-------
EPA 520/1-82-005
U. Del. Report No. CMS-C-1-78
MATERIALS FOR CONTAINMENT
OF LOW-LEVEL NUCLEAR WASTE
IN THE DEEP OCEAN
by
Stephen C. Dexter
Associate Professor
of Ocean Engineering
and Materials Science
College of Marine Studies
University of Delaware
Prepared August, 1978
Revised April, 1980
and June, 1983
This report was prepared as an account of
work sponsored by the Environmental
Protection Agency of the United States
Government under Contract No. WA-6-99-2767-J
Project Officer
Robert S. Dyer
Analysis and Support Division
Office of Radiation Programs
U.S. Environmental Protection Agency
Washington, D.C. 20460
ongoing effort to evaluate the low-level radioactive waste packaging
techniques and materials used in the past, with a view towards developing
packaging performance criteria and recommendations for improved packaging
materials and design.
As part of this development process, the Office has initiated a
generic study to describe the various corrosion and degradation processes
that would be expected to affect a waste package in the deepsea. This
report presents some preliminary results of that ongoing study. Based
upon these preliminary results, an estimate can be made of the expected
lifetime of a conventional metal drum containing a concrete matrix. This
information can then be compared with the empirical results of the
separate, detailed analyses of various low-level radioactive waste
packages recovered from the ocean under the direction of this Office, and
any differences will be examined.
Readers of this report are encouraged to inform Mr. David E. Janes,
Director, Analysis and Support Division (ANR-A61), Office of Radiation
Programs, U.S. Environmental Protection Agency, Washington, D.C. 20460,
of any errors, omissions, or other comments pertinent to improving this
document.
Gien L. Sjcblom, Director
Office of Radiation Programs
-------
-------
ABSTRACT
During the period between 1946 and 1970 the United States carried
out sea disposal of low-level radioactive waste. A low-level waste
package was developed consisting of a plain carbon structural steel 55-
gallon cylindrical drum into which the radioactive material and a
concrete matrix was admixed. Tie present report examines the electro-
chemical processes which corrode metals in saltwater, factors affecting
corrosion rates in the deep sea, and mechanisms which degrade concrete
immersed in seawater. Environmental data is reviewed on the rates and
interactive nature of these corrosion and degradation mechanisms. For
the purposes of this report, failure of the container is considered to
have occurred as soon as seawater is allowed to contact the waste.
Based on this criterion, it is concluded that the lifetime of conven-
tional containers is likely to be short compared to the ten half-life
(50-300 years) minimum isolation period that has been considered for key
radioactive components of low-level waste, such as strontium-90, and
cesium-137. The concept for an improved cost-efficient container
capable of isolating wastes for 50 to 75 years with a high degree of
reliability is presented. However, it is emphasized that our present
knowledge of the mechanisms and rates of deterioration of structural
materials in the deep ocean is insufficient to guarantee the integrity
of even the improved containers for upwards of 300 years.
-------
-------
TABLE OF CONTENTS
1.0 INTRODUCTION. . . . , 1
2.0 DETERIORATION MECHANISMS 3
2.1 CORROSION OF METALS 3
Basic Mechanism 3
Structural Steel 5
Stainless Steels 7
Factors Affecting Deep Ocean Corrosion Rates 9
2.2 DETERIORATION OF CONCRETE 17
3.0 ENVIRONMENTAL DATA 17
3.1 STRUCTURAL STEEL 17
3.2 STAINLESS STEELS 23
3.3 CONCRETE 26
3.4 EFFECTS OF RADIATION 27
4.0 LIFETIME OF CONVENTIONAL CONTAINERS 29
5.0 ALTERNATIVES TO PRESENT MATERIALS OF CONTAINMENT 35
5.1 INNER CORE 36
5.2 CONCRETE LAYER 41
5.3 OUTER SHELL 42
6.0 REFERENCES 45
-------
-------
1.0 INTRODUCTION
The two classes of materials that have historically been used for
packaging low-level nuclear waste for ocean disposal are plain carbon
structural steel and concrete. These were usually arranged in the form
of a two- or three-layered container starting on the outside with a 55-
or 80-gallon structural steel oil drum from which one end had been
removed. First, the drum was partially filled with concrete. The waste
form itself was pressed into the wet concrete and the drum then filled
with more concrete. Some types of low-level liquid waste were mixed
directly with the concrete in the steel drum. Occasionally the waste
form itself was a contaminated or activated piece of plain carbon or
stainless steel, which was pressed into the wet concrete. In this
latter case, the steel or stainless steel waste form would act as an
additional containment barrier in that radionuclides would be released
to the environment only through corrosion of the steel or stainless
steel itself. Thus, a third class of materials which must be con-
sidered, in addition to the concrete and steel mentioned above, are the
stainless steels.
It is assumed throughout this report that typical low-level nuclear
wastes must be isolated from man and his food chain for a minimum period
of time corresponding to ten half-lives of the waste. For example, for
low-level wastes containing Cs-137 or Sr-90 this containment period
would be a minimum of 300 years, and 280 years respectively. Other
wastes having shorter half-lives (e.g. Co-60) would need to be isolated
for upwards of 50 years. It is further assumed that public opinion will
not long tolerate any waste disposal program in which there is even a
small percentage of containment failures that result in premature
-------
release of the waste to the marine environment before the end of the
appropriate minimum containment period.
All three classes of materials mentioned above are known to be
susceptible to deterioration in seawater environments. The degradation
mechanisms begin immediately upon immersion of these materials in sea-
water, and appreciable deterioration by corrosion can occur in as little
as one or two years for steel and stainless steel, and in 5 to 20 years
for concrete. These times are much shorter than the 50 to 300 year
minimum containment period for the radionuclides discussed above. The
purposes of this report are, therefore, to (1) summarize the current
literature on the seawater degradation of these materials, (2) assess
what the lifetime of these materials is likely to be in the deep sea and
in the configurations used for packaging nuclear wastes, and (3) suggest
possible alternative packaging techniques and materials.
From an engineering point of view, the purpose of the outer steel
container is primarily to serve as a convenient mold for the concrete.
Before it starts to corrode appreciably, the steel container also
provides a partial barrier between the concrete and seawater. In past
ocean dumping this was a minor function for steel because the packages
left the concrete exposed directly to seawater at the open end of the
drum. Therefore, functionally, it matters little whether or not the
steel corrodes. Deterioration of the concrete and the metals of the
waste form itself are a more serious matter, as they are the materials
which are called upon to isolate the waste.
In the following sections, the appropriate deterioration mechanisms
will be outlined, the available environmental data on degradation rates
will be summarized, the expected lifetime of the conventional types of
-------
containers will be estimated from these data, and several alternatives
to the present methods of containment will be discussed.
2.0 DETERIORATION MECHANISMS
2.1 CORROSION OF METALS
Basic Mechanism
Degradation of metallic materials in marine environments by cor-
rosion can take several forms, but all of them are electrochemical in
nature. That is, the process involves the simultaneous transfer of
electrical charges in the metal, and ions in the electrolyte (seawater
in this case) into which the metal is placed.
The corrosion process takes place at two distinctly different types
of sites on the metal surface. At the first of these sites, called the
anode, chemical oxidation takes place, and metal ions leave the surface
of the metal to enter the solution as illustrated in Figure 1. It is
this loss of metal ions at the anode that leads to structural damage of
the metal during corrosion. Of equal importance is the second site,
called the cathode, where chemical reduction takes place and dissolved
oxygen from the seawater is reduced to hydroxyl ions (OH ). As iron
I i
atoms are oxidized to Fe at the anode they liberate electrons which
travel through the metal and are consumed in the reduction at the
cathode (Figure 1). A simple chemical equation describes each of these
reactions as follows:
— I I
At the anode: 2Fe - 4e -> 2Fe
At the cathode: 02 + 2H20 + 4e~ -> 40H
Combining the two, we can write the overall reaction as:
2Fe + 0 + 2H 0 -> 2Fe"H" + 40H~ -> 2Fe(OH>2
The product of this reaction, Fe(OH ), is a hydrated ferrous oxide or
-------
4 OH
FIGURE 1 Electrochemical reactions at anodic
and cathodic sites on structural steel corroding
in seawater.
-------
ferrous hydroxide. It is insoluble and forms a loosely adherent green
or greenish black surface film on steel in seawater. The outer surface
of the film, which is in contact with a ready supply of dissolved
oxygen, is further oxidized to ferric hydroxide by the reaction:
2Fe(OH) + H20 + 1/2 C>2 + 2Fe(OH)3
This is orange to reddish brown in color and is familiar to us as common
rust.
Structural Steel
On plain carbon structural steel, the anode and cathode areas
described above shift about continuously on the metal surface. The
result is called "uniform corrosion". The surface becomes generally
rusted and the metal thickness is reduced almost uniformly over the
entire exposed surface. It is this type of corrosion to which the outer
steel drum of past nuclear waste packaging containers is subjected.
Uniform corrosion of structural steel in seawater will be accelerated if
the steel is brought into direct contact with a more noble (cathodic)
metal such as a copper alloy, stainless steel, or titanium. This type
of attack, called galvanic corrosion, or electrolysis, will be described
in the following paragraphs.
When two different metals are placed in direct contact in seawater,
a difference in electrochemical potential exists between them. This
potential difference causes a corrosion current to flow. The more
active of the two metals becomes the anode and begins to corrode more
rapidly than it would by itself. The more noble metal becomes the
cathode and is protected from corrosion. The accelerated corrosion on
the less noble (anodic) metal is called galvanic corrosion. Table I
2
shows a "galvanic series" of common metals and alloys in seawater.
-------
TABLE I
GALVANIC SERIES OF METALS AND ALLOYS
IN SEAWATER
Alloy
Potential vs.
Satd. Calomel Electrode
ANODIC
END
Magnesium
Zinc
Aluminum Alloys
Cast Iron
Steels (structural and alloy)
Stainless Steels (active-i.e. in crevices & pits)
Al - Bronze D
Naval Brass
Red Brass
Copper (ETP)
Inhibited Admiralty Brass
Manganese Bronze A
Silicon Bronze
90-10 Copper Nickel
Type 400 Stainless Steels (passive)
Lead
70-30 Copper Nickel
17/4 pH Stainless Steel (passive)
Silver
Monel 400 (Nickel-copper alloy)
Type 300 Stainless Steels (passive)
Stainless Alloy 20 cb3
Titanium and Titanium Alloys
Inconel 625
Hastelloy C-276
Stainless Alloy 6X (passive)
Tantalum
Platinum
Graphite
CATHODIC
END
-1.60 to -1.63V
-0.98 to -1.03V
-0.70 to -0.90V
-0.60 to -0.72V
-0.57 to -0.70V
-0.35 to -0.57V
-0.30 to -0.42V
-0.30 to -0.40V
-0.20 to -0.40V
-0.28 to -0.36V
-0.25 to -0.34V
-0.25 to -0.33V
-0.24 to -0.27V
-0.21 to -0.28V
-0.20 to -0.28V
-0.19 to -0.25V
-0.13 to -0.22V
-0.10 to -0.20V
-0.09 to -0.14V
-0.04 to -0.14V
-0.00 to -0.15V
+0.05 to -0.15V
+0.06 to -0.05V
+0.10 to -0.04V
+0.10 to -0.04V
+0.32 to -0.15V
about +0.2V
+0.35 to +0.2V
+0.3 to +0.2V
-------
For any combination of metals, the galvanic series can be used to
predict which metal will suffer accelerated attack (anode) and which
metal will be protected (cathode). If any two metals or alloys from
Table I are placed in contact in seawater, the one higher on the series
will be the anode, and the severity of the damage to it will be directly
proportional to the magnitude of the potential difference between the
two metals, or simply the separation of the two metals or alloys on the
chart. Two examples illustrate this point. If steel is coupled to
Monel 400 (nickel-copper alloy), the steel will be anodic and will be
damaged severely because the two alloys are far apart on the chart. If
copper and silicon bronze are coupled, damage will be slight because
they are close together on the chart; that is, the two metals have a
small potential difference.
It is possible to take advantage of galvanic corrosion by coupling
a material one wishes to protect from corrosion to a more active metal.
For instance, steel can be protected from corrosion in seawater by
coupling it to zinc. The zinc in this case is called a sacrificial
anode, and the steel will be protected as long as any zinc remains. The
protection can be extended indefinitely by periodically replacing the
zinc anodes if the structure is accessible for maintenance.
Stainless Steels
The behavior of stainless steels is quite different from that de-
scribed above. Stainless steels contain 12% or more of chromium, which
make them "passive," or resistant to uniform corrosion. Their passivity
in seawater is caused by the formation of a tightly adherent surface
film of metal oxide. Other alloying elements in addition to chromium
are often added to make this passivity more stable in a wide range
-------
of environments or to enhance the mechanical and physical properties, or
both. Nickel and molybdenum are two alloying elements commonly added to
stainless steels intended for seawater service. The class of stainless
steels most frequently used around marine environments in the past, the
300-series stainless steels, has approximately 16 to 18% chromium, 8 to
10% nickel and up to 3% molybdenum.
When used in seawater, stainless steels resist uniform corrosion
but the chloride ions in seawater tend to cause breakdown of the passive
O /
film at discrete sites on the metal surface. ' The underlying metal at
these sites becomes anodic, while the vast maj.ority of the surface
remains passive and becomes cathodic. Once formed, the anode and
cathode areas do not shift location, as they do on plain carbon steel,
but remain fixed. In addition, the high concentration of metal chlo-
rides that builds up at each anodic site causes the pH of the electro-
lyte at the anode to become more acidic, and accelerates the rate of
3 4
attack at these sites. '
The result is a nigh rate of penetration of the metal at the anodic
sites while the rest of the surface remains unattacked. This mode of
corrosion, often leading to a series of small holes which perforate the
metal, is called "pitting." As a general rule, the fewer pits there are
on a given surface, the larger and deeper they will be. As the number
of pits increases, the individual pits become smaller and shallower.
Stainless steels are also susceptible to another form of localized
attack called crevice corrosion. It takes place within crevices found
around threaded fasteners, under washers, gaskets, and 0-rings, between
the faces of riveted lap joints, and similar places of restricted
geometry that are shielded from the water movements that continually
-------
wash the remainder of the surface with fresh, oxygenated seawater. The
water inside the crevice becomes stagnant and depleted in oxygen, and
the surface in contact with this water becomes the anode, while the
remaining surface outside the crevice is the cathode. The mechanism is
similar to that of pitting. The pH inside the crevice becomes acidic
and the anode reaction becomes self-accelerating. Corrosion starts on
the interior surfaces and hollows out the part, as illustrated in Figure
2, until its strength becomes insufficient and it fails.
Factors Affecting Deep Ocean Corrosion Rates
The parameters generally recognized as having an influence on the
corrosion rates of steel and stainless steels in seawater are: dis-
solved oxygen, temperature, salinity, pH, relative velocity of the
O / 7
water, and the presence of fouling organisms. ' " Figure 3 shows the
typical variability of some of these parameters with depth in the
Northeastern Pacific Ocean.6 Let us consider their effects first on
plain carbon structural steel and then on stainless steels. For pur-
poses of this discussion, the deep ocean shall be defined as any depth
greater than 2000 meters.
On structural steels, the corrosion rate is a direct function of
the dissolved oxygen concentration of the seawater. This effect is
so strong that it tends to overshadow everything else. The corrosion
rate of steel is also a direct function of temperature. When oxygen
concentration is the rate controlling factor, as it usually is in sea-
water, the corrosion rate of steel is estimated to be halved by each
30°C decrease in temperature.4 Nevertheless, the influence of tem-
perature on corrosion rates is weaker than that of dissolved oxygen. An
-------
10
CORROSION OF
BOLT HEAD
CORROSION OF
WASHER
CORROSION OF
BOLT THREADS
CORROSION
OF NUT
FIGURE 2. Crevice corrosion sites around a mechanical fastener.
-------
11
surface r*
500
— 1000
a.
LU
a
1500
2000
SALINITY
0
8
Oxygen, ppm
0
68 10 12
i Temperature, °C
14
16
_L
18
6.4 6.6
6.8
I
7.0
I
7.2
I
7.4 7.6 7.8 8.0 8.2
I
33.0 33.2 33.4
33,6 33.8 34.0 34.2 34.4 34.6 34.8
Salinity, ppt
FIGURE 3. Variability of oceanographic parameters in
the Pacific Ocean off Port Hueneme California (After Reinhart6)
-------
12
example of the overriding effect of dissolved oxygen in the deep ocean
is shown in Figure 4, where the corrosion rate, expressed as mils per
year (mpy), increases again with increasing dissolved oxygen concen-
tration at greater depths despite a continual decrease in temperature.6
The salinity and pH of seawater have only a small effect on the
corrosion rate of steel. When considered as a function of salinity, the
corrosion rate reaches a maximum at: a value of about 35 °/oo (parts per
thousand) salt content and decreases gradually at higher and lower
4
values. This maximum is the result of two competing tendencies.
First, the conductivity of seawater increases with increasing salinity,
resulting in a greater separation between the anode and cathode areas,
and consequently a less protective corrosion product (rust) film.4 This
accounts for the initial increase in corrosion rate with increased
salinity. Second, the solubility of dissolved oxygen decreases with
increasing salinity. This latter effect becomes predominant at about 35
/oo salinity, thus accounting for the gradual decrease in corrosion
4
rate at higher salinity. Over the normal range of open ocean salin-
ities, however, the changes in corrosion rate with salinity are so small
as to be completely overshadowed by the effects of dissolved oxygen.5"7
Within the normal slightly alkaline (pH 7.4-8.4) range of open ocean pH,
the corrosion rate of steel is independent of pH. Below pH 4, the
corrosion rate increases rapidly with increasing acidity. Above pH 10
the corrosion rate decreases, and approaches zero at pH values greater
than 12.4
At water velocities exceeding 3 to 4 meters per second, the cor-
2
rosion rate of steel rapidly increases. This is especially true in
cases where there is appreciable suspended particulate matter in the
-------
13
SURFACE
DISSOLVED OXYGEN
CARBON AND LOW
ALLOY STEELS
-c 1,000
A AISI 1010 STEEL
2,500
OXYGEN (ml/l)
or
CORROSION RATE (mpy)
FIGURE 4. Corrosion rate of structural steels as a
function ot depth and dissolved oxygen. (After Reinhart6).
-------
14
seawater that might lead to abrasion of th€
corrosion product films. Deep ocean currents,
much smaller velocities than this (3 to 30 cm per second is typical8)
and are not expected to play a substantial role in determining the
corrosion rate.
Macroscopic marine fouling organisms do not usually have an adverse
effect on the corrosion rate of structural steels. In fact, a complete
cover of hard-shelled sedentary (sessile) fouling organisms may decrease
the corrosion rate by acting as a barrier through which dissolved oxygen
must diffuse to reach the metal surface.4'6'7 Such a complete layer of
organisms is rare in the deep ocean. There are fewer biofouling organisms
present at great depths, and those that are present near the bottom tend
to attach as opportunistic individuals, causing only a slight pertur-
bation of the corrosion rate in their immediate vicinity.9'10 Fouling
is expected, therefore, to have only a minor effect on the overall cor-
rosion rate of steels in the deep ocean.
Marine microorganisms can influence the corrosion rate of steels
only in restricted areas where they can change the chemistry of the
surroundings. For instance, the corrosion rate of steel can be appre-
ciable in anaerobic bottom sediments if sulfate-reducing bacteria are
present. The mechanism of this type of attack will be considered later
in this report.
These same parameters have different effects on the corrosion rate
of stainless steels, which corrode by pitting and crevice corrosion
rather than by uniform attack.4'6 The factors that influence these
localized forms of attack are primarily those that allow or discourage
differences in the environment from place to place on the metal surface.
-------
15
Dissolved oxygen has a major effect only when differences in it are per-
mitted to develop along the metal surface. This happens most readily in
stagnant or zero velocity conditions or when crevices, such as shown in
Figure 2, are built into the mechanical design. Seawater velocities of
1.5 to 2 meters per second can suppress the initiation of pitting
corrosion altogether by keeping the surface uniformly immersed in oxy-
2
genated water. Water velocity has little effect on crevice corrosion,
however, because the corroding surface is shielded from the flow of
water.
The above effects are in direct contrast to the case for plain
carbon steels, where an increase in velocity brings more oxygen to the
corroding surface and increases the corrosion rate. Another direct
contrast to the case for plain carbon steels is that of fouling organ-
isms. On plain carbon steels these organisms tend to decrease the
corrosion rate by acting as a diffusion barrier for dissolved oxygen.
On stainless steels, however, unless the fouling layer is complete and
uniform, it creates localized oxygen-shielded regions and promotes
crevice corrosion beneath the organisms.
Temperature has the same effect on stainless steels that it has on
plain carbon steels, and to a first approximation, the salinity and pH
of seawater have little effect on the rate of pitting and crevice
corrosion of stainless steels.
Historically, when stainless steels have been present in low-level
nuclear waste packages from ocean disposal, they have been as part of
the waste form itself embeddec in the concrete. Thus, the environment
to which stainless steels have- been subjected is neither open ocean
seawater nor bottom sediment, but rather seawater-saturated concrete,
-------
16
the pH of which is normally 32 to 14. This alone should suppress the
initiation of pitting, because- hydroxyl (OH ) ions have been found to
4
be effective inhibitors for pitting in stainless steels. The alkaline
pH will also inhibit corrosion if the core is made of structural steel.
As long as the dissolved oxygen level is maintained uniformly low and
the pH uniformly high over the- entire surface of the metal, very little
corrosion might be expected whether the metal is made of structural
steel or stainless steel.
There are two ways in which this desirable low oxygen, high pH
environment may be lost. The first way is if the concrete itself
deteriorates according to one of the mechanisms discussed in the next
section. If the concrete becomes fractured and allows oxygenated sea-
water to contact part of the metallic waste form, corrosion will start.
For stainless steel, the exposed area will become cathodic and the re-
maining surface still shielded by the concrete will be susceptible to
crevice corrosion. For structural steel, uniform corrosion will start
on the surface exposed to ambient seawater and the wedging action of the
corrosion products, which are more voluminous than the steel from which
3
they form, will tend to pry the concrete away and progressively expose
more steel.
The second way the desirable environment can be lost involves
changes in internal chemistry that; occur as the concrete absorbs sea-
water over a period of time. Recent evidence indicates that as the
pores (or voids) in the concrete become saturated with seawater, chlo-
ride ions from the absorbed seawater break down the protective film on
the metal surface and initiate corrosion. As corrosion proceeds, the
-------
17
metal-chloride corrosion products, which are acidic, both reduce the pH
of the pore water, and produce the wedging action mentioned above. Both
these effects tend to produce still more corrosion.
2.2 DETERIORATION OF CONCRETE
The primary mechanism of degradation of portland cement type
12
concretes in seawater is sulfate attack. In waters with more than
about 2000 ppm sulfate, destructive chemical reactions can take place
12 13
with the cementing constituents of the concrete. ' Since the average
concentration of SO ions in the Atlantic Ocean is 2810 ppm, this type
12
of attack can and does take place.
Destructive attack can follow either or both of the following two
12
modes: (1) cracking due to the formation of a calciura-sulfoaluminate
hydrate called ettringite, or (2) surface softening due to the formation
of gypsum. The first type of attack can take place in portland cements
containing more than 5% of tricalcium aluminate (3CaO'Al?0^). The
sulfates react with hydrated 3CaO*Al 0 in the presence of calcium
hydroxide (Ca(OH) ) to form ettringite (3CaO-Al 0 •3CaO-SO •32H 0).
Ettringite has a strong tendency to absorb water. This causes it to
expand and results in cracking of the concrete. In the second type of
attack, acidic sulfates cause the conversion of Ca(OH)? to gypsum
(CaSO *2H O"1 according to the following reaction:
Ca(OH) + MgSO -7H 0 -> Mg(OH) + CaSO •2H 0 + 5H 0
This is accompanied by surface softening and spalling of the concrete.
The types of attack discussed above are aggravated by high porosity as
well as by high 3CaO*Al 0 and Ca(OH) contents in the cement phase.
-------
18
3.0 ENVlROraffiNTALJDATA
3.1 STRUCTURAL STEEL
There is a considerable volume of data on the corrosion of plain
carbon structural steel in natural seawater. The bulk of these data are
for shallow coastal water exposures. In general, the shallow water
data show initial uniform corrosion rates as high as 380 micrometers per
year (llm/y), decreasing with tirae to a constant value of 75 to 125 Mm/y
after two or three years of exposure. 2>5'«-« Jhe ^ ^ ^ ^
uniform corrosion rate decreases depends upon how quickly the barrier
film of corrosion products, carbonaceous deposits, and fouling organise
builds up, and upon how effective this film ls in preventing dissolved
oxygen from reaching the bare metal surface. Once this film has forced,
factors which may cause the corrosion rate to vary outside the 75 to
125 Wy limlts are pollution (especially by sulfides) , high water
velocities, or variations in water temperature. In addition, very low
dissolved oxygen can reduce the rate to less than 75 pm/y.
The corrosion is not always entirely unifo™. Located attack In
the for. of broad shallow pitting has frequently been reported with
penetration rates of up to 10 tu.es the uniform corrosion rate.
Occasionally, a very deep p,t WLU be observed. For instance, a pit
with a depth of 4 mm was observe,, on one panel that had been exposed
for 16 years in the Pacific Ocean at the Panama Canal Zone.17 The
Plttlns type of attack tends to be more severe in polluted seawater
and on material from which the mill scale is not cleaned off properly/''
1" the deep ocean, which is the primary focus of this report, the
corrosion rates of ,,laln carbon steels
i, surface waters.5'^
-------
19
waters, as it does in shallow water, starting at 50 to 100 ym/y and
decreasing to 12 to 50 ym/y. Several factors favor lower corrosion
rates in the deep ocean. The most prominent of these is low dissolved
oxygen. The decrease in dissolved oxygen with depth is more pronounced
in the northeastern Pacific Ocean than it is in the northwestern
Atlantic, where the minimum in dissolved oxygen is generally not as low,
and where the value at great depths can be as high as, or higher than it
5 18
is at the surface. ' Table LI shows that for plain carbon steel ex-
posed at similar depths in the Atlantic and Pacific Oceans, the corrosion
rate is doubled in the Atlantic because of the higher dissolved oxygen.
For a given oceanic site, therefore, the concentration of dissolved
oxygen must be known before any prediction of the corrosion rate can
be made.
Other factors favoring low corrosion rates in the deep ocean are
low water velocities and low temperatures. The decrease in corrosion
rate with time is usually more gradual in the deep ocean than it is
in surface waters because protective surface films form more slowly
under deep ocean conditions. The low temperature and more acidic pH
5 19
retard the rate of deposition of carbonaceous type mineral scales, '
9 10
and fouling films form slowly Lf at all. '
TABLE 11
CORROSION RATE OF STEEL AFTER 3 YEARS EXPOSURE
Location Depth Temperature
Pacific Ocean 1675 m 6.65°C
Atlantic Ocean 1515 m 4.L8°C
Dissolved
Oxygen
1.5 ml/ 9,
5.7 ml /I
Corrosion
Rate Reference
25 ym/y 14,15
50 ym/y 14
-------
20
Since the low-level nuclear waste containers that have previously
been deposited in the ocean are sitting on the bottom partially embedded
in the sediments, corrosion rates in these sediments are an important
consideration. Generally, corrosion rates of steel in the sediments
will not differ greatly from those in the water immediately above
the sediments.5 However, there are several factors that can cause
corrosion rates in the sediments to be different from those in the
water. These factors include: (a) a difference in concentration of
some chemical species, such as dissolved oxygen, between the sediments
and the water, (b) the presence or absence in the sediment of sulfate
reducing bacteria under anaerobic (no dissolved oxygen) conditions and,
(c) the presence of bottom currents strong enough to keep appreciable
particulate matter in suspension. Each of these factors will now be
discussed in more detail, first for an unpainted steel cylinder, and
then for a painted steel cylinder similar to the radioactive waste con-
tainers under consideration.
The dissolved oxygen concentration is often lower in the sediments
than in the water. This difference produces what is commonly called an
oxygen concentration cell. Whenever such a concentration difference
exists, it sets up a potential difference resulting in the area of
lower oxygen concentration becoming anodic and the area of higher oxygen
concentration becoming cathodic. For an unpainted steel cylinder, the
portion of the surface exposed to the lower oxygen concentration (i.e.,
the portion buried in the sediments) would become anodic and suffer a
moderately accelerated rate of attack compared to that portion exposed
4
directly to the water.
-------
21
If dissolved oxygen is completely absent from the sediments, the
4,20
corrosion rate of steel in the sediments would normally be negligible
but may be accelerated in the presence of anaerobic, sulfate-reducing
bacteria. Such bacteria are almost universally found in the sediments
under shallow natural waters, where the dissolved oxygen is used up by
20
decaying organic matter. In deep sea conditions where the amount of
decaying organic matter may be very low, however, the presence of
sulfate-reducing bacteria may be more variable. In the presence of
these bacteria, corrosion rates of steel can be increased by as much as
several orders of magnitude. Under anaerobic conditions, the cathodic
reaction shifts from the reduction of oxygen described earlier (Section
2.1) to reduction of hydrogen according to the reaction:
H+ + e~ -> H°
Without sulfate reducing bacteria present, the cathodic surface quickly
becomes covered with a monolayer of neutral hydrogen atoms (H°), and
corrosion stops. Sulfate reducing bacteria, however, strip off the
hydrogen layer, thus allowing corrosion to continue, a process called
21
cathodic depolarization. The bacteria utilize hydrogen in their
metabolism to reduce sulfates from decaying organic matter to sulfides.
The sulfides, in turn, form an iron-sulfide scale on the steel which is
cathodic to a bare steel surface. The sulfide scale thus leads to an
20
additional galvanic corrosion of the bare steel. Sulfate-reducing
bacteria are rendered inactive, and their effect on corrosion is
stopped, by the presence of even trace amounts of dissolved oxygen.
Thus, the mechanism of corrosion described above requires both com-
pletely anaerobic conditions and a source of sulfates.
-------
22
For an unpainted steel cylinder, the differences discussed above
between the water and sediments will cause accelerated corrosion of
that portion of the steel exposed to the sediments. For a painted
steel drum, however, the inverse may be found due to the presence of
the paint itself. The paint exposed to the water is in a more dynamic
environment and may deteriorate more quickly than that exposed to the
sediment, due to surface abrasion and biological fouling. This allows
corrosion to begin more quickly on the surfaces of the druir exposed to
seawater, while the sub-surface paint coating continues to protect the
drum. This appeared to be the case for the container recovered from
the northwestern Atlantic Ocean during the 1976 Atlantic dumpsite survey.
It suffered very little corrosion on the portion of the cylindrical
surface that had been buried in the sediments because the paint coating
22
remained in good condition. Coatings, however, cannot be made perfect
enough to provide dependable long-term (greater than about 15 years)
protection as will be discussed later in this report.
Higher rates of corrosion can also occur just above the bottom
sediments if the seawater currents close to the bottom are sufficiently
strong to keep appreciable paniculate matter in suspension. In this
case, the sediment particles gradually abrade away corrosion product
and mineral scale or paint films, thus constantly exposing fresh bare
metal. This often happens in shallow coastal waters as a result of
tidal current or wave action. It would be unusual to encounter deep
ocean currents of sufficient magnitude to cause such abrasive action.
However, turbidity currents set up by underwater landslides moving
large volumes of sediment at considerable velocity may produce this
abrasive action.
-------
23
In light of the above discussion, the important factors to be
considered when evaluating corrosivity to conventional structural
steel containers placed in deep sea dumpsites are: the dissolved
oxygen concentration in the water and sediments; the presence or
absence of sulfates and sulfate-reducing bacteria; and bottom currents
or turbidity currents.
3.2 STAINLESS STEELS
Stainless steels are not susceptible to uniform corrosion in
seawater. They corrode by localized breakdown of their protective
passive film at pits and crevices as discussed in an earlier section
of this report. The resulting penetration of the surface is deeper
and more rapid than it is on plain carbon steel.
Many types of stainless steel are available today, and a number
of these perform well in both the marine atmosphere and in the splash
zone above mean high tide level. Nearly all of the conventional alloys,
however, perform poorly when fully submerged in quiet seawater for periods
exceeding one or two months. ' ' Several recently developed stainless
alloys have superior performance when fully submerged in seawater. These
alloys will be discussed in a subsequent section.
With one exception, it has not been possible to identify which
of the available stainless st.eel alloys have historically been disposed
of as nuclear waste in the ocean. That exception is the pressure vessel
of the N/S Seawolf reactor, which was dumped in the Atlantic 2800m
nuclear waste dumpsite in 1959, and was constructed of Type 347
0 ^
stainless steel. For the purposes of this report, therefore', it
will be assumed that only tht- more resistant of the commonly available
-------
24
grades (300-series stainless steels) were used and our discussion will
be limited to stainless steel alloys 302, 303, 304, 316, and 347. The
nominal compositon of these alloys is given in Table III.
As explained in an earlier section, the environment surrounding the
inner stainless steel waste form is initially seawater-saturated concrete
at a pH of 12 to 14 rather than ambient seawater itself. There are
no long-term data available for seawater corrosion of stainless steel
in concrete, but it is expected that as long as the concrete remains
intact and the chloride ion concentration remains low, corrosion rates
will be low due to the uniform alkaline environment. If the concrete
should spall or crack, however, and expose the stainless steel to ambient
seawater, or if chlorides become sufficiently concentrated in the pore
water, accelerated attack on the stainless steel may take place.
TABLE III26
NOMINAL COMPOSITIONS OF 300-SERIES STAINLESS STEELS
Alloy
% Cr
% Ni
% Fe
% Other Major
Alloying Elements
Comments
302
303
304
316
17-19
17-19
18-20
16-18
8-10
8-10
9-12
10-14
Balance
Balance
Balance
Balance
-
0.2 S or Se Free machining
grade, least
resistant .
-
2 - 3 Mo Most resistant
of group
347
17-19
9-13 Balance
1 - 2 Mn
Weldable grade
-------
25
All the 300-series stainless steels suffer rapid localized attack
by pitting and crevice corrosion in quiescent seawater. Rates of pene-
tration in shallow water as high as 11.5 mm/y by pitting and 1.3 mm/y by
5 27
crevice corrosion have been reported. ' In the deep ocean, the
penetration rates tend to be lower (up to 4 mm/y by pitting and 0.3
to 0.75 mm/y by crevice corrosion) due to the lower dissolved oxygen
5 27
and temperature. ' The rate of attack can be strongly influenced by
seawater velocity. An increase in seawater velocity inhibits pit
initiation but accelerates pit growth after the pit has already formed.
But a stainless steel form encased in concrete would be shielded from
such effects. The rates cited above could be considerably higher if
less corrosion resistant grades of stainless steel such as the 400-series
were involved.
Crevice corrosion can be avoided by welding parts together rather
than using threaded fasteners. Unless the welding is done properly,
however, the material can be made sensitive to another type of corrosion
called intergranular attack or weld decay. This occurs in the heat-
affected zone of the weldment. While the metal is at high temperature
during welding, the chromium in the alloy is scavenged out of solid
solution by an internal precipitation reaction, leaving a network of
chromium-rich particles embedded in an alloy that is no longer stainless
(because of the depleted chromium). Corrosion then proceeds rapidly
in the depleted regions on either side of the weld, while the remainder
3
of the part remains unaffected. Intergranular attack can be minimized
by using specially stabilized grades of stainless steel such as types
304L (low-carbon), 316L, 321 and 347 for any part where welded joints
-------
26
are anticipated. These stabilized grades have resistances to pitting
and crevice corrosion comparable to other type 300-series stainless
steels but are more expensive because of the addition of stabilizing
elements such as tantalum and columbium (Types 321 and 347), or because
of a more complex manufacturing process (Types 304L and 316L). Because
of the difficulties discussed above, the conventional 300 series stain-
less steels are seldom recommended for applications in seawater re-
quiring full immersion for durations exceeding about six months, unless
some corrosion protection measures are feasible.
3.3 CONCRETE
The rate of deterioration of concrete in seawater varies with the
formulation of the material. Most of the published data on deteriora-
12
tion rates comes from blocks of concrete immersed in Los Angeles Harbor
in 1905 and subsequently inspected after 27 and 67 years, and a similar
28
series of tests near Trondheim Harbor, Norway, started in 1936 to
1943. The sulfate type of attack described earlier leading to cracking
or surface softening was most severe in tests on high permeability
concretes where losses in compressive strength of up to 10% were ob-
served.12'28 The sulfate type of attack was negligible for high quality
formulations using portland pozzolan or portland blast-furnace slag
cements of low permeability, low alkalanity and low 3CaO-Al20^ content.
Losses in compressive strength due to sulfate attack and/or water ab-
sorption in these high quality concretes rarely exceeded 10% and a
1 '•> 90
strength gain was often reported. ""' Damage by rock-boring mollusks
was limited to superficial surface scouring if a granite aggregate was
used.
-------
27
The factors of most significance to the problem of nuclear waste
containment are water absorption and permeability. Most concretes
tested at depth were estimated to absorb a total of 1 to 3 percent by
weight of water over an extended period of time (several months or
29
more). Superimposed on this absorption, all concretes were permeable
to seawater to some extent. Permeabilities have been calculated from
the changes in buoyancy with time for hollow concrete spheres moored at
27 30
various depths in the ocean. ' This permeability to seawater leaves
open the possibility of a hollow concrete container eventually becoming
filled with seawater and the subsequent leaching of contaminated water
back into the environment. For these reasons, unmodified concrete alone
should not be relied upon to isolate even low-level waste for a period
as long as 300 years, whether the waste is mixed in with the wet con-
crete, or placed in a cavity inside a concrete shell.
3.4 EFFECTS OF RADIATION
There are no data in existence on the effects of either high or
low-level radiation on corrosion rates in the deep ocean. The data that
are available in the open literature are concerned with the effects of
3] 32
radiation on corrosion rates in nuclear reactor environments. ' For
stainless steels, these data indicate that radiation generally does not
'52
accelerate the corrosion rate." There can even be a slight retardation
of the corrosion rate of stainless steels if the radiation intensity is
large enough to break down the water surrounding the corroding metal.
Breakdown of the water provides more oxygen for repairing the passive
32
film, thus inhibiting corrosion. The opinions relative to the effects
32 33
of radiation on plain carbon steel are mixed. * There is some evi-
32
dence that the corrosion rate is accelerated under high neutron flux,
-------
28
but this is a much different environment than that to which a low-level
radioactive waste container would be exposed in the deep ocean.
It is probably safe to conclude that the radiation intensity pro-
duced by low-level waste is unlikely to cause a significant increase in
the corrosion rate of either plain carbon or stainless steels in the
deep ocean where the temperature remains low and the pressure high.
Under deep ocean conditions, it would take a level of radioactivity
approaching that found in and immediately around nuclear reactor cores
to cause a change in the corrosion rate.
-------
29
4.0 LIFETIME OF CONVENTIONAL CONTAINERS
Based on the data presented in the previous section, an attempt
will now be made to estimate the useful and reliable lifetime over which
one can confidently predict that the conventional type of low-level
nuclear waste container made from steel, concrete and stainless steel
will be able to isolate the waste from man and his food chain in the
ocean. For the purposes of this report, the criterion of failure for
any given layer of the container will be that it either allows water to
contact the next layer in such a way that it will start to deteriorate,
or it allows water to contact the waste itself. In order for the en-
vironment to be contaminated, water must not only reach the interior of
the container but must also find its way back out. The time scale of
this process depends upon the failure mode of the container, and can
range from a few seconds for a container that fails by implosion to
several decades or more for one that fails by seepage (leaking or
leaching) through concrete and/or by corrosion of a metallic inner waste
form. In the following discussion, the entire container will be con-
sidered to have failed as soon as it allows water to contact the waste,
and the time for the contaminated water to return to the environment
will be neglected.
In some cases, there will be so many unknown factors that it will
be impossible to accurately estimate the lifetime. In such instances,
the only prudent course of action will be to identify the worst possible
combination of conditions and the most rapid deterioration mode, assume
that if there are a large number of containers these worst-case con-
ditions will occasionally be met, and base the estimated lifetime on
this worst case.
-------
30
Consider first the outermost mild-steel shell. If we assume a
worst-case deep-sea corrosion rate of 100 ym/y for the first year
decreasing to a constant 50 ym/y thereafter, a 1.5 mm wall thickness
will be entirely consumed in 29 years if the corrosion is uniform.
Usually> however, there is some broad shallow pitting in which pene-
tration rates are up to 10 times faster than the uniform corrosion rate.
Thus, the wall could be penetrated locally in as little as about three
years. In most instances, the steel will last considerably longer than
this, as was the case for the containers recovered during the 1976
22
Atlantic dump site survey. But there is no assurance that this will
be the case for all containers.
In actual practice, the steel was not meant to isolate the concrete
at all as one end was left open. Even if the end were closed, however,
one should not rely on the mild carbon steel to keep seawater from the
concrete for more than three years.
Consider next the layer of concrete. It has been estimated that an
originally dry, hollow sphere with 15 cm thick walls of high quality
concrete would be penetrated by seawater (as evidenced by liquid drop-
lets apppearing on the inside surface) in one to two months under a
34
pressure head corresponding to a depth of 3,660 meters. It has also
been estimated that the same concrete wall would be saturated with
o /
seawater within six months. By the present definition of failure,
therefore, a conventional type container having nonmetallic waste mixed
in concrete, and the concrete exposed to seawater on one end would have
failed by allowing water from the environment to contact the waste
within two to six months after deployment. If the open end of the outer
steel container were closed, the time to penetration could be extended
-------
31
by three years, but this is still short compared to the desired 50- to
300-year isolation period.
For containers with structural steel or stainless steel inner waste
forms, the initial saturation of the concrete by seawater would not
constitute failure because it would not immediately create a situation
in which the structural steel or stainless steel would begin to corrode.
Corrosion may start shortly thereafter, however, if chloride ions become
sufficiently concentrated in the pore water of the concrete, or if the
concrete deteriorates further by cracking, spalling or sulfate attack.
Cracking and/or spalling of the concrete can sometimes take place
35
even before deployment of the container due to shrinkage during curing.
This problem can be aggravated by the almost inevitable rough shipboard
handling which is well known to all seagoing personnel. The outer steel
drum helps to minimize such damage, but even with special handling
precautions, one would expect the concrete in a small percentage of
conventional type containers to be cracked prior to disposal.
Sulfate attack takes place gradually over a time period of tens of
-i 9 10 90
years. ' ' Based on our accumulated experience as outlined in
section 2.2, one would not expect a 15 cm wall thickness of high quality
concrete, in the absence of chloride ions, to be sufficiently damaged by
sulfate attack to allow an underlying steel or stainless steel waste
form to corrode for at least 75 years, and perhaps as long as the
desired 300 years. The difficulties are that (1) in the configuration of
conventional low-level nuclear waste containers, the cracking problem
may occasionally render the concrete prematurely useless as a corrosion
barrier for any interior metals, and (2) not enough is currently known
about the chloride ion problem to allow us to predict how long it will
-------
32
take to initiate corrosion. Almost certainly, however, the time will be
short compared to even the shortest desired containment period (50
years).
Let us now examine what might be done to increase the reliability
of these containers in their conventional configuration. The methods
available for increasing the lifetime of marine structures fall into
four categories. These are: (1) isolation of the structure from the
environment by coatings, (2) elimination or control of corrosion by
cathodic protection, (3) use of chemical inhibitors (which can be ruled
out in the present open ocean situation because of the volume of in-
hibitor that would be required) and (4) use of materials more resistant
to deterioration, which will be taken up in the following section on
alternative packaging methods.
Coatings protect against corrosion either by isolating the metal
O £
substrate from the environment, or, in the case of metallic coatings,
by creating a galvanic corrosion couple where the coating is sacrificial
9 (->
to the substrate. Nonmetallic marine coating systems have been
improved over the last 10 to 15 years, but none of the currently
available systems are capable of protecting the substrate from seawater
under fully submerged conditions for more than about 10 years without
periodic maintenance.
Sacrificial metallic coatings such as zinc (galvanizing) have been
used to protect steel successfully in seawater for short periods of
time. As a rough rule of thumb, a zinc coating 0.025 cm thick is
sufficient to protect steel from corrosion in seawater for one year.
Heavier coatings can extend the protection to two or three years. As
was the case with nonmetallic coatings, however, longer-term protection
-------
33
requires periodic maintenance, and this is not practical for a structure
sitting unattended on the seafloor. Thus, coatings are considered
capable of extending the lifetime of the present type of low-level waste
container for only about 10 to 15 years.
Cathodic protection is the most effective means available for
controlling metallic corrosion. It is the only way to reduce the
corrosion rate to zero and maintain it that way over an extended period
4
of time. Cathodic protection is achieved by applying an external
electrochemical current to the corroding system. The polarity of the
external current is opposite to that of the natural corrosion current.
The applied current opposes the tendency for metallic ions to enter the
4
seawater, and if made strong enough, will stop corrosion altogether.
The applied current must be supplied by either an external electronic
power supply or by a sacrificial anode such as aluminum, zinc or mag-
nesium. Unfortunately, either type of system requires that the struc-
ture be accessible for periodic: maintenance. External power supplies
need a source of current and must be monitored periodically to make sure
that the proper amount of current is being supplied. Too little current
allows some corrosion to take place, while too much current can damage
the structure being protected. Galvanic anodes become exhausted and
must be replaced every one to three years. For these reasons, cathodic
protection is not a practical way of controlling corrosion on structures
which must be able to survive for more than a few years unattended in
the deep ocean. Thus, it is not expected that cathodic protection will
contribute significantly to extending either the average lifetime or the
reliability of conventional low-level nuclear waste containers in the
ocean.
-------
34
Based on the discussions above, the performance of an average low-
level nuclear waste container as previously configured for ocean dumping
is expected to be as follows. Starting from the open end of the con-
tainer, the concrete should become saturated with seawater within about
six months. The condition of the concrete then becomes critical. For
most of the containers in which the concrete is not cracked or other-
wise damaged, the inner waste form, be it structural steel or stainless
steel, should not start to seriously degrade for 75 years or longer.
The careful application of marine coatings might extend this time by
about 15 years. It is even conceivable that a substantial portion of
the containers might succeed in isolating their waste from the environ-
ment for the desired 300-year period. It is probable, however, that the
concrete in a small percentage of containers will become cracked before
disposal in the ocean due to shrinkage upon curing, or rough handling or
both. In this case, the inner waste form should not be relied upon to
resist corroding for more than a few days if made of structural steel,
or a few months if made of stainless steel. These times are short
compared to even a 50-year containment period.
While marine coatings are capable of adding to the lifetime of the
average container whose concrete does not crack, coatings can do nothing
about the cracking problem itself. Coatings are thus incapable of
improving the reliability of the present containers. The containers are
likely to have the same rate of early failures with or without coatings.
In order to significantly increase both the reliability of the
containers and the public's confidence in the waste disposal system, it
will be necessary to change either the materials of construction or the
configuration of the container or both.
-------
35
5.0 ALTERNATIVES TO PRESENT MATERIALS OF CONTAINMENT
Before any scheme for nuclear waste disposal can be considered to
be a viable alternative, it must be capable of winning public confidence.
In the case of low-level waste disposal on the deep-sea floor, that
confidence must be based upon the public's perceived reliability as
well as the actual reliability of the waste package. The problem of
nuclear waste is such an emotional issue that the public's confidence,
based upon the perceived reliability, is likely to be low unless the
actual reliability is very nearly 100%. If ocean disposal of low-level
waste is to be resumed, convincing evidence must be presented to show
that the probability of a total package failure, allowing the seepage
of contaminated water into the environment, is negligible over the
necessary containment period, even for a large number of packages.
In order to achieve such confidence, a multi-layered package having
the following characteristics Is recommended: (1) the waste should be
encapsulated in an inner core made of an impervious material that will
resist corrosion for the duration of the required containment period
even if inadvertently exposed to seawater during or shortly after
disposal. (2) The inner core should be surrounded with at least
15 cm of an incompressible nonmetallic material that will neither
crack nor absorb seawater but that will provide a uniform environment
for the core if seawater does get in. (3) The package should be one
solid incompressible mass to eliminate the possibility of implosion
under pressure. (4) The outer shell should be of a noncorrodirig
material capable of withstanding shock loading and rough treatment
at sea.
-------
36
For the 50-year containment period, it should be possible to build
such a waste package. It is not the purpose of this paper to present a
detailed design, but some recommendations based upon the characteristics
outlined above can be made. A schematic cross section of a conceptual
package design is snown in Figure 5.
5.1 INNER CORE
Consider first what would happen if the inner core were made of
plain carbon steel or one of the 300 series stainless steels. In either
case, corrosion could begin as soon as chloride ions from the aeawater
caused the pH of the pore water in the concrete to fall below about 4.
For structural steel, this will be uniform corrosion. Sucn a core might
last a relatively long time compared to stainless steels. For example,
if a wall thickness of 6.5 mm of a particular structural steel is needed
to resist deep ocean pressures, then at normal deep ocean corrosion
rates, a total wall thickness of 13 mm should last about 125 years
before the inner core implodes or begins to leak. A total wall thick-
ness of 2.5 cm might be expected to last the desired 300 years if the
corrosion remained uniform. If there were any non-uniformities to the
distribution of the corrosion over the surface however, the penetration
rate would be faster than that for uniform corrosion. The increase in
the penetration rate would be proportional to the degree of non-uniformity
and could vary from 1.5 to 100 times the uniform corrosion penetration
rate.
If the core material were a 300 series stainless steel, either
pitting or crevice corrosion or both could take place, although crevice
corrosion is more likely to be the dominant mode in this case because
-------
Low Level Waste
Imbedded in Cement
Polymer
Impregnated
Concrete
37
Unalloyed Titanium
Inner core
Outer Shell
(see below)
Concrete
Solid Polyethylene jacket
Polyethylene foam
Outer Polyethylene skin
Figure 5. Schematic cross section of
proposed low-level waste package.
-------
38
of the restricted geometry. For either the crevice corrosion or pitting
modes, it is difficult to predict how fast the stainless steel core
would be penetrated, as the corrosion rate depends on a number of
unknown factors. One of the more important of these is the amount of
stainless steel surface area that is exposed to ambient seawater. This
area will become the cathode, and the larger it is, the more rapidly
will the rate-controlling oxygen-reduction reaction be able to proceed.
The exposed area could range from less than 1 square millimeter in the
case of a small crack up to several square centimeters if a large chunk
of concrete spalls off. Assuming the worst case conditions, where a
substantial area of 0.5 cm thick stainless steel is exposed, the data
presented earlier in this report suggest that penetration of the con-
tainer could take place in as little as 5 years by a pit on the cylin-
drical surface or 35 years by crevice corrosion. These numbers could be
adjusted up or down for different wall thicknesses, but again, they are
shorter than the desired lifetime of 50 to 300 years.
In addition, there is another mode of attack which could be even
more rapid. This would involve corrosion at the point where a stainless
steel core was sealed after emplacement of the waste. If sealing were
accomplished by bolting the end cap to a flange containing a rubber CD-
ring or a gasket, then crevice corrosion may take place preferentially
in the groove holding the 0-ring or gasket in place as has often been
the case with pressure casings for oceanographic instrumentation. Once
corrosion has tunneled under the 0-ring, the container will leak,
resulting in failure by the present definition. Such failure in stain-
less steel instrument casings similar in size to a hypothetical waste
container core have sometimes occurred within 6 to 18 months of seawater
-------
39
service. In oceanography, these casings are used only for relatively
short-term immersions (1 hour to 1 month) and are given periodic pre-
ventive maintenance. Such maintenance is not presently available for a
low-level waste container once it has been placed on the sea floor. One
can conclude then that plain carbon structural steel and the conventional
300 series stainless steels are not suitable for use as inner core
materials. There are, however, other metals with increased corrosion
resistance.
The metals and alloys most resistant to both uniform and localized
corrosion in seawater include gold, platinum, tantalum, unalloyed
titanium and some higher strength titanium alloys, Hastelloy C-276,
Inconel 625, and several recently developed stainless steels. The
corrosion behavior of all of these has been studied extensively in
2 4-6 37-39
seawater. ' ' Gold and platinum can be eliminated because of
their low strength and prohibitively high cost. Tantalum, while offer-
ing excellent corrosion resistance, is also ill-suited for use as a
structural material due to its high cost, low availability and low
strength. It is used mostly as a liner in containers for severely
corrosive chemicals. The titanium alloys, Hastelloy C-276, Inconel 625
and the new stainless steels, however, are all possible candidates for
use as inner core materials.
Unalloyed titanium and two high strength titanium alloys are
commonly used in marine environments. The two alloys are Titanium 6 Al,
4V (6-4) and Titanium 6 Al, 2 Cb, 1 Ta, 0.8 Mo (6-2-1). The 6-4 alloy
has an excellent strength-to-weight ratio but would have to be down-
graded for nuclear waste applications because of a slight possibility of
stress corrosion cracking in seawater environments. The 6-2-1 alloy
-------
40
has good strength and was formulated as a replacement for alloy 6-4 to
avoid the cracking problem. Lt is a much newer alloy, however, and
there is less long-term seawater experience with it. For these reasons,
unalloyed titanium is probably the preferred choice from the titanium
alloys group. Its moderate strength of 50,000 to 90,000 psi should be
adequate for the present application, and it has withstood more than 30
years of service in seawater environments at 0°C to 30°C without any
reported instances of corrosion of any type. The major disadvantage of
all the titanium alloys is their relatively high cost as compared to
structural steel.
Hastelloy C-276 and Inconel 625 are both nickel-based chromium-
molybdenum alloys, which have long-term seawater corrosion resistances
and costs roughly comparable to those of the titanium alloys.2'37'39
Use of these materials, or one of the titanium alloys, would maximize our
confidence in the long-term integrity of the inner core. The titanium
alloys have the added advantage of being the lowest in density of any of
the candidate inner core materials. Their low density would help to
minimize the overall package weight, a matter of considerable importance
in shipboard handling.
A series of stainless alloys has recently been developed having
greatly improved resistance to crevice corrosion and pitting in seawater38
as compared to that of the conventional 300 series stainless steels.
These are the ferritic and duplex (part ferritic and part austenitic)
stainless steels. The most resistant of these are the ferritic stain-
less steels with 27 to 30% Chromium, 3 to 4% Molybdenum and less than 1%
Nickel. These alloys have crevice corrosion resistances under labora-
tory conditions comparable to that of Hastelloy C-276, although there is
-------
41
o o
far less service experience with the newer alloys. Other ferritic
alloys with 25 to 28% Chromium, 2 to 4% Molybdenum and up to 4% Nickel;
plus some of the duplex alloys with 20 to 26% Cr, 5 to 25% Ni and 3 to
7% Mo may also prove to have adequate crevice corrosion resistance.
These newer ferritic and duplex alloys have the advantage of being less
costly than Hastelloy C-276, Inconel 625 and the titaniums. Their
disadvantages are that there is yet far less service experience with
them and they are not yet as widely available as the more established
alloys. Both these latter two factors are expected to change dramati-
cally, however, within the next 5 years.
There are, thus, a number of alloys now available with greatly
improved corrosion resistance compared to the conventional 300 series
stainless steels. Any one of the alloys discussed above could probably
be used successfully as an inner core material. The final choice of the
best alloy will depend on many factors outside the scope of this report,
and must await a more detailed package design.
In order to eliminate the possibility of implosion of the inner
core under pressure, no matter what its material of construction, it is
recommended that once the waste has been placed in the core, cement be
injected under pressure to fill any voids and create a solid mass. If
the waste were liquid, it could perhaps be mixed with cement, injected
into the core and then allowed to set.
5.2 CONCRETE LAYER
The concrete used outside the core should be of high quality, with
a low permeability and a granite aggregate as described previously.
Water absorption of the concrete can probably be eliminated by using one
of the polymer-impregnated concretes that are under development at the
-------
42
U.S. Naval Civil Engineering Laboratory, Port Hueneme, California,40 and
the Brookhaven National Laboratory, Upton, New York.41
5.3 OUTER SHELL
The outer shell should serve two basic functions and perhaps a
third. It should serve first as a mold for the concrete, and second as
a shock absorber to shield the concrete from the effects of rough
handling. It is recommended that the shell be constructed of a rigid
polyethylene foam sandwich material as shown by the inset in Figure 5.
The outer skin and foam layers will provide shock resistance. These
layers will be collapsed onto the inner solid polyethylene jacket by the
great hydrostatic pressure of the deep ocean, but this will be of no
consequence as they will already have served their function. The
polyethylene should be stabilized against ultraviolet radiation using
carbon black to prevent deterioration by sunlight between the time of
manufacture and of emplacement. Marine exposures of polyethylene for up
to seven years usually have resulted in no detectable degradation,10 and
polyethylene has one of the lowest rates of water absorption of all the
polymers that have been tested. In addition, polyethylene is relatively
immune to marine borer attack when not immediately adjacent to a piece
of wood. This material is far less likely to degrade over a 50-year
period than is a conventional structural steel shell. A third possible
function would be for the inner polyethylene jacket to serve as an
additional moisture barrier if some way could be found to seal the outer
shell after insertion of the core and concrete.
Such a container could be either cylindrical or spherical in shape.
A sphere would be preferable because, under hydrostatic compression,
tensile and shear stresses would be minimized. The cylinder, however,
-------
43
might be more economical and convenient, since tubular mill forms of
most materials are readily available. In addition, the cylindrical
shape would be easier to handle at sea. There could be a high degree of
confidence in such a container of either shape for 50 to 75 years
because there is experience with: (1) concrete in seawater for that
length of time; (2) titanium, Hastelloy C-276 and Inconel 625 in sea-
water for about half that time; and (3) the newer candidate materials
for at least a few years. It is not possible, however, given our
present knowledge of the mechanisms and rates of deterioration of
structural materials in the deep ocean, to guarantee the integrity of
even this improved container for 300 years. Selected individual con-
tainers might well last that long or longer, but there will be uncer-
tainty involved. Therefore, when the desired period of isolation of
even low-level nuclear waste in the ocean exceeds the 50- to 75-year
period over which actual material exposure tests exist, it is recom-
mended that the waste be disposed of in such a manner that the disposal
environment itself provides another barrier to migration of the waste,
rather than relying strictly on the packaging materials themselves to do
that job.
While a full discussion of additional options to the present
methods of nuclear waste disposal is beyond the scope of this report,
one would be remiss in making this recommendation without indicating
briefly that there is at least one possible additional method already
under consideration. Disposal of low-level radioactive waste in stable
42
geologic formations of the deep seabed would seem to be an option
compatible with the concept of not requiring or expecting the physical
container to permanently isolate the waste. Such depositories are being
-------
44
discussed and evaluated mainly for high-level waste disposal. However,
their most attractive feature - the potential of being able to chemi-
cally bind the waste and thus to isolate it from man and his marine food
chains for long periods of time after the deterioration of the original
container - should apply to low- as well as to high-level waste.
At the present time then, the most logical course of action might
be to: (1) attempt to develop cost-efficient nuclear waste packaging
systems which could demonstrably isolate and contain the wastes for
specified periods of time, preferably fifty to three hundred or more
years; and (2) couple this packaging system to a carefully selected
environmental isolation system (deep-sea disposal site) to provide for
an additional measure of isolation and containment of the waste after
the man-made packaging system has begun to release its contents.
-------
45
6.0 REFERENCES
1. Dyer, R. S., 1976, "Environmental Surveys of Two Deepsea Radio-
active Waste Disposal Sites Using Submersibles", International
Symposium on the Management of Radioactive Wastes from the
Nuclear Fuel Cycle, IAEA, Vienna, IAEA-SM-207/65, Vol. II, p.
317-338.
2. Tuthill, A. H. and C. M. Schillmoller, 1971, "Guidelines for
Selection of Marine Materials", 2nd Edition, May, 1971, The
International Nickel Co., Inc.
3. Fontana, M. G., and N. D. Greene, 1967, Corrosion Engineering,
McGraw-Hill.
4. Uhlig, H. H., 1971, Corrosion and Corrosion Control, Second Edi-
tion, Wiley-Interscience.
5. Fink, F. W., and W. K. Boyd, 1978, "Corrosion of Metals in Marine
Environments", Metals and Ceramics Information Center, Battelle
Columbus Laboratories, Columbus, Ohio, MCIC Report No. 78-37,
March, 1978. (See also the earlier DMIC Report No. 245, May,
1970, AD 712585.)
6. Reinhart, F. M., 1976, "Corrosion of Metals and Alloys in the Deep
Ocean", U.S. Naval Civil Engineering Laboratory, Technical
Report R-834, February, 1976. (See also LaQue, F. W., 1975,
Marine Corrosion, Wiley-Interscience.)
7. Dexter, S. C., and C. Culberson, 1980, "Global Variability of
Natural Sea Water", Materials Performance, 19 No. 9, pp. 16-
28.
8. Knauss, J. A., 1978, Introduction to Physical Oceanography, Prentice
Hall, Inc., pp. 166-193.
-------
46
9- Dexter, S. C., 1974, "Microbi dogica! Fouling and Its Control in
Coastal Water and thu Deep Ocean", Woods Hole Oceanographic
Institution Technical Report WHOI-74-64, September 1974.
Unpublished manuscript.
10. Muraoka, J. S., 1964_67, "Deep_0cean Biodeterioration of Materials -
Parts 1 to VI", U.S. Naval Civil Engineering Laboratory, Port
Hueneme, CA, Technical Reports: R329 (November, 1964) AD608939;
R393 (August, 1965) AD619014; R428 (February, 1966) AD631078;
R456 (June, 1966) AD636412; R495 (November, 1966) AD'642838;
R525 (May, 1967) AD65U24.
H. Dehghanian, C., and C. E. Locke, 1982, "Electrochemical Behavior of
Steel in Concrete as a Result of Chloride Diffusion into
Concrete: Part 2", Corrosion, Vol^J8_No^, p. 494. (See
also Part 1 of the same study in Ref. No. 3 of the quoted
article, and Paper No. 52 by Hartt and Voshardt presented at
CORROSION/81, Toronto, Canada, April, 1981.)
12. Mehta, P. K., and H. H. Haynes, 1975, "Durability of Concrete in
Seawater", Journal of the Structural Divison, ASCE, 101, No.
ST8, Proc. Paper 11516, August, 1975, p. 1679-86.
13. Haynes, H. H., and R. S. Highberg, 1976, "Concrete Properties at
Ocean Depths", Presented at ASCE National Water Resource and
Ocean Engineering Convention, San Diego, April, 1976.
14. Reinhart, F. M., 1966, "Corrosion of Materials in Hydrospace", U.S.
Naval Civil Engineering Laboratory, Technical Report R-504.
15. Reinhart, F. M., 1967, "Corrosion of Materials in Hydrospace, Part
I - Irons, Steels, Cast Irons and Steel Products", U.S. Naval
Civil Engineering Laboratory, Technical Note N-900, July,
1967.
-------
47
16. Southwell, C. R. , and J. D. Bultman, 1975, "Corrosion of Metals in
Tropical Environments", Part 10, Final Report of 16-Year
Exposures, Naval Research Laboratory, NRL Report 7834,
January, 1975.
17. Southwell, C. R., and A. L,. Alexander, 1968, "Corrosion of Structural
Ferrous Metal in Tropical Environments - 16-Year's Exposures
to Sea and Fresh Waters", presented at 1968 NACE Conference,
Cleveland, Ohio.
18. King, C. A. N., 1962, An Introduction to Oceanography, p. 89-90.
19. Park, K., 1966, "Deep-Sea pH", Science 154, p. 1540-1542.
20. Miller, J. D. A., 1970, Microbial Aspects of Metallurgy, American
Elsevier, p. 63, New York.
21. Uhlig, H. H. , 1971, Cojr_ro_sio_n and Corrosion Control, Second-Edition,
Wiley-Interscience, p. 96.
22. Dexter, S. C. , 1979, On. ISoajrd Corrosion Analysis o_f a. Recovered
Nuclear Wa_st_e Container, U.S. Environmental Protection Agency,
Office of Radiation Programs, Report No. ORP/TAD-79-2, Washington,
DC 20460.
23. Lennox, T. J., Jr., M. H. Peterson, and R. E. Groover, 1968,
"Marine Corrosion Studies - The Corrosion Characteristics and
Response to Cathodic Protection of Several Stainless Steel
Alloys in Quiescent Seawater", Naval Research Laboratory
Memorandum Report No. 1948, AD 684073, November, 1968.
24. Lennox, T. J., Jr., and M. H. Peterson, 19.76, "Inherent Corrosion
Resistance and Response to Cathodic Protection in Seawater of
Recently Developed Stainless Steel Alloys", Naval Research
Laboratory Report No. 8016, August, 1976.
-------
48
25. Dyer, R. s. , Personal Communication, April, 1980. From: Special
Subcommittee Hearing on Industrial Radioactive Waste Disposal
of the Joint Committee on Atomic Energy, 1959, p. 3092.
26. The Metals Handbook, 1961, Eighth Edition, Vol. j
for Metals, Metals Park, Ohio, p. 409.
27. Reinhart, F. M., 1971, "Corrosion of Materials in Hydrospace, Part
VI - Stainless Steels", U.S. Naval Civil Engineering Laboratory,
Technical Note N-1172, September, 1971.
28. Gjorv, 0. E., 1972, "Long-Time Durability of Concrete in Seawater"
Proceedings - American Concrete Institute, Vol. 68, p. 60-67.'
29. Haynes, H. H., R. s. Highberg, and B. A. Nordy, 1976, "Seawater
Absorption and Compressive Strength of Concrete at Ocean
Depths", U.S. Naval Civil Engineering Laboratory, Port Hueneme,
California, Technical Note N-1436, April, 1976.
30. Haynes, H. H., and R. s. Highberg, 1978, "Deep Ocean Study of
Concrete Spheres", presented at the FIP Eighth World Congress,
April-May, 1978, London.
31- Jenks, G. H., 1957, "Effect of Radiation on Corrosion", presented
at HRP Civilian Power Reactor Conference, Oak Ridge, May,
1957, U.S. Atomic Energy Commission Report TID-7540.
32. Byalobzheskii, A. V., 1970, Radiation Corrosion, Translated from
the Russian by Israel Program for Scientific Translations,
Jerusalem.
33. Uhlig, H. H., 1971, Corrosion and Corrosion Control, Wiley-mterscience,
p. 145.
34. Haynes, H. H., Personal Communication, November, 1977. Estimate
based on experience at the U.S. Naval Civil Engineering
Laboratory, Port Hueneme, California.
-------
49
35. Haynes, H. H., Personal Communication, November, 1977.
36. Hunger, C. G., 1975, "Marine Coatings", in F. L. LaQue, Harine
Corrosion, Wiley-Interscience, p. 283-317.
37. Reinhart, F. M., 1967, "Corrosion of Materials in Hydrospace, Part
II - Nickel and Nickel Alloys", U.S. Naval Civil Engineering
Laboratory, Technical Note N-915, August, 1967.
38. Streicher, M. A., 1983, "Analysis of Crevice Corrosion Data from
Two Sea Water Exposure. Tests on Stainless Alloys", Materials
Performance, Vol. 22, No. j, p. 37.
39. Dexter, S. C., 1979, Handbook of Oceanographic Engineering Materials,
Wiley Interscience, New York.
40. Haynes, H. H., and Eckroth, W. V., 1979, "Lightweight Concrete
using Polymer-Filled Aggregate for Ocean Applications - An
Exploratory Investigation", U.S. Naval Civil Engineering
Laboratory, Port Hueneme, California, Technical Note: TN No.
N-1565, December, 1979.
41. Kukacka, L. E., 1977, "Production Methods and Applications for
Concrete Polymer Materials", Presented at AIChE Symposium on
Degradation of Concrete and Ceramics, New York, NY, November,
1977.
42. Hollister, C. D. , 1977, "The Seabed Option", in Oceanus, VpJl_._J20,
No_, _1, "High Level Nuclear Wastes in the Seabed?", Woods Hole
Oceanographic Institution, Woods Hole, Massachusetts.
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 520/1-82-005
2.
4. TITLE AND SUBTITLE
Materials for Containment of Low-Level Nuclear Waste
in the Deep Ocean
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
December 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Stephen C. Dexter
8. PERFORMING ORGANIZATION REPORT NO.
CMS-C-1-78
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
College of Marine Studies
University of Delaware
Lewes, Delaware 19958
11. CONTRACT/GRANT NO.
Contract No. WA-6-99-2767-J
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Radiation Programs
U.S. Environmental Protection Agency
401 M Street, SW
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
ANR-461
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The United States ocean dumping regulations developed pursuant to PL 92-532,
the Marine Protection, Research, and Sanctuaries Act of 1972, as amended, provide
for a general policy of isolation and containment of low-level radioactive waste
after disposal into the ocean.
In order to determine whether any particular waste packaging system will meet
this general requirement, and for how long, it is necessary to know what materials
were used and how those materials will behave in the deep sea.
This report discusses the mechanisms of marine corrosion of structural and
stainless steels, the degradation mechanisms acting on concrete in a marine
environment, the interaction between metal and concrete combined as a low-level
radioactive waste package under deep sea conditions, and the effect of
environmental parameters such as dissolved oxygen, temperature, and water velocity
on deep sea degradation processes. A concluding discussion presents various
improved metal alloys and concrete additives which may result in greater resistance
of a low-level radioactive waste package to deep sea deterioration processes.
17.
DESCRIPTORS
Ocean Dumping/Ocean Disposal
Low-Level Radioactive Waste Packaging
Deepsea Corrosion
Deepsea Deterioration of Concrete
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Unlimited Release
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
58
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
------- |