominal 0.0476 inch thickness) mild steel drum would require 25-37
years in this disposal environment before corrosion would cause the container
to lose its effectiveness as a barrier to activity migration.
-XT-
-------
I. INTRODUCTION
Sea disposal of low-level radioactive waste began in the United States
in 1946, under the licensing authority of the Atomic Energy Commission (AEC).
In 1962, the first commercial land disposal site was licensed in Beatty,
Nevada, and as land disposal operations expanded, ocean dumping was phased
out in 1970.
Most of the radioactive wastes were packaged in used 55 gallon drums
filled with concrete so that the average package density was sufficiently
greater than sea water to assure sinking. Generally, the drums were capped
with "clean" concrete and were disposed without tops.
Although radioactive wastes were dumped in various areas of the Pacific
and the Atlantic Oceans, three deepsea disposal sites received the majority of
wastes dumped between 1946 and 1959. The location of these sites and the
estimated amounts of waste disposed are given in Table 1.
Recently, there has been renewed interest in sea disposal as a waste
management alternative to land burial of low-level radioactive wastes. The
U.S. Environmental Protection Agency (EPA) has been designated to establish
controls governing ocean disposal of wastes. To develop ocean disposal regu-
lations, it is considered important to assess past packaging techniques and
to determine the effects of ocean environments on the waste packages. In
1974 and 1975, EPA conducted survey studies at the Pacific-Faralions 900 m
and 1700 m depth sites and the Atlantic 2800 m depth site' ' using submersibles
to locate and identify radioactive waste packages.
As part of this effort, Brookhaven National Laboratory (BNL) conducted
container corrosion and matrix leach-rate and degradation studies on an
80-gallon radioactive waste package retrieved from the Atlantic 2800 meter
radioactive waste disposal site in July 1976.
This report includes analytical methods, results, conclusions, and
preliminary estimates relevant to the expected life of the metal container -
concrete matrix packaging system in the ocean environment.
-1-
-------
II. RETRIEVAL, STORAGE, AND TRANSPORT OF THE 80-GALLON RADIOACTIVE WASTE PACKAGE
On July 31, 1976, an 80-gallon radioactive waste drum was retrieved from
the Atlantic 2800 meter disposal site, centered at coordinates 38°30'N,
72°06'W. This site, Figure 1(a) is located approximately 120 miles (190 km)
east of the Maryland-Delaware Coast, and occupies an area of 98.8 square
miles (256 km2).
A. Retrieval Operation
The retrieval operation was a coordinated effort involving the R/V Lulu,
support ship for the deep submersible ALVIN, the DSV ALVIN, and the R/V Cape
Henlopen, escort ship for the ALVIN underwater survey operations.
During the morning of July 29, 1976, ALVIN located a clearly labeled
80-gallon radioactive waste drum, deemed suitable for recovery, at a depth
of 9131 ft (2783 m). The location of the drum is shown in Figure l(b).
The actual drum retrieval operation required a specially designed drum
attachment device, a computerized navigation system, and a synthetic lift
line. Figure 2 shows a schematic diagram of the hoist system. The plan was
to position a lift line close enough to the waste package to allow ALVIN to
attach them together for retrieval aboard the R/V Cape Henlopen.
A 1500 Ib. clump anchor and a transponder were lowered by the synthetic
Kevlar line* to a depth approximately 50-100 meters from the ocean bottom in
close proximity to the waste package. At this point, the R/V Lulu was able
to guide the R/V Cape Henlopen towards the container by monitoring the trans-
ponder position with an acoustic array previously placed on the ocean bottom.
(3)
The details of this procedure have been described previously/ ; During the
early morning of July 30th, the clump anchor was positioned on the ocean
floor 295 ft (90 m) from the waste package. To maintain the position of the
anchor, the end of the Kevlar line was buoyed off to two-48 inch diameter
floats.
* Manufactured by Philadelphia Resin, Inc.
-2-
-------
Immediately after positioning the anchor, ALVIN dove carrying the drum
grab mechanism and a 100 m length of 1 inch diameter nylon line. ALVIN
attached the grab to the drum, and then ran the 100 m length of nylon line
between the drum and the clump anchor. The clump anchor was detached after
one end of the line was attached to the grab and the other to the lift line
below the transponder in preparation for the final hoist procedure.
The retrieval procedure started aboard the Cape Hen!open at 0200 on
July 31st with the recovery of the two 48-inch floats. At 0800 the waste
package surfaced and the radiation level was monitored before being lifted
aboard, as shown in Figures 3(a) and 3(b). Once aboard it was carefully
documented photographically, sampled for corrosion products and biological
growth, and finally sealed in a shipping container, which was flushed with
dry argon. The elapsed time from the moment the waste package broke the
surface until it was sealed in the shipping container was two hours.
B. Storage and Transfer
The container used for storage and transfer of the drum to BNL was de-
signed for the shipment of H47 jet engines. It is a hermetically sealed
cylinder of welded heavy gauge steel construction having outside dimensions
75 inches long x 40 inches wide x 43 inches high, as shown in Figure 4(a).
The container stands horizontally on heavily reinforced legs and has a
gasketed, bolted flange which is on the plane of the central axis of the
cylinder. Thus, the top and bottom form half cylinders which are fastened
at the mid-section (Figure 4(b)). Hermetic sealing is certified at interior
pressures of 7.5 psia to 30 psia. The container was modified to provide
access for introducing an inert cover gas (argon) and removal of the initial
air atmosphere. The purpose for inert gas was to remove oxygen and thus
minimize corrosion of the metal drum in transit. The oxygen content in
the sealed container was reduced to ^ 0.5% by successive evacuation (-5 psig)
and argon pressurization (+1 psig).
A BNL health physicist aboard the R/V Cape Henlopen maintained radio-
active surveillance from the point of recovery in the Atlantic Ocean to
arrival at BNL.
-3-
-------
III. ANALYSIS OF THE WASTE FORM
A. Description of the Retrieved Waste Package
The retrieved waste package* comprised an eighty gallon mild steel drum
filled with cement concrete. The waste form was 50.1 inches (127 cm) in
length with a diameter of 22.9 inches (50.2 cm). According to identification
markings stamped into the lower drum head, the 55 gallon drum was made from
18 gauge (nominal 0.0476 inch thickness) steel and was manufactured in June
1959. The drum did not have a lid and that portion of the exposed drum lip
which was not in intimate contact with the concrete was almost completely
separated from the drum body by corrosion. Markings scratched into the ex-
posed top surface of the concrete are shown in Figure 5. These markings
indicate that the Army Chemical Corps was the source of this waste package
which had a radiation dose level of 40 and 3 mr/hr at the surface and at one
meter, respectively, at the time of disposal. The surface dose rate measured
after recovery ranged from 0.1 to 4.0 mr/hr with the highest reading at the
open end portion of the drum buried in the sediment. Background levels were
measured at a distance of one meter. The drum was designated package 28,
dated 1961, and weighed 1682 pounds at the time of disposal. The markings
also seem to indicate that the package contains cobalt-60.
After sufficient time had been allowed for drainage after recovery, the
retrieved drum weighed 1600 pounds. This is an apparent weight loss of
approximately 5% since disposal. Weight loss may be attributed to several
factors: (1) dissolution of calcium hydroxide (a hydration product) and some
of the cement phase from the waste form, (2) loss of water by evaporation
during concrete curing, (3) erosion of the waste form in disposal and (4) the
accuracy of the initial weighing. Since the weight of the package was written
in the concrete surface, the concrete was apparently "wet" during weighing.
Because hydration of the cement phase requires substantial time (hours or days),
both hydration and evaporation compete during curing for water and appreciable
loss of unreacted water due to evaporation is possible.
*In this and subsequent discussion, waste form refers to the solidified solid
enclosed in the eighty gallon mild steel drum which is the waste container.
The waste package is comprised of the waste form and its container.
-------
Figure 6 is a schematic of the retrieved drum and the orientation system
used in describing the waste package. In this system, one point on the drum
circumference at the open end was arbitrarily assigned to be the zero degree
reference point. Looking at the open end of the drum, an angle (9) measured
clockwise about the longitudinal axis of the drum is used to describe any
,radius about this axis. In taking metal samples and concrete cores, coordinates
of any point in the waste form are described by (0, x, r) where x is the dis-
tance (in inches) along the longitudinal axis from the open end concrete
surface and r is the distance (in inches) along a radius from the circumferential
surface towards the longitudinal axis. The diameter of the waste package
(22.9 inches) is such that one inch along the circumference is equivalent to
a theta (0) of 5 degrees. Note in Figure 6 that the sediment line is indi-
cated for both the open and closed ends of the drum. Above these lines (as
indicated in the figure) the drum was buried in sediment at the time of
retrieval.
B. Radiography
Information supplied by the site where the waste was packaged indicates
that the 80 gallon drum was formed by welding one half of a 55 gallon drum
to the end of another 55 gallon drum to increase its length. The added
length was necessary to accommodate a sealed metal container used to encapsu-
late demineralized resin or filter material containing cobalt-60. This con-
tainer was centered in the 80 gallon mild steel drum and the surrounding space
was filled with concrete containing radioactive cesium and possibly cobalt-60.
The waste package was radiographed to determine the exact position of the in-
ternal vessel since knowledge of the location was necessary prior to proceeding
with the concrete coring operation.
The radiographs were produced by the Consolidated Testing Laboratories,
Inc., New Hyde Park, New York, at Brookhaven National Laboratory using a 45
curie cobalt-60 source. Two series of radiographs were taken along the length
of the drum; one in which the source was positioned on the drum circumference
along the 180 degree longitudinal axis with the film located along the zero
-5-
-------
degree axis and a second series with the source located on the 90 degree
longitudinal axis with the film along the 270 degree axis. The second
series, along an axis displaced 90 degrees from the first, was necessary to
define the location and shape of internal items. Fiducial markers were placed
on the drum to allow the subsequent positioning of one radiograph relative to
another. Initially, thirty minute exposure times were used to locate the
upper flange of the inner vessel. Subsequently, one hour exposures were em-
ployed to enhance detail. Figure 7 illustrates a montage of the radiographs
in which the film was positioned along the 270 degree longitudinal axis.
Figure 8 shows the flanged end of the inner container alone.
Interpretation of the radiographs by BNL personnel satisfied the prime
objective, the location of the inner vessel. This permitted the calculation
of the maximum depth cores that could be taken at various positions without
impacting the inner vessel. The radiographs indicated that the inner steel
vessel was approximately 39 inches (100 cm) in length, with an outside
diameter of 6.3 inches (16 cm) and a wall thickness of 0.25 inches (0.64 cm).
The flanged end was located approximately 5.5 inches from the open end of the
waste package. These values proved to be quite close to the actual dimensions
of the inner container measured after it had been removed from the concrete.
In addition to the inner container, the radiographs indicated the presence
of other objects, including two pipes located near the inner container. The
radiographs also indicated a large indentation along the longitudinal axis
of the inner container. This was ascribed to bending of the wall of the inner
container due to implosion during descent.
C. ConcreteCoring
After the eighty gallon steel drum was removed with a power chisel, the
concrete waste form was cored. Cores were taken to determine the type,
quantity, and distribution of contained activity in the waste form. In
addition, the presence of the core holes facilitated the subsequent removal
of the inner steel vessel.
-6-
-------
Figures 9-12 show sequential views (0°, 90°, 180°, 270° longitudinal
axes) of the concrete waste form after removal of the steel drum. Observa-
tion disclosed that two concrete pourings were made. An interface five inches
from the closed end (bottom) of the waste form delineated the two pourings.
Since the radiographs indicated that the non-flange end of the inner steel
vessel was also approximately five inches from the closed end of the waste
form, it was assumed that the first concrete pouring was made to position
the inner vessel that distance from the end of the waste form, probably for
shielding purposes. The integrity of the concrete waste form was good. Some
dissolution of the cement phase on the open end and on the sides of the waste
form near the open end had occurred leaving aggregate exposed. However, the
exposed aggregate still adhered to the waste form indicating the dissolution
of the cement phase was not extensive. Some deposits resulting from corrosion
of the steel drum were apparent on the portion of the waste form near the open
end. Circular markings in Figures 9-12 resulted during sampling of meta] from
the steel drum for corrosion studies.
*
Coring of the concrete waste form was performed using a Target concrete
hole saw with a dual speed motor (500/1000 rpm) on a swivel base. Impregnated
diamond core bits (2-1/4 inch diameter) were used to produce two inch diameter
concrete cores. Although water is normally used during concrete coring as both
a lubricant and to flush out coring debris, water was not used during the
coring of the retrieved waste form since the water potentially could remove
activity from the core and also create substantial volumes of contaminated
liquid waste. A commercial teflon spray lubricant was occasionally applied
to the outside of the coring bit. An attempt was made to take two inch long
cores from each core hole until the inner steel container was closely approached.
After a core hole was started, drilling would continue until the bit had pro-
gressed approximately two inches into the concrete. At this point, the
drilling was stopped and the coring bit backed out which usually broke off
the concrete core at the two inch depth. Drilling would begin again until
* Robert G. Evans Company, Kansas City, Missouri 64130
-7-
-------
the core bit had cut an additional two inches into the concrete (four inch
total depth) at which point drilling would stop and the two inch long core
of the concrete from a depth of 2-4 inches into the waste form was removed.
The coring operation was performed inside a hot cell (primarily for dust
control purposes) as shown in Figure 13. Cores were taken along the 0°, 90°,
180°, and 270° longitudinal axis as shown in Figures 14-17 respectively.
Although individual cores could be described by the (0, x, r) coordinates,
the core holes were also designated by a letter for ease of identification.
The center of any particular letter designated core hole was the same dis-
tance from the open end concrete surface (x) independent of the theta (9) for
that axis. Figure 18 gives the x distances for the letter designated core
holes. Cores were described as, for example, 90C6, which indicates the core
was taken along the 90° longitudinal axis, from core hole C (x = 10 inches)
and to a depth (r) of six inches. Since cores were normally two inches in
length, the material in this core was obtained at a depth of 4-6 inches. A
core designated OC7.5, however, had a length of 1-1/2 inches since drilling
for the prior core (OC6) stopped at a depth of six inches.
Core drilling was done slowly to prevent the bit from heating up ex-
cessively and thus releasing diamonds and also to provide an indication of
the presence of foreign objects before substantial damage occurred. During
coring, the presence of wide mesh wire, two steel pipes and a cavity which
occurred along the 0° longitudinal axis at a depth of 7.5 inches for much of
the length of the waste form were observed. A vacuum system was used to pick
up the dust and coring debris created.
After core samples were taken for activity analysis and compression testing,
the concrete waste form was dissected using a power chisel to free the inner
steel vessel. The progress of this dissection is shown in Figures 19-22. In
Figure 19, the five inch thick concrete pouring at the closed end of the waste
form has been removed. Note the presence of the two concentric rings of wide
mesh wire and the two steel pipes. At this point, work was continued at the
open end of the waste form. Figure 20 shows the open end of the waste form
with the upper eight inches of concrete removed, exposing the inner container
flanged end and the two steel pipes which traverse almost the entire length
-8-
-------
of the waste form. The upper flange is observed in Figure 21 at a different
angle showing the indentation which runs the length of the inner container on
the side which imploded. The concrete debris after removal of the inner con-
tainer are shown in Figure 22. The lower end of the steel pipe in the upper
portion of this photograph was cut off during concrete removal. Figure 23
is a photograph of the inner steel vessel after removal from the waste form.
This vessel had a mass of 96 Ibs (43.5 kg) with a length of 39 inches (99 cm),
and a body diameter of 6.5 inches ( 16.5 cm). The 0° marking on the edge of
the flange corresponds to the 0° longitudinal axis of the concrete waste form.
Note the implosion of the inner container along the 0° axis which occurred
during or after descent to the sea floor. Figure 24 shows the inner container
from the 180° orientation. The flange was removed from this container in an
enclosed area using non-sparking tools because of the possibility of hydrogen
pressurization due to radiolysis. No pressurization of the container was noted,
Three wound filter assemblies were contained within as shown in Figures 25 and
26. Metal samples cut from this vessel for corrosion studies indicated a con-
tainer wall thickness of 0.25 inches (0.64 cm). In addition, the vessel con-
tained 1.74 liters of liquid which was collected for subsequent analysis. The
cation constituents of this liquid were determined using a Perkin Elmer model
360 atomic absorption spectrophotometer. Sulfate content was measured with a
Technicon Autoanalyzer II. This composition is compared in Table 2 to the
composition of bottom seawater obtained in the vicinity of the retrieved waste
package. These analyses indicate that the liquid in the inner steel vessel
had significantly lower sodium, magnesium, and sulfate contents than the bottom
seawater and that these constituents were not present in the same ratios. As
such, it was determined that the liquid in the inner seal container was not
seawater.
D. Radi oc hemi ca1_Ana1ys i s
The concrete cores taken to determine the radionuclide distribution in
the waste form were dissolved in aqua regia prior to analysis. The aqua
regia was made by combining 3 parts of 12 M^ hydrochloric acid, 1 part of
16 M nitric acid, and 1 part of distilled water by volume. Distilled water
was added to inhibit the interaction of the two acids during storage. Since
-9-
-------
the concrete was composed of portland cement, sand, and quartz aggregate,
only the cement phase dissolved in the aqua regia. However, the activity in
the cores did go into solution since it is associated predominantly with the
cement phase and any activity on the aggregate is cleaned off during the dis-
solution. No activity was noted in any aggregate samples counted.
In the dissolution process, the concrete core to be dissolved was
weighed and then placed in a glass beaker to which aqua regia and a Teflon
stirring bar were added. The solution was stirred until the sample was dis-
solved and present in the solution as a suspended floe with the exception of
the aggregate which settled out on the bottom of the beaker. The solution
containing the floe was then poured into a 500 ml volumetric flask, which in
all cases exceeded the volume of the dissolved sample. The beaker and aggregate
were washed with additional aqua regia to remove any residual solution and/or
activity. This rinse was also added to the volumetric flask. Sufficient
additional aqua regia was then added to the volumetric flask to bring the
liquid level up to the calibrated volume. The liquid in the volumetric flask
was mixed thoroughly and a fifteen milliliter sample taken for analysis. The
fifteen milliliter samples thus obtained were placed into twenty plastic screw
cap polyethylene counting vials for analysis. It was observed that the sus-
pended floe solution was very stable, facilitating the removal of homogeneous
aliquots. The core aggregate was weighed after drying to determine, by sub-
traction, the cement mass in the core.
Core dissolution samples were analyzed using an Ortec coaxial Ge(Li)
detector. The detector was horizontally mounted with an integral FET pre-
amplifier whose signal was fed into an Ortec 472A spectroscopy amplifier.
The detector has an efficiency of 20% with a resolution of 2 keV at 1.33 MeV.
The energy spectrum was analyzed using a Tracor Northern** TN 1700 multi-
channgel analyzer in the pulse height analysis mode. A hardwired peak search
routine (All) was used for peak identification and peak area determination.
* Ortec, Inc., 100 Midland Road, Oak Ridge, Tennessee 37830
** Tracor Northern, 3551 W. Beltline Highway, Middleton, Wisconsin 53562
-10-
-------
Strontium-9Q was measured using a low-level beta counter to determine the in-
growth of yttrium-90 after a radiochemical separation of strontium. Plutonium
analysis was performed by alpha spectroscopy using a surface barrier silicon
detector after radiochemical separation and electrodeposition on to a disk.
Cesium-137, cesium-134, and cobalt-60 were found in the concrete cores
as indicated in Tables 3 and 4. Table 3 expresses the activity concentration
in terms of curies/gram of core mass, while the values in Table 4 indicate
the curies/gram of cement mass in the respective cores. The data in Table 4,
that is the activity concentration relative to the core cement mass, are
shown plotted as a function of position in the waste form in Figures 27-36.
Concentration in curies/gram of cement mass is believed to be more meaningful
than considering the entire core mass because the activity is associated with
the cement phase and the aggregate portion of the cores vary considerably
depending upon the location from where the core was taken. These figures
show that the cesium-137 (t^ = 30.2 years) concentration increases with core
depth and reaches an approximately equal level for the 4-6 inch and 6-7.5
inch cores depths. The cesium-137 concentration levels were highest in the
0° and 270° core orientations. Cesium-134 (t% = 2.1 year) was present in
-12
concentrations below 10 curies/gram cement in most cores. It was measured
above this level only in the 0° and 270° orientations and primarily in 4-6
inch depth cores. Since disposal (fifteen years), the cesium-134 has gone
through approximately seven half-lives. With no loss due to leaching, only
0.64% of the initially contained quantity could be present at the time of
analysis. Cobalt-60 (t^ =5.3 years) was found to be present in approxi-
mately equal concentrations for all orientations and core depths. The OH2
core was analyzed to determine its strontium-90 and plutonium content. This
12
core contained less than 2.4x10 curies/gram strontium-90 (limit of detection)
-13
and 6.7x10 curies/gram plutonium-239. These values should be evaluated re-
lative to background fallout levels.
The results from the concrete coring can be considered in terms of annular
volume elements. The average cesium-137 content of the concrete in annular
volume elements has been determined and is shown in Table 5. The thickness
of these volume elements correspond to the position of core series depths, i.e.,
-n-
-------
0-2, 2-4, 4-6, and 6-7.5 inches (this final volume element has a thickness
of 1-1/2 inches). While these elements only consider the concrete to a
depth of 7.5 inches from the surface, this represents 88% of the drum volume,
and the majority of the volume not considered was occupied by the inner
concrete container. An average volume weighted cesium-137 activity concen-
tration of 2.48 x 10" curies/gram was measured to a depth of 7.5 inches.
Since the density of the concrete waste form (neglecting the inner concrete)
3 5
is 2.21 g/cm , the sum of the concrete annular masses is 6.52 x 10 g,
-4
resulting in a total cesium-137 content of 1.62 x 10 curies at the time of
analysis. This number could be related to the original cesium-137 content
of the waste form (considering decay) to determine the radionuclide release
during disposal if the total initially contained activity were known. How-
ever, if the original activity distribution was homogeneous and the inner
container released no activity, an estimate of the minimum cesium-137 release
can be made. This is accomplished by noting that the activity concentration
increases from the outside to a depth of four inches and that for the 4-6 and
6-7.5 inch depth cores, the average activity concentration are approximately
equal with a volume weighted average of 5.65 x 10" curies/gram. As such,
leaching can be assumed to have removed activity only from the outer four
inches of the waste form and the constant activity concentration at core
depths of 4-7.5 inches represents the initial waste form concentration after
decay. Using a decay time of 15 years from waste form disposal to analysis,
cesium-137 decays to 70,9% of its original quantity (t^ = 30.2 years). This
suggests an initial homogeneous activity concentration of 7.97 x 10" curies/
gramoratotal waste form activity of 5.20 x 10" curies of cesium-137 at the
time of disposal. The calculated loss of activity (corrected to the time of
disposal) from the outer two volume elements to produce an activity concen-
-10 -4
tration of 7.97 x 10 curies/gram is 2.92 x 10 curies. The waste form
is calculated to have lost 2.92 x 10 curies of cesium-137 from a total
content of 5.20 x 10" curies (both corrected to the time of disposal). This
corresponds to a release of 56.2% of the cesium-137 contained at the time of
_3
disposal. The calculated bulk leach rate, Lg, of the waste form is 2.38 x 10
O
g/(cm -day) where Lg is defined by:
-12-
-------
4 = 2 an x m (1)
T* 3
_ n_ = cumulative fraction release of the species of interest
A (corrected to the time of disposal)
m = waste form mass, g
2
S = external geometric surface area, cm
t = cumulative time since disposal, days
Note that cesium is one of the most Teachable radionuclides in a cement
waste form. The release rates for other radionuclides, particularly cobalt-60
are typically appreciably lower. This calculation shows that for cesium-137
the release is dominated by Teachability and not waste form dissolution.
The inner flanged container when opened was found to contain 1740 mini-
liters of liquid. This liquid was counted on the Ge(Li) detector after the
addition of 24 ml of 12 M_ HCT and 2 ml of distilled water. The measured
specific activities in this liquid were 2.00 x 10" curies/ml Cs-137,
5.48 x TO"10 curies/mT Cs-T34, 2.37 x TO"9 curies/ml Co-60 and 1.47 x 10"8
curies/ml Sr-90.
During removal of the waste form from its shipping container at BNL,
approximately 7.5 liters of liquid was present in the bottom of the shipping
container. Upon analysis, this liquid was found to contain 3.53 x 10"
curies/ml Cs-T37, 1.55 x 10"11 curies/ml Cs-134, 2.4 x TO"13 curies/ml Sr-90
and 1.4 x TO curies/ml Pu-239. The one sigma counting uncertainty for
the plutonium analysis was ± 50%.
E. Concrete Integrity
Concrete cores were taken to evaluate the integrity of the concrete as
measured by its compressive strength using ASTM Standard C 39-72, "Method
of Test for Compressive Strength of Cylindrical Concrete Specimens." A
Soiltest* CT-710 compressive tester was used to make the compression strength
measurements. Table 6 lists the compressive strengths of the cores tested,
which averaged 1,710 psi. These cores had a diameter of 1.73 inches and
varied in length from 2.17-3.58 inches. The sample diameter is less than
*Soiltest, Inc. Evanston, Illinois 60602
-13-
-------
the minimum diameter typically used in compression testing (2 inches). Core
drilled samples for construction material compression strength verification
typically use larger core diameters (3-6 inches). Small sample diameter can
lead to misleading low compression strength values as the drilling operation
may introduce substantial imperfections such as surface microcracking, es-
pecially when water cooling and flushing is not employed during coring.
Inaccuracies can also occur if the size of the concrete aggregate approaches
the diameter of the core sample. The concrete was also tested using an impact
test hammer (Soiltest, CT-320). With this method, a weighted hammer is impacted
against the concrete surface and the hammer rebound measured. This rebound is
directly related to the compression strength. Rebound is expressed in terms of
compression strength by the use of conversion tables. Available conversion
tables are applicable to ordinary construction concrete. For other types of
concrete (that differ appreciably in composition in terms of type and quantity
of aggregate, cement and water), one must first establish the relationship
between rebound and compression strength to derive an appropriate conversion
table. Applying this method and construction concrete conversion tables to the
concrete waste form, an average compression strength of 4,100 psi was obtained.
As such, the core samples used in actual compression testing may have been ad-
versely affected during the drilling operation, although this can not be defi-
nitively verified since the original concrete composition is not known.
It is difficult to estimate if the strength of the concrete has decreased
as a result of ocean disposal since no control exists for comparison purposes.
Certainly, the concrete exhibits good integrity in that it has a reasonably
high compression strength and does not indicate appreciable mechanical degra-
dation such as exfoliation or cracking. The only visually observed degradation
consisted of a small amount of cement phase dissolution near the open end of
the waste form which exposed some aggregate. This aggregate, however, remained
bound to the cement matrix. Minimum average concrete compression strength
values are available as a function of water/cement weight ratio. ' A value of
4,300 psi is obtained with a water/cement weight ratio of 0.5 (4,900 and 3,800
for w/c = 0.45 and 0.55 respectively). This is a common concrete mix, and as
such, it may indicate no significant change in waste form compression strength
since disposal.
-14-
-------
Any deterioration of the concrete would most likely be attributed to sulfate
attack. Solutions containing sodium, calcium and magnesium sulfates can react
chemically with hydrated lime and calcium aluminate in cement to form calcium
sulfate and calcium sulfoaluminate. These reactions lead to an increase in
volume, accompanied by mechanical disruption of the cement phase. This ex-
pansion type of degradation, typically associated with on-land sulfate attack,
is not observed with cement in seawater. The cause for the lack of expansion
is not known. Instead, seawater sulfate attack, particularly by sodium or
calcium sulfate, is of a surface softening type. This is a localized attack
and will not progress unless the soft surface layer is removed (wave action,
abrasion, etc.). This may account for the cement phase dissolution near the
open end of the waste form. In any case, the concrete did not exhibit signifi-
cant deterioration as a result of sulfate attack.
The concrete waste form will lose integrity as the cement phase dis-
solves. The maximum weight loss that could be ascribed to cement phase
dissolution (hydrated silicate and aluminate compound and calcium hydroxide)
is 5% after fifteen years in disposal. As mentioned previously, other factors
may have contributed to this apparent weight loss. The most satisfactory ex-
planation for the weight loss is measurement error at the time of disposal.
As a result, the assumption of a 5% weight loss over 15 years due to cement
phase dissolution is conservative. Assuming a constant 0.33%/yr weight loss
due solely to cement phase dissolution, a period of 300 years would be re-
quired for the cement phase to completely dissolve. During this time, the
waste form would demonstrate a gradual loss of integrity. This estimate of
the time during which the waste form will retain at least some integrity is
conservative. The dissolution rate is likely to decrease substantially with
time as the relatively soluble calcium hydroxide is removed leaving the less
soluble hydrated calcium silicate and aluminate compounds which bind the sand
and aggregate together and are the primary contributors to concrete integrity.
Conversely, although a substantial decrease in the rate of weight loss is
expected as relatively soluble products are removed, some increase above this
low rate may occur as the waste container corrodes away exposing more of the
waste form surface area to attack.
-15-
-------
IV. CORROSION ANALYSIS OF THE WASTE CONTAINER
Corrosion analysis of the eighty gallon carbon steel drum was performed
using the experimental procedures recommended by NACE^ ' whenever practicable.
The particular analyses employed have attempted to (1) describe the qualita-
tive nature of the corrosive attack, (2) give a quantitative estimate of the
attack and (3) describe the chemical and metallurgical correlations to the
corrosive attack.
A. VisualInspection
The waste package was first removed for inspection from its storage and
transfer container on September 14, 1976. For the prior forty-five days
since retrieval, the waste package was stored in a dry argon atmosphere.
(A dry argon storage atmosphere was also utilized for subsequent intervals
between sampling for corrosion studies.) A photographic survey of the waste
container outer surface was performed at this time. The photographs included
alphabetic labels whose coordinates were recorded, in terms of x, r, and 0.
The coordinates (x, r and 0) have as their ordinate a point defined by the
intersection of the cylindrical axis and the plane of the rim at the concrete-
exposed end of the container. The coordinate, x, is the distance parallel
to the cylindrical axis running into the container. Looking at the concrete
end, e, is defined as degrees of clockwise rotation about the cylindrical
axis. This coordinate system is shown schematically in Figure 18. A map of
the container surface is illustrated in Figure 37. In this figure, the con-
tainer is shown rolled out into a plane after cutting along the 180° longitu-
dinal axis. Figure 38 is a photograph of the waste package while Figure 39
is a schematic illustrating specific features of the waste container. The
sediment line lies along FK in Figure 39. The container was constructed
from two cylindrical carbon steel drums welded at x = 15-1/2", the weld line
being along CD. The container has three chimes, AB, GH, and IJ.
The upper portion of the container, between x = 0 and x = 17-1/3" (EF)
appears, from visual examination, to be attacked more severely than the re-
maining portion of the container. At the time of examination, much of this
upper region contained loosely adhering material covering a dark red to black
-16-
-------
scale directly adhering to the metal. Figures 40 and 41 show the severe
attack as it appears on this upper portion of the container. The sediment
side of this upper container region appears in Figure 40. Here much of the
loosely adherent material is absent, thereby exposing the red-black substrate
scale. On the sea-facing side of the container, Figure 41, more of the loose
material had remained upon the surface. As shown in Figures 3{a) and 38, the
rim located on this portion of the container is nearly severed as a result
of the adjacent corrosive attack.
In contrast to the upper portion of the container, the mid-section ex-
hibits relatively little general attack. The mid-section is that region
bounded by CD and IJ as labelled in Figure 39. Much of an original surface
coating on this portion has remained. In this area, however, severe local
attack has occurred. As seen in Figure 42, the attack in this region is
characterized by local pitting within the coated region, and general
attack adjacent to the chimes. Figure 43 illustrates the attack specific to
the neighborhood of the chimes. In fact, complete perforation was observed
in some areas adjacent to the chimes, where corrosion product had deposited.
Further examination of the corrosion adjacent to the chimes, observed
when the container was cut open and the inner surface viewed, indicated points
of perforation. Some of these points are specific to the region immediately
adjacent to the chimes, as viewed from the interior of the container. Figure
44 shows this chime specific attack perforating the sediment-facing side of
the container, while Figure 45 shows a similar attack located on the sea-side
of the container.
Figure 46 shows the entire inner surface of the carbon steel sheath.
The upper portion in Figure 46 has retained much concrete which adheres to
a rough red scale on the container. The lower portion of the container
contains less adhering concrete. The metal interior surface in this region
is black and appears to be smooth. The relative adherence of the concrete
to these two portions of the container sheath correlates with the increased
spall ing of the waste form where it contacted these regions. Figure 47 shows
the concrete waste form denuded of the carbon steel sheath. Clearly, the
-17-
-------
region contacting the upper portion of the sheath (and closer to the end
exposed to the sea-water) is more severely degraded.
Some further observations can be made from the inside view of the sheath.
Figure 48 shows attack which has perforated a region adjacent to a longitudinal
weld in the upper portion of the container. Similarly, though not severely
enough to produce perforation, the corrosion has also selectively attacked
the region adjacent to the longitudinal weld in the lower container, Figure
49.
At the lower portion of the cylinder, corrosion has produced more exten-
sive and typically deeper pits than those formed within the midsection. This
is shown in Figure 50.
The metal cap on the lower end of the container is shown in Figures 51
and 52, Although there is little attack on this surface of the container,
specific pitting has occurred on the rim and in the center. An interesting
instance of coated steel corrosion, namely "filiform" corrosion is observed
in this section, shown in Figure 52. This is a form of under coating tunneling
attack which forms thread-like traces under the coating. Its active corrosion
cell maintains dissolution within the dark head, producing ferrous ions and
ferrous hydroxide, while the tail contains the oxidized ferric ion and hydrous
ferric oxides and hydroxides. '
B. Corrosion Rates as a Function ofPosition
Metal!ographic examination of a selected number of trepanned metal
specimens was performed to allow a more selective and quantitative assess-
ment of the corrosive attack. Metal samples were trepanned from the carbon
steel sheath along two longitudinal lines, one at 25° ± 5°, and one at
220° ± 5°. This selection of sampling sites allowed a uniform sampling to
be made from the portions of the container exposed to the sediment and the
sea environment respectively. In order to minimize contamination of the
samples, neither lubricant nor water was used for the trepanning operation.
The trepanning was accomplished with a 1-1/2" i.d. hole saw and an air
chisel. Prior to examination, the trepanned samples were stored in glass
vials in a desiccator over anhydrous MgSCL.
-18-
-------
Characteristic 0.05 in2 (0.3 cm2} to 0.08 in2 (0.5 cm2) sections from
the trepanned specimens were selected for an analysis of their cross section
dimension. These samples were selected uniformly from over the container
surface. Selected samples were cut from the trepanned metal specimens and
cold mounted in a metal!ographic epoxy as cross sections. The mounted speci-
mens were coarsely ground by several tenths of a millimeter, followed by a
finer abrasion with 240 grit cloth. Cross sections prepared in this manner
were then photographed at a calibrated 50x magnification. While this pre-
paration does not allow observation of metallurgical microstructure, it
does provide a means of rapidly determining the average cross section di-
mension of a rough surfaced specimen. Corrosion free specimens from within
rimfolds provided reference dimensions. From these data the corrosive
attack as a function of container location was quantitatively determined, as
well as the average metal thinning due to corrosion.
Metal thickness as a function of position on the container is presented
in Figure 53. The error bars show the magnitude of the standard deviation of
this measurement. Their magnitude is a measure of the surface roughness at
the observed 50x magnification used in the microscopic measurement. Approxi-
mately 0.1" of cross-section length was sampled for each point in Figure 53.
Figure 53a illustrates the dimension for the sea-side of the container at
220 , while Figure 53b presents similar data for the sediment-side of the
container at 25°, The unattacked thickness of the container was determined
from analyses of the metal in the rim folds at both ends to be 0.047 in (0.12 cm)
for the upper container, and 0.039 in (0.10 cm) for the lower. The x = 0
end of the container, i.e., the upper container shows significantly more
general attack than that shown by the lower container. However, the lower
container shows more severe local attack, even though there are several points
between x = 20 and x = 50 where there is no measured dimension loss. Comparing
now the thinning data for the sea-side to that for the sediment side, there
is a small but significant difference in the thinning, the sediment side ex-
hibiting more attack.
-19-
-------
Since the corrosion rate of carbon steel in sea water is essentially
constant/ ' an average corrosion rate for the material in this container may
be estimated and compared to rates obtained under defined conditions. An
observed rate is defined as (d - d)/. where d is the initial dimension of the
u u u
cross section of the container and d is the observed dimension at the time of
the container recovery, 15 years,(t), after being placed in the marine environment.
Since there are points on the container where attack has completely penetrated
the 0.039 in (0.1 cm) sheath, a lower limit for local corrosion can be placed
at 0.0026 in/yr (0.067 mm/yr), while the average rates for the general attack
on the upper container are 0.0013 ± 0.0002 and 0.0019 ± 0.0002 in/yr
(0.032 ± .006 and 0.049 ± .006 mm/yr) for the respective sea facing and sedi-
ment facing sides. As can be seen in Table 7, these values range somewhat
under those observed for carbon steel in surface water. The corrosion rate
for carbon steel in deep water at a site off the coast of California has been
observed to be linear in oxygen concentration, following the equation:
Corrosion Rate (microns/yr) = 21.3 + 25.4 (02) + 0.356 (T)
where 0?, the oxygen concentration is in ml/1, and the temperature is in
C.^ ' Using this equation, the corrosion rates calculated for 1 ml/1, 2 ml/1,
4 ml/1 and 6 ml/1 oxygen at 0° are tabulated in Table 7. The oxygen content
ig\
at this depth and location ranges between 5 and 6 ml/1. ' The average
attack observed on the upper container is slightly less than that rate cal-
culated assuming 1 ml/I 0^. Since this is a somewhat low estimate for oxygen
concentration at this depth, the apparent inhibition must be attributed to
some other factor such as an initial surface finish which is clearly in
existence on the lower container, or a decrease in corrosion kinetics at long
times due to a uniform deposit of scale and sediment. '
C. Microscopic Examination of__L_oca1 Effects
While the dimension analysis provides quantitative information on general
corrosion, the more specific forms of attack and the metallic microstructure
have been observed by microscopic techniques, primarily by metallographic
analysis of selected sheath cross sections. The salient features of the
local attack and metal structure of this container shown by these observations
-20-
-------
are discussed in this section. The micrographs in Figures 54 and 55 show
the metallic grain structures developed by nital etch of the respective
upper drum and the lower drum cross section.
The metallurgies of these two portions of the container clearly differ
with respect to grain size; the grain size of the upper container is larger.
While it is difficult to attribute any of the enhanced general corrosion
rate of the upper container to this particular microstructure, an influence
upon local corrosion of the grain structure within a metal fold of the lower
container is seen in Figure 56. Here the attack prefers the ends of the
grains which are elongated by the cold work of the fold.
The chimes within the lower portion of the container, although exhibiting
enhanced attack, show no apparent microstructural features which differ from
the rest of the lower container. As can be seen in Figure 57, the attack at
the chime surface appears more or less general, but with some shallow pitting.
Although the midsection of the container has escaped from severe general
attack, it has exhibited severe pitting even to the point of perforation.
Therefore, the initiation of pitting at points where much of the coating
has remained was investigated.
The coating surface consists of a 55 pun lamina over a 3-5 ym interfacial
scale lying on the metal substrate. Figure 58 shows a photograph of the
coating, scale, and metal. Non-dispersive x-ray fluorescence analysis of
the interfacial region shows the coating to contain Ti, Cr, Zn, and Si,
while the only element apparent within the scale interface is iron. Care-
fully removing a portion of the coating by abrasion reveals a black surface
similar to the Fe^O. observed upon the inside of the carbon steel sheath.
It can be concluded that this scale interface is not the conceivable result
of a zinc phosphate or chromate conversion coating, but is, in fact, a
Fe^CL scale.
Figure 59 shows the cross section of a typical pit formed under the
coating within the midsection. The microphotograph was made so as to high-
light the coating and corrosion product. The coating and the interfacial
-21-
-------
scale is elevated above the pitted region. Remaining within the pitted
region is a corrosion product which appears to be wedged under the edge of
the delaminating coating. The delamination structure is better illustrated
in the scanning electron micrograph appearing in Figure 60. Pictured here
is the coating at the top of the photograph. Failure took place to the left,
where the interfacial oxide and coating had lifted.
The significant point concerning this interface rests in the observation
that in the neighborhood of a pit the adhesion of the coating system to the
substrate fails at the interface of the scale to the metal, rather than that
of the coating to the scale. This is illustrated in Figure 61. Here adhesion
of the oxide to the coating remains while the oxide has lifted from the metal
substrate.
Although lifting of the coating and scale appears to be influenced by the
wedge of corrosion product formed within the crevice under the coating, dis-
bonding of the metal/oxide interface appears even at a distance away from the
actual lifting of the coating. This suggests an electrochemical mechanism in
addition to a purely mechanically enhanced disbonding.
An Fe00. reduction mechanism for the propagation of the coating failure
fll)
by disbonding similar to that proposed^ ' for coatings over thinner ferric
oxides is not inconsistent with these results. By this mechanism, the anodic
dissolution proceeds by the following reaction:
Fe - 2e~ * Fe2+ E = + 0.440 - 0.028 log(Fe2+) (Ref. 14)
and subsequent oxidation:
02 (dissolved) + H20 + Fe2+ + Fe(OH)3 (solid) + 2H+
within an occluded region of a coating break. This provides a driving force
and H activity for the reduction of the interfacial Fe-0.:
2e~ + 8H+ + Fe304 -> 3Fe2+ + 4HgO
E = + 0,980 - 0.236 pH - 0.08861 log (Fe2+) (Ref. 14)
to the extent sufficient to destroy the adhesion of the interfacial Fe^O,
oxide.
-22-
-------
Whether the interfacial oxide was initially present, or whether it had
formed during the immersion period is a significant question. There is
evidence that the interfacial oxide formed during the immersion period. The
lack of a similar interfacial oxide within the rim fold where the coated
surface was sealed from the environment suggests that the interfacial oxide
"was formed after immersion. Indeed, coatings are not complete barriers to
(12)
the flow of corrosion reactants, ' but serve to slow the oxygen and electro-
ns)
lyte penetration.v ; Therefore, a formation due to slow oxidation of the
metal to Fe~CL at the interface is consistent with previously observed corrosion
processes at coating interfaces. To take this hypothesis a step further, the
stresses formed within the coating, due to the oxide formation, particularly
at the chimes, could lead to coating rupture, thus initiating the more rapid
pitting or general attack adjacent to the chimes.
Although neither coating nor remnant of a coating was observed upon the
upper portion of the container, several aspects of the surface morphology
must be noted. Figure 62 illustrates a typical cross section showing' the
attack which appears at the surface of the upper container. The surface
appearing at the top of the photograph is the sea-facing surface. The lower
portion of the photograph shows the surface morphology of the metal at the
concrete/metal interface. The sea-facing surface exhibits general attack
characterized by quite shallow pits. The corrosion product adhering to this
surface was very loose and did not survive the sample preparation. The mor-
phology of attack at the sea surface is to be contrasted with that observed
at the concrete side. On the concrete side of the carbon steel a more adher-
ent and compact oxide has formed and remained within pits. These pits have
greater curvatures than those observed on the sea-side of the container.
This oxide appears black and presumably is the Fe~CL as characterized by XRD
for the black appearing surface oxide (Table 8). As seen in Figure 62, this
oxide takes on a laminated structure where the attack leads to perforation.
The formation of this compact, adherent Fe.,0. is significant to the deteriora-
tion of concrete at concrete/metal interfaces since the volume expansion due
to the density ratio of FegCL/Fe of 2 will compress the concrete causing it
to spall.
-23-
-------
D. Chemical Analysis
Table 8 presents the results of the x-ray diffraction of samples of
loosely adhering corrosion product, scale and surface films. Correlation of
predominant lines to corrosion products was done with reference to the ASTM
cards. In Table 8, the species correlated by the observed XRD lines are
presented in order of their predominance in the specimen whose position and
description is also tabulated.
Several observations can be made from this qualitative data. On the
upper, more rapidly corroding surface, a loose mixture of the ferric oxides
is formed above the substrate with the hydrated a form predominating, while
adhering directly to the metal is the j - FepCL. (In the lower portion of
the container the y - Fe2^3 Predominates in the loose product.) The adherent
dark scale described in preceding sections and found on the interior of the
sheath at the container midsection shows a well defined FeJX diffraction.
The hydrated alpha-ferric oxide is the more stable corrosion product;
however, it is more slowly formed from the initially precipitated iron com-
plexes. ' The loose material clinging to the rusted metal on the upper
container is primarily the a-FeOOH, while the y-FepO.- and yP^O^'FLO pre-
dominate within the scale adhering directly to the surface of the upper
container, and loosely at points of local attack upon the lower container.
This suggests that corrosion on the lower portion of the container may be
a more recent event, its initiation resulting from a relatively late coating
breakdown.
The ferric compounds exist at sites where the subsequent oxidation of
the primary ferrous species is rapid,^ ' while at sites between the sheath
and the concrete the Fe30, is formed due to the slower oxidation of the pri-
mary ferrous ions or metal surface.
The compositions of the two parts of the container are shown in Table 9.
The differences in the compositions cannot sufficiently explain the different
corrosion rates exhibited by the two containers. ' Differences in corrosion
rates must be attributed to the surface finish of the respective portions of
the container sheath.
-24-
-------
V. CONCLUSIONS
(1) Little dissolution of the concrete waste form in the ocean environ-
ment occurred as evidenced by a maximum waste package weight loss of approx-
mately 5%. Measurement error at the time of disposal is the most satisfactory
explanation for the apparent weight loss. A conservative estimate that assumes
"a constant 0.33%/yr weight loss due to cement phase dissolution predicts that
it would require a minimum of 300 years in this environment before the concrete
waste form would lose its integrity and provide no barrier to activity release.
(2) The measured compression strength of the concrete waste form is in
the range expected for comparable concrete formulations. This indicates the
absence of appreciable sulfate attack which is also supported by the obser-
vation that negligible deterioration of the waste form surface has occurred.
(3) The concrete waste form contained Cs-137, Cs-134 and Co-60. Based
on the Cs-137 distribution in the waste form, a bulk leach rate for this
3 2
radionuclide of 2.4 x 10" g/(cm -day) was calculated. This corresponds to an
average cesium-137 release rate of 3.7% per year.
(4) While the inner container which enclosed three wound filter elements
imploded due to the pressure differential during or after descent, water analy-
sis indicated that the container did not leak and hence radionuclides were con-
tained.
(5) Corrosion rates for general attack on the upper portion of the steel
drum (assuming a constant rate with no induction period) were 0.0013-0.0019
in/yr (0.032-0/0.049 mm/yr), A lower limit for the rate of local pitting
corrosion of 0.0026 in/yr (0.067 mm/yr) was determined. Based upon observations
after 15 years in ocean disposal, general thinning attack appears to be the most
important process. Using these rates of general attack, an 18 gauge (nominal
0.0476 in thickness) mild steel drum would require 25-37 years before corrosion
would cause the container to lose its effectiveness as a barrier to activity
migration.
(6) Variations in the corrosion attack between the upper and lower
portions of the drum are ascribed to differences in surface finishes on the
respective portions of the drum. While the coating on the lower portion of
the drum successfully inhibited the initiation of general attack, instances
-25-
-------
of severe local attack leading to pitting and perforation adjacent to the
drum chimes were observed.
(7) The waste container limits seawater exposure and movement through
the waste form in disposal. As a result, the rate of activity loss through
leaching and the rate of cement phase dissolution are decreased.
-26-
-------
References
1. Dyer, R.S., "Environmental Surveys of Two Deep Sea Radioactive Waste
Disposal Sites using Submersibles". Proceedings ofan International
Symposium on Management of Radioactive Hastes from the Nuclear Fuel
Cycie. international Atomic Energy Agency, Vienna, (1976) p. 317-338.
2, Dexter, Stephen C., Cruise Report on R.V. Cape Henlopen 1976 Atlantic
Radioactive Waste Dumpsite Survey, December 1976.
3. Hunt, M.M., W.M. Marquet, D.A. Moller, K.R. Peal, W.K. Smith, and
R.C. Spindell, 1974, "An Acoustic Navigation System", Woods Hole
Oceanographic Institution Technical Report, WHOI 74-6, unpublished
manuscript,
4. U.S. Department of the Interior, Bureau of Reclamation, Concrete Manual,
7th Edition, (1966), p. 148.
5. National Association of Corrosion Engineers, Technical Practices Committee,
"Recommended Practice, Collection and Identification of Corrosion Products",
NACE Standard RP-01-73, February 1973, Houston, Texas.
6. Fontana, M.G. and N.D. Greene, Corrosion Engineering ,. 47-48, McGraw-Hill,
New York, (1967).
7. Tamada, A., M. Tammura and G. Tennyo, "Corrosion Behavior of Low Alloy
Steels in Sea Water" Fifth Int. Congress on Metallic Corrosion, 1972,
786, NACE, Houston, Texas, 1974,
8. Reinhart, F.M. and J.R. Jenkins, "The Relationship Between the Concentration
of Oxygen in Seawater Corrosion of Metals" Proc. 3rd Congr. on Marine
Corrosion and Fouling, p. 562, et. seq. 1972.
9. National Science Foundation, GEOSECS Atlas, to be published.
10. LaQue, F.S., Marine Corrosion , p. 97.
11. Gonzalez, O.P., P.M. Josephic and R.A. Oriani, J. Electrochem. Soc., 121,
29 (1974).
12. Mayne, J.E.O., Research (London), 5,, 278 (1952).
13. Kendig, M.W. and H. Leidheiser, Jr. "The Electrical Properties of Protective
Polymer Coatings as Related to Corrosion of the Substrate", J. Electrochem
Soc.,123 (7), 986 (1976).
14. Pourbaix, Atlas of Electrochemical Equilibria.
-27-
-------
15. Kakehi, T. and H. Yoshino, "Corrosion of Steel in Polluted Sea Waters",
Proc. 5th Int. Cong, on Metallic Corrosion, 1972, NACE, (1974), p. 76.
16. Misawa, T., K. Hashimato and S. Shumodiara, Corrosion Science, 14, 131
(1974). ' ~
-28-
-------
TABLE 1
Major U. S. Radioactive Waste Disposal Sites
(1)
IND
I
Site
Atlantic
Atlantic
Pacific
Farallon Island
(Subsite A)
Farallon Island
(Subsite B)
Coordinates
38°30 ' N
72°06'W
37°50'N
70°35'W
37°38'N
123°08'W
37°37'N
123°17'W
Depth
(m)
2800
3800
900
1700
Distance
from Land
(km)
190
320
60
77
Years
Dumpsite
Used
1951-56
1959-62
1957-59
1951-53
1946-50
1954-65
Estimated No.
of 55-Gallon
Drums Dumped
14,300
14,500
3,500
44,000
Estimated Activi
in Drums at Time
Packaging (Ci)
41,400
2,100
1,100
13,400
ty
of
-------
TABLE 2
Analysis of Liquid found in the Inner
Steel Container and Bottom Seawater
Composition, ppm
Constituent Inner Conta i ner Li quid Bottom Seawater
Sodium 1,250 7,350
Magnesium 2,6 500
Sulfate 245 2,100
Na/Mg ratio 480 15
Na/S04 ratio 2.6 3.5
-30-
-------
TABLE 3
Core Specific Activity - Curies/Gram ± g (%)
Core Mass
Core (Grams) Cs-137 Cs-134 Co-60
OA4 52.08 1.28E - 09 ± 0.4% 2.02E - 12 ± 51.3% 8.74E - 12 ± 12.8%
OA6 189.37 1.61E - 09 ± 0.4% 5.31E - 12 ± 13.7% 7.56E - 12 ± 10.4%
OB2 139.50 3.26E - 11 ± 4.0% — 4.26E - 12 ± 15.0%
OB4 129.76 3.64E - 10 ± 1.1%
OB6 119.05 3.91E - 09 ± 0.4% 1.65E - 11 ± 9.0% 4.56E - 12 ± 25.2%
OC2 179.48 8.30E - 12 ± 11.1% — 2.70E - 12 ± 13.5%
, OC4 169.20 1.73E - 11 ± 5.3% -- 2.30E - 12 ± 22.2%
r OC6 151.66 6.87E - 12 ± 2.6% — 2.44E - 12 ± 19.7%
OC7.5 63.75 3.02E - 10 ± 1.1% — 4.46E - 12 ± 15.7%
OD2 157.33 3.11E - 11 ± 4.5%
OD4 153.39 2.65E - 10 ± 1.2% — 1.61E - 12 ± 50.0%
OD6 57.13 6.38E - 10 ± 0.9% — 3.43E - 12 ± 39.3%
OD7.5 92.65 5.39E - 10 ± 1.1% — 3.81E - 12 ± 39.2%
OE2 197.80 2.44E - 11 ± 5.0%
OE4 126.67 1.14E - 10 ± 2.2% — 1.95E - 12 ± 50.0%
OE6 117.16 8.66E - 10 ± 0.8% 1.96E - 12 ± 44.1% 1.39E - 12 ± 50.0%
OE7.5 92.31 1.14E - 09 ± 0.7% — 1.44E - 12 ± 28.8%
OF4 171.79 1.72E - 10 ± 1.4% -- 2.11E - 12 ± 34.3%
OF6 110.03 5.87E - 10 ± 1.0% 3.68E - 12 ± 28.3% 4.62E - 12 ± 29.6%
062 133.45 7.42E - 12 ± 11.7% — 1.85E - 12 ± 50.0%
-------
TABLE 3 (Cont'd)
Core Specific Activity - Curies/Gram ± a (%)
ro
i
Core
OG4
OG6
OG7.5
OH4
OH6
OH7.5
90C4
90C6
90E2
90E4
90E6
90G2
90G4
90G6
180A2
180A4
180A6
180D3.5
180F4
180 F6
270C2
270C4
Core Mass
(Grams)
158.37
169.88
100.86
92.58
150.72
54.56
133.22
103.53
173.08
120.36
91.24
120.41
114.59
100.81
179.96
129.57
149.06
107.34
99.81
54.80
148.92
125.31
1.17E
5.52E
4.53E
1.31E
5.96E
2.26E
2.46E
7.84E
6.07E
4.37E
3.18E
1.59E
6.43E
7.75E
1.06E
3.82E
1.05E
5.51E
2.14E
1.98E
2.75E
1.02E
Cs-137 Cs-134
- 11 ± 8.1%
- 11 ± 2.7%
- 10 ± 1.1%
- 10 ± 2.4%
- 10 ± 0.8%
- 10 ± 0.6%
- 12 ± 12.2%
- 12 ± 7.1%
- 12 ± 9.4%
- 12 ± 12.6%
- 12 ± 20.0%
- 12 ± 16.8%
- 12 ± 18.7%
- 12 ± 8.5%
- 11 ± 6.9%
- 12 ± 6.2%
- 10 ± 1.9%
- 12 ± 9.8%
- 11 ± 7.0%
- 11 ± 6.3%
- 12 ± 14.8%
- 10 ± 1.5%
Co-60
1.91E -
1.45E -
3.71E -
4.82E -
2.28E -
1.81E -
2.30E -
3.41E -
1.49E -
4.73E -
6.61E -
6.76E -
3.21E -
5.07E -
2.79E -
6.71E -
4.74E -
1.96E -
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
11 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
12 ±
36.1%
50.0%
33.9%
31.8%
28.8%
50.0%
13.4%
14.8%
35.9%
10.1%
15.6%
11.0%
13.7%
13.8%
15.6%
16.8%
31.3%
19.9%
-------
TABLE.3 (Cont'd)
Core Specific Activity - Curies/Gram± a
Core Mass
Core (Grams) Cs-137 Cs-134 Co-6Q
270C6 114.59 1.40E - 10 ± 0.6% 2.12E - 12 ± 22.2%
270E2 151.06 4.48E - 12 ± 20.5%
270E4 152.92 2.82E - 11 ± 4.2% — 2.67E - 12 ± 35.0%
270E6 85.99 3.71E - 10 ± 1.0% 1.49E - 12 ± 50.0%
270G2 121.86 7.30E - 11 ± 2.8% -- 4.74E - 12 ± 31.4%
270G4 113.23 3.47E - 11 ± 3.0% — 2.30E - 12 ± 20.2%
270G6 135.41 4.09E - 10 ± 0.7% 4.25E - 13 ± 88.3% 2.57E - 12 ± 16.2%
co
CO
i
-------
TABLE 4
Cement Specific Activity - Curies /Gram ± a (%)
Cement Mass
Core (Grams) Cs-137 Cs-134 Co-60
OA4 17.08 3.90E - 09 ± 0.4% 6.17E - 12 ± 51.3% 2.66E - 11 ± 12.8%
OA6 60.52 5.05E - 09 ± 0.4% 1.61E - 11 ± 13.7% 2.37E - 11 ± 10.4%
OB2 46.75 9.73E - 11 ± 4.0% — 1 .27E - 11 ± 15.0%
OB4 38.67 1.22E - 09 ± 1.1%
OB6 69.02 6.75E - 09 ± 0.4% 2.84E - 11 ± 9.0% 7.87E - 12 ± 25.2%
OC2 48.13 3.10E - 11 ± 11.1% — 1.01E - 11 ± 13.5%
OC4 65.44 4.47E - 11 ± 5.3% — 5.95E - 12 ± 22.2%
OC6 55.63 1.87E - 11 ± 2.6% - 6.65E - 12 ± 19.7%
OC7.5 25.81 7.49E - 10 ± 1.1% — 1 .10E - 11 ± 15.7%
OD2 71.50 6.84E - 11 ± 4.51
OD4 28.08 1.45E - 09 ± 1.2% — 8.77E - 12 ± 50.0%
OD6 20.79 1.75E - 09 ± 0.9% — 9.42E - 12 ± 39.3%
OD7.5 29.99 1 .67E ~ 09 ± 1 .1% — 1 .18E - 11 ± 39.2%
OE2 62.10 7.76E - 11 ± 5,0%
OE4 49.02 2.94E - 10 ± 2.2% — 5.03E - 12 ± 50.0%
OE6 47.68 2.13E - 09 ± 0.8% 4.81E- 12 ± 44.1 3.43E - 12 ± 50.0%
OE7.5 25.82 4.07E - 0.9 ± 0.7% — 5.15E - 12 ± 28.8%
OF4 83.22 3.55E - 10 ± 1.4% -- 4.35E - 12 ± 34.3%
OF6 31.20 2.07E - 09 ± 1.0% 1.30E - 11 ± 28.3% 2.06E - 11 ± 29.6%
OG2 24.94 3.97E - 11 ± 11.7 — 9.85E - 11 ± 50,0%
-------
TABLE 4 (Cont'd)
Cement Specific Activity - Curies/Gram ± a (%)
CO
en
Core
064
066
067.5
OH4
OH6
OH7.5
90C4
90C6
90E2
90E4
90E6
9062
9064
9066
180A2
180A4
180A6
1800D3.5
180.F4
180F6
270C2
270C4
Cement Mass
(6rams)
42.61
30.77
76.62
16.01
35.03
13.30
35.40
32.57
37.44
38.03
33.42
32.56
32.31
32.90
33.42
22.52
33.87
31.16
27.59
10.45
37.59
34.26
Cs-137
4.33E -
3.05E -
5.96E -
7.55E -
2.56E -
9.27E -
9.25E -
2.49E -
2.81E -
1.38E -
8.69E -
5.87E -
2.28E -
2.37E -
5.69E -
2.20E -
4.61E -
1.90E -
7.74E -
1.04E -
1.09E -
3.72E -
11 ± 8.1%
10 ± 2.7%
10 ± 1.1%
10 ± 2.4%
09 ± 0.8%
10 ± 0.6%
12 ± 12.2%
11 ± 7.1%
11 ± 9.4%
11 ± 12.6%
12 ± 20.0%
12 ± 16.8%
11 ± 18.7%
11 ± 8.5%
11 ± 6.9%
11 ± 6.2%
10 ± 1.9%
11 ± 9.8%
11 ± 7.0%
10 + 6.3%
11 ± 14.8%
10 ± 1.51
Cs-134
—
—
—
—
—
—
--
--
—
—
—
--
--
--
—
__
--
—
—
--
—
__
Co-60
7.09E - 12 ± 36.1%
8.00E - 12 ± 50.0%
4.89E - 12 ± 33.9%
2.79E - 11 ± 31.8%
1.12E - 11 ± 28.8%
7.41E - 12 ± 50.0%
1.28E - 10 ± 13.4%
1.08E
4.13E
1.80E
2.34E
2.18E
1.85E
2.23E
9.60E
2.43E
2.48E
7.79E
11 ± 14.8%
12 ± 35.9%
11 ± 10.1%
11 ± 15.6%
11 ± 11.0%
11 ± 13.7%
11 ± 13.8%
12 ± 15.6%
11 ± 16.8%
11 ± 31.3%
12 ± 19.9%
-------
TABLE 4 (Cont'd)
CementSpecific Activity - Curies/Gram ± a (%)
Cement Mass
Core (Grams) Cs-137 Cs-134 Co-60
270C6 25.08 6.37E - 10 ± 0.6% 9.69E - 12 ± 22.2%
270E2 42.52 1.59E - 11 ± 20.5%
270E4 32.52 1.32E - 10 ± 4.2% — 1.26E - 11 ± 35.0%
270E6 20.90 1.53E - 09 ± 1.0% 6.14E - 12 ± 50.0%
270G2 31.81 2.80E - 10 ± 2.8% — 1.82E - 11 ± 31.4%
270G4 29.40 1.34E - 10 ± 3.0% -- 8.85E - 12 ± 20.2%
27066 23.30 2.38E - 09 ± 0.7% 2.47E - 12 ± 88.3% 1.49E - 11 ± 16.2%
OJ
en
i
-------
TABLE 5
Cesium-137 Content in Annular Volumes of
the Concrete Waste Form as a Function of Depth
Annular element
(core depth) , in.
0-2
2-4
4-6
6-7.5
El ement
Volume, cm^
1.07 x 105
8.63 x 104
6.60 x 104
3.62 x 104
Average
Cesium-137
curies/gram concrete
2.02 x 10"11
1.54 x 10"10
5.83 x ID'10
5.32 x 10"10
Cesium-137
content, curies
4.78 x 10"6
2.94 x 10"5
8.50 x 10"5
4.26 x 10"5
TOTAL
2.96 x 10'
1.62 x 10
-4
Average concrete density =2.21 g/cm .
-37-
-------
TABLE 6
Concrete Core Compressive Strength
Distance From
Core Location, i
8
3
4
5
5
-4
Core
n. Location
180 D
180 E
180 F
180 F
180 F
180 H
Core
Diameter, in.
1.73
1.73
1.73
1.73
1.73
1.73
Core
Length, in.
2.24
2.28
3.15
2.17
3.58
2.17
Comprehensive Strength
psi
1720
1910
1700
1680
1720
1510
Impact Hammer Method (Average) 4100
-38-
-------
TABLE 7
Mild Steel Corrosion Rates in Ocean Environments
Corrosion Rate of
Carbon Steel in/yr Ref.
Clean Surface Waters
off Coast of Japan
Projected from Five Year
Tests in Surface Waters
Emperical Formula
1 ml/I 02, 0°C
2 ml/I 02, 0°C
4 ml/I 02, 0°C
6 ml/1 02, 0°C
General Attack in Upper
Container
Sea Side
Sediment Side
Local Attack in Container
0.002
0.0023
0.0019
0.0028
0.0048
0.0069
0.0013 ± 0.0002
0.0019 ± 0.0002
>0.0026
(15)
(7)
(8)
(8)
(8)
(8)
-39-
-------
TABLE 8
X-ray Diffraction Identification of Surface S^ragings
7 44 220° Y - Fe203
10 27 220° Y - Fe203 • H20, a - Fe203 • H20
32 17 45° a - Fe203 • HgO, Y - Fe203 • H20, y - F
4 13 207° a - Fe203 • HgO, Y - Fe203 • H20
40 10 180° a - Fe203 • H20 Loose Surface Material
M-25 0 25° Y - Fe203 Surface Scraping
M-25 0 25° Y - Fe203 Surface Diffraction
M-6 0 Fe~04 Surface Diffraction of
Black Inner Surface
-40-
-------
TABLE 9
Trace Element Analyses of Container Materials
Weight Percent
£ Mn Sj_ Cu_ £r N1_
Upper Container 0.097 0.30 0.0023 0.123 0.013 0.024
Lower Container 0.090 0.35 0.0046 0.067 0.007 0.019
Weld in Lower 0.096 0.33 0.0038 0.086 0.017 0.029
Container
-41-
-------
Q
\OUMP
SITE
A TLANTIC
OCEAN
40°
bo'
38°
OO'
36°
~oo'
74°OO
72° OO
Figure l(a).
Approximate Location of the Atlantic 2800 m Radioactive
Waste Disposal Site.(2)
-42-
-------
72°
6'w
9
A
18
a
a
1C
Al
ri4'W 72°
2
a
o4
7
> 11
D
12 'W
3 15
02D
^*"V"V^f— 1
Vj/v,/^^^!
\
CO
RE
1
D
|C
D
IQ'W
19
D
NTAINER
COVERY
SITE
^
[
L
08' W
DUMP
SITE
BOUND
14
D
D Box Cores
!i Current Meter Mo
5 Water Samples
oe'w
04'W
8
A
_^~--~^
ARY
A7
orings
i ,
H
02'W 72°00'W 71°
a_38°
37'N
38°
35'N
33'N
3l'N
29'N
27'N
38°
25'N
23'N
58' W
Figure 1b. Location of the Retrieved 80 - Gallon Radioactive Waste Package
43
(2)
-------
TWO 48 FLOTATION SPHERES
TAG LINE
AND FLOAT
^SYNTACTIC FOOTBALL FLOATS
FOR POSITIVE TRANSPONDER
BUOYANCY
LANYARD PULL
FOR RELEASE
HOOK
TRANSPONDER
SAFETY RELEASE HOOK
100 M
I" NYLON LINE
TEE
HANDLE
2000 LB.
WORKING LOAD
SNAP HOOK
O_
DRUM
1000 LB.
500 LB,
CLUMP ANCHORS
SUBMERSIBLE
TAG LINE
Fiqure 2. Schematic Diagram of the Hoist System Used for the Retrieval
of the Waste Package from the Ocean Floor at a Depth of 2783 meters.
(2)
-44-
-------
Figure 3(a). Open End of Radioactive Waste Package Immediately after Surfacing
from the Atlantic 2800 m Disposal Site.
-45-
-------
Figure 3(b). Side View of Radioactive Waste Package Prior to Being Brought Aboard.
-46-
-------
Figure 4(a).
A Modified H47 Jet Engine Container Used for the Encapsulation
and Shipment of the Retrieved Radioactive Waste Package.
Figure 4(b).
An Open H47 Jet Engine Container Showing Rubber Faced Interior
Clamp Rings Provided to Hold the Radioactive Waste Package
Against Shock and Vibration.
-47-
-------
Figure 5. Surface Markings on the Exposed
Concrete Face of the Waste
Package.
-48-
-------
.— ^ ~7 n
£ f . U
lc 1 S s"
1 0,O
. — - ..
-*i K-
OPEN
END
Q.
—
E
r>
(E
Q
o:
o
*^J""
0
-z.
E
LU
s
X
o
— oo.c.
t
^
0.75"
*l k-
3| (E
*!
;
;
o
u- ..... Fn
oc.
u
I
C.
1.9
11 ,1
270
80 GALLON RETRIEVED DRUM
AXIAL BARREL
SEAM
ISO'
OPEN END
180°
CLOSED END
270
Figure 6. Orientation System Used to Describe the Waste Package.
-------
I
on
O
rt-
a
"O
Figure 7. Montage of Waste Package Radiographs. The Top of the Figure
Shows the Flanged End of the Internal Container Which Was
Located Approximately 5,5 Inches from the Open End of the
Waste Package. Note the Concavity along the Length of the
Container (on the right edge in this figure) Which Resulted
from Implosion During or After Descent to the Seafloor.
Film was Positioned along 270° Axis.
-------
Figure 8. Radiograph of the Flanged End of the Internal Container.
-------
Figure 9. View of the Concrete Waste Form
with the Steel Drum Removed
(0° axis).
-52-
-------
Figure 10, View of the Concrete Waste Form
with the Steel Drum Removed
(90° axis).
-53-
-------
Figure 11. View of the Concrete Waste Form
with the Steel Drum Removed
(180° axis).
-54-
-------
Figure 12. View of the Concrete Waste Form
with the Steel Drum Removed
(270° axis).
-55-
-------
Figure 13. Core Drilling of the Concrete Waste Form. The Worker on the Right
is Moving the Drill Bit into the Waste Form While the Worker on the
Left Holds the Pneumatic Chisel Used to Remove the Steel Drum.
-56-
-------
U1
••si
I
£
27.i:- G-
: • :'''. •:,;;•' -.'. :' f-'--.:
Figure 14.
-------
Ol
co
I
Figure 15. Letter Designation of Core Locations along the 900 Longitudina!
-------
I
Cn
UD
I
Figure 16. Letter Designation of Core Locations along the 180° Longitudinal
Axis.
-------
I
en
O
i
Figure 17. Letter Desi
--—•" Ayic;
ignation of Core Locations along the 270° Longitudi
nal
-------
i
01
x(inches) =02 6
1 I
OPEN END
<^J
CORE HOLE
DESIGNATION
O O
A B
10
1
O
c
18 26
| I
O O
D E
34
1
O
F
42 46
1 1
O O
G H
49.3
22.8 270
CORE COORDINATES DESCRIBED BY (9,x,r)
6, DEGREES, MEASURED CLOCKWISE ON OPEN END
x , INCHES, DISTANCE ALONG CENTERLINE FROM OPEN END
(ALSO DESCRIBED BY CORE HOLE DESIGNATION)
r, INCHES, DEPTH OF CORE FROM CONCRETE SURFACE
180
Figure 18. Schematic Showing the Letter Designation of Core Holes as Related
to (e,x,r) Location Coordinates.
-------
Figure 19, Closed End of the Waste Form after
Removal of the Five Inch Thick
Initial Concrete Pouring.
-62-
-------
Figure 20. Waste Form after Removal of the Upper
Eight Inches of Concrete from the
Open End, Exposing the Flanged End
of the Inner Container.
-63-
-------
Figure 21. View of the Exposed Flange End of the Inner Container. Note the Gap
between the Container and the Concrete Formed by Implosion of the
Inner Container Wall during Descent.
-64-
-------
crt
Figure 22. Concrete Waste Form after Removal
of the Inner Container.
-------
I
Ol
en
i
Figure 23. Inner Steel Vessel after Removal from the Haste Form. Note
Implosion of the Container Walls along the 0° Longitudinal
AX1 S *
-------
24.
Inner
Steel
Sel
a Jon
9 the
180°
Axis,
-------
Figure 25. Inner Steel Vessel with the Cover Removed, Showing the Enclosed
Wound Filter Assemblies.
y:fe:,:, -68-
-------
Figure 26. Close up View of the Inner Container Cover, Flange and Enclosed
Filter Assemblies.
-69-
-------
!0~8 in
10
-9
LU
S
LU
O
< in-l
QC IU
CO
yj
(T
O
10'
10
-12
0
ORIENTATION
CESIUM-137
CORE DEPTH
o 0-2 inch
a 2-4 inch
0 4-6 inch
> 6-7.5 inch
ABC
i i i
i i i
CORE HOLE DESIGNATION
D E F G H
10 20 30 40 50
DISTANCE,x, INCHES
60
Figure 27. Cesiutn-137 Content in Concrete Cores along the 0° Longitudinal
Axis.
-70-
-------
UJ
LJ
O
O
CO
UJ
cr
u
ID
'9
-IO
10
-"
10
-12
90° ORIENTATION
CESIUM-I37
CORE DEPTH
o 0-2 inch
n 2-4 inch
0 4-6 inch
CORE HOLE DESIGNATION
10 20 30 40 50
DISTANCED, INCHES
60
Figure 28. Cesium-137 Content in Concrete Cores along the 90° Longitudinal
Axis.
-71-
-------
10'
r9
Ld
O
c/)
UJ
O
10
-II
10
r!2
180° ORIENTATION
CESIUM -137
CORE DEPTH
a 2-4 inch
0 4-6 inch
CORE HOLE DESIGNATION
D F
10 20 30 40 50
DISTANCE, x , INCHES
60
Figure 29. Cesium-137 Content in Concrete Cores along the 180° Longitudinal
s.
-72-
-------
10
-8
10
,-9
UJ
^
UJ
o
or
(D
\
CO
UJ
cr
Z>
o
10
rIO
10"
1 I I
270° ORIENTATION
CESIUM- 137
CORE DEPTH
o 0-2 inch
n 2-4 inch
0 4-6 inch
CORE HOLE DESIGNATION
E G
10 20 30 40 50
DISTANCE, x, INCHES
60
Figure 30. Ceslum-137 Content in Concrete Cores along the 270° Longitudinal
Axis.
-73-
-------
io-'°
1-
UJ
Ld
O
CURIES/GRAM
5 6
i i
r\3 —
ID'1 3
- i I i i i I -
- 0° ORIENTATION ~
CESIUM-134
— —
— —
" X>
/M
-/\
l A
^ / CORE DEPTH -
n 2-4 inch
0 4-6 inch
CORE HOLE DESIGNATION
"ABCDEFGH
ii i i i i '
it i i i ii
i i i i i
10 20 30 40 50 60
DISTANCED, INCHES
Figure 31. Cesium-134 Content in Concrete Cores along the 0° Longitudinal
Axis.
-74-
-------
10
-10
H
Z
UJ
5
LU
O
10
-I I
UJ
tr
=>
o
10
H2
10
-13
270° ORIENTATION
CESIUM-134
CORE DEPTH
0 4-6 inch
C
i
10
CORE HOLE DESIGNATION
E G
T
20 30 40 50
DISTANCE,x, INCHES
60
Figure 32. Cesium-134 Content in Concrete Cores along the 270° Longitudinal
Axis.
-75-
-------
1 I I
0° ORIENTATION
COBALT - 60
l-
UJ
^
LU
O
CORE DEPTH
o 0- 2 inch
n 2- 4 inch
04-6 inch
> 6- 7.5 inch
ID
O
10'
10
-12
A B
-------
10
-9
z
LU
s
LJ
O
< 10
tr
e>
LU
CC
=>
O
10
-10
10
H2
90° ORIENTATION
COBALT-60
CORE DEPTH
o 0-2 inch
n 2-4 inch
04-6 inch
CORE'HOLE DESIGNATION
E G
10
20 30 40 50
DISTANCE, x , INCHES
60
Figure 34. Cobalt-60 Content in Concrete Cores along the 90° Longitudinal
Axis.
-77-
-------
LU
S
Ld
O
O
\
CO
UJ
cr
Z)
o
0"
10
-12
"A
80° ORIENTATION
COBALT-60
CORE DEPTH
a 2-4 inch
0 4-6 inch
CORE HOLE DESIGNATION
D F
10 20 30 40 50
DISTANCED, INCHES
60
Figure 35. Cobalt-60 Content in Concrete Cores along the 180° Longitudinal
Axis.
-78-
-------
h-
yj
LU
o
5
cr
X.
LU
CC
Z)
o
10
-10
10
IO
H2
CORE DEPTH
o 0-2 inch
n 2-4 inch
0 4-6 inch
270° ORIENTATION
COBALT-60
CORE HOLE DESIGNATION
E G
10 20 30 40 50
DISTANCED , INCHES
60
Figure 36. Cobalt-60 Content in Concrete Cores along the 270° Longitudinal
Axis.
-79-
-------
CD
O
I
180'
T
o"-
II
50.1 —
90'
A
O1
270
H
B
J
-O-
D
Figure 37. A Map of the Surface of the Container.
-------
00
. •'" •"';' '•• ." . ••• .-V"-•'•'• .^-'-." '1
Figure 38. The Entire Length of the Container.
-------
I
oo
/
E
\ C E C
i
1
1
1
1
1
1 F ~~--"
1
1
1
1
1
i
3 D !•
5 ]
^__
"""*"•*••»•.
H
•
" -— »
"~*""**— *. ^.^
J
K
CONTAINER SCHEMATIC
Figure 39. A Schematic illustrating Specific Features of the Container
as Seen in Figure 38.
-------
00
CO
I
Figure 40. An Exterior View of the Upper Portion of the Container
Sediment Facing Side.
-------
I
CO
-pi
I
Figure 41. An Exterior View of the Upper Portion of the Container - Sea
Facing Side.
-------
CO
on
Figure 42. An Exterior View of the Mid-Section of the Container.
-------
CO
cr>
t
1"
H
Figure 43. An Exterior View of General Attack Adjacent to a Chime
within the Mid-Section.
-------
Chime
1"
CO
Figure 44. Sediment Side Perforation Adjacent to a Chime as Viewed from
the Inside of the Carbon Steel Sheath.
-------
Chime
1"
Figure 45. Sea Side Perforation Adjacent to a
Chime as Viewed from the Inside of
the Carbon Steel Sheath.
-88-
-------
I
00
Figure 46. The Interior Surface of the Carbon Steel Sheath.
-------
o
I
Figure 47. The Concrete Waste Form. The Upper Portion of the Form is
to the Right in the Photograph.
-------
i"
Figure 48. Attack Adjacent to the Weld in the Upper Container.
-------
'-:<"
ro
1"
I 1
Figure 49. Attack Adjacent to the Weld in'the Lower Container.
-------
co
I
' ' ' : •••:•:•' .... • .'.•:•-
Figure 50. Macroscopic Pits Covering the Carbon S
End of the Container.
-------
(O
4->
c
o
-------
1
l£»
en
Figure 52. An Example of Filiform Corrosion.
-------
10
01
I
0.10
o 0.05
*•»
CO
CO
LU
o
I-
<
UJ
O.I
0.05
220°±10
10
10
20 30
X, INCHES
40
50
50
Figure 53. The Sheath Thickness vs. Container Position.
-------
::^miK
W-^S
^...vfvi-*-.;,,;.^
.•'.'••.•••' ••\,i.,s:'-. •• ••.''. '•"•'
®S§S?'
' Ji*-v-4::- :-X '-.V'V
.',»•.. :
^
0.01
II
Figure 54. A Typical Metallographic Cross Section
of the Upper Container.
-97-
-------
-.
- ;• '3 - <- • * • v; .^-^ ,
•
Figure 55. A Typical Metallographic Cross Section
of the Lower Container.
-98-
-------
Figure 56. Attack Within a Rim Fold.
-99-
-------
0.01
It
Figure 57. Attack at a Chime.
-100-
-------
Figure 58. The Coated Interface,
-101-
-------
II
Figure 59. A Pit Formed Within the Coated
Region of the Mid-Section of
the Container.
-102-
-------
.
0.001
.-;.
Figure 60. Scanning Electron Micrograph of
the Disbonding of the Interfacial
Oxide.
-103-
-------
Figure 61. Micrograph of the Attack upon
the Upper Container.
-104-
-------
Figure 62. A Perforation Formed at the Upper
Container Sheathing.
-105-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA 520/1-82-009
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Analysis and Evaluation of a Radioactive Waste Package
Retrieved from the Atlantic 2800 Meter Disposal Site
8. REPORT DATE
May, 1982
Date of
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
P. Colombo, R. M, Neilson, Jr., and M. W. Kendig
8. PERFORMING ORGANIZATION REPORT NO.
BNL 51102
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Nuclear Waste Management Research Group
Department of Nuclear Energy
Brookhaven National Laboratory
Upton, New York 11973
1O. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Interagency Agreement No.
EPA-IAG-D6-0166
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Radiation Programs
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
ANR-461
13. SUPPLEMENTARY NOTES
16. ABSTRACT Qn juiy ^lt 1976, an 80-gallon radioactive waste package was retrieved from a
low-level radioactive waste dumpsite in the Atlantic, located at a depth of 2800 meters
120 miles (190 Km) east of the Maryland-Delaware coast at coordinates 38°30'N, 72°06'W.
This was the first such recovery of a radioactive waste package from an ocean dumpsite
and was conducted by the EPA Office of Radiation Programs. The drum was transported to
the Brookhaven National Laboratory where container corrosion, and matrix leach rate and
degradation rate analyses were conducted. The drum was dumped approximately 15 years
prior to recovery and was found to contain a sealed steel vessel containing some liquid
and wound filter assemblies. The integrity of the concrete matrix had not degraded
appreciably, and it is estimated that in the ocean dumpsite recovery environment it
would require a minimum of 300 years before the concrete waste form would lose its
integrity and provide no barrier to radioactivity release. The concrete waste form con-
tained eesium-134, cesium-137, and cobalt-60 in both the concrete matrix and the inner
steel vessel. The inner steel vessel did not leak; hence, the radioisotopes were con-
tained. The estimated annual rate of leaching of the cesium-137 radioisotope measured
in the concrete matrix was 3.7% per year.
Corrosion attack on the metal container varied between the upper portion of the
drum exposed to ocean water, and the lower portion of the drum exposed to sediment. Genr
eral thinning attack appears to be the most important corrosion process. It is estimat-
ed that an 18 gauge mild steel drum in this ocean dumpsite would require 25-37 years
before corrosion would cause the metal container to lose its effectiveness as a barrier
to radioactivity migration.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
Ocean Dumping/Sea Disposal
Radioactive Waste Disposal/Nuclear Waste
Disposal
Radioactive Waste Packaging-Concrete
Deepsea Corrosion
18, DISTRIBUTION STATEMENT
Unlimited Release
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
118
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |