ML19351E566
| ML19351E566 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 12/02/1980 |
| From: | Fredericks K, Giacobbe F METROPOLITAN EDISON CO. |
| To: | |
| Shared Package | |
| ML19351E556 | List: |
| References | |
| NUDOCS 8012100458 | |
| Download: ML19351E566 (24) | |
Text
.
\\
f
()
EVALUATION OF THE LINER INTEGRITY OF THE TMI UNIT 2 EPICOR II RADWASTE SYST_.S F. S. Giacobbe K. H. Fredericks Rev. 1, December 2, 1980 80123ooq57
+
l l
9 The contents of this study are based upon available data at the time of issuance.- Data which may have a direct bearing upon the conclusions is expected in the immediate future. This data will include, but not be limited to, the Epicor Proprietary Disclosures data from liquid samples to be taken from selected prefilters and the results of various resin de-gradation. studies which are currently in progress.
The right to alter any portion is-reserved.
a f
r
Introduction integrity of the Epicor II Spent Resin Liners as af fected by 1hc this time was various internal environments which are believed to exist at Ef fluent data was provided by TMI for the sixty-five lihers which ev alua ted.
contained spent resin as of August, 1980.
This data was reviewed by' the GPUN Systen. Laboratory 02emistry Section in order to -provide a basis for detennining When the nature of the environment which would be present during storage.
additional liners are proc ;ssed or when additional data regarding the internal is developed, this data vill be evaluated and a revised report environment available Analysis of the data currently / ed to a classification of the internal l
is sued.
env ironment into four categories. These categories grouped the various environ-ments according to the corrosiveness to the ASTM: A36 carbon steel liner base cate rial. A complete analysis of the environments are contained in Attach-ment I.
The most aggressive environment (category 4) postulated is a dilute, air saturated, hydrochloric and boric acid solution with a pH of approximately 2.5 and very low solids / salt concentration. Based on the dewatering technique utilized (I), it has been demonstrated that a maximum of 1.5 gallons of this solution can be expected to be on the floor of the liners.
Although ' corrosion resistant coatings had been applied to the liner internals no nondestructive examinations were performed on the coatings af ter application.
Subsequent examination of liner coatings onsite with a holiday detector found them to contain numerous defects, the analysis of the liner integrity, therefore, was carried out assuming that the 1.5 gallons of solution
~
were in direct contact with a pinhole in the coating and that the carbon steel This case substrate is currently being acted upon by the internal environment.
would produce the earliest liner penetration as opposed to a liner with no coating, with a large patch of coating removed or a defect free coating.
(1) TMI 2 Spent Resin Liner Dewatering Study, April 1980.
=
3 Introduction (continued)
Tests conducted onsite have also shown that the liners cannot be from overpressurizing the
. pressurized above 2 psig, thus eliminating concerns liners due to evolution of gases from the resins.
Another case which needs to be addressed but will not be discussed contact of the resin beads with the coating and/or in this report is direct This will be evaluated 4
the carbon steel when there is no free-standing water.
and will be reported on at a later date.
Conclusions
-1.
Final effluent chemistry data for sixty-five liners have been corrosive-reviewed and categorized in four groups. representative of their inherent ness to the liners. These categories are:
i (1)
Solutions with pH > 5 and cenductivities < 50 umhos.
i (2) Solutions with pH > 5 and conductivities of 50-4000 umhos.
(3)
Solutiona with pH 3.5-5.0 and conductivities < 70 unhos.
2 (4)
Solutions with pH 2.5-3.5 and conductivities of 100-700 umhos.
2.
Category (4) is the most aggressive solution. Four liners fall into this category as follows:
Liner Designation Size Metal Thickness PF-16 4' x 4' 1/2" - 5/8" PF-17 4' x 4' 1/2" - 5/8" PF-18 4' x 4' 1/2" - 5/8" PF-19 4' x 4' 1/2" - 5/8" Under the current worst case conditions postulated, small pinhole the earliest in 15 months for a penetration of liners could be predicted'at 1/2" liner and 19 months for a 5/8" liner. This, however, is conservative considering the steady-state corrosion rate assumed is the initial high rate which would be expected *o decrease to some lower value with time.
Weie rh a
av.
y
+
.; Conclusion (continued)
Category (3), although the solution is slightly higher in pH than 3
category (4), it is still sufficiently low that a slight drop in pH,will probably occur within pits producing corrosion rates similar to category (4).. Liners in category (3) are as follows:
Liner Designation Size Metal Thickness
. PF-40 4' x 4' 1/2" - 5 /8 "
. PF-41 4' x 4' 1/2" - 5 /8"
. PF-42 4' x 4' 1/2" - 5/8"
. PF-43 4' x 4' 1/2" - 5 /8" These liners under worst case conditions could also be perforated in the 15-19 month period.
For purpose of this analysis " worst case" has been defined as 4.
.01% hydrochloric acid solution at 750 - 900 F in direct contact with the a
carbon steel substrate via a pinhole defect in the coating. The solution is air saturated, however, no intrusion of air into the container is predicted, is consumed in therefore, with tt=e the oxygen concentration will decrease as it The corrosive attack is expected to proceed in a pitting the corrosion process.
mode.
5.
Categories (1) and (2), which encompass fif ty-seven liners, have pH values in excess of 5.
These liners are expected to have a life in excess of 25 years not counting any contribution to life by the coating and assuming tha t resin degradation will not lower the pH.
Attachment II desc ribes the corr,osion of these liners in more detail.
6..
For those liners where solutions are in contact only with the coating and no defects are assumed, life expectancy for the coating would he conservatively on the order of ten years. This being the typical life of a coating in contact with desineralized water with no coating maintenance.
Conclusion (continued) 7.
Long term resin degradation is expected to release borate anicus plus amines which az e pr edic ted to establish a buf fering action and possibly raise the pH of the solution.' In this event the environment will become less aggressive to the carbon steel substrate (see Attachment I).
Discussion The corrosion of the carbon steel substrate for the types of environ-ments defined will be most significantly af fected by the hyd rogen ion concentra-tion (pH) and oxygen concentration of the solution. The oxygen concentration for the internal environments defined, however, will be the overriding f actor influencing both short term and long term corrosion rates.
As the conc en tra tion of oxygen within the sealed liner is depleted the corrosion rate will be decreased accord ing ly. This factor will be most significant for liners in categories (1) and (2) where the corrosion rate-is suf ficiently slow that aggressive pitting is not an tic ipa ct.1 but rather localized corrosion may occur initially with a gradual spreading out of the corrosion with thne.
For category (4) liners, maximum corrosion rates for dilute aerated hydrochloric acid solutions would be on the order of 400 mils per year (mpy) as discussed in ' Corrosion and Corrosion Control', 2nd Ed.
Knowing that the i'
thickness at the bottom of the 4 X 4 liners is 500 mils (nin), pinhole perfora-i
. tion could occur in as soon as 15 months assuming no competition for oxygen from J
other locations within the liner which could decrease the available oxygen thus i
slowing the corrosion rate and that the environment is able to maintain the initial high corrosion rate. Maintaining this high corrosion rate for lonc periods is unlikely but until the steady state rate can be adequately defined it is the most conse.rvative approach.
In regard to category (3) liners, although the environment is somewhat
~
less aggressive if one considers pitting as the primary mode of corrosion then l
a.
??k
,a
Discussion-(continued) at pH 3.5, it would be necessary to assign a corrosion rate similar to that for category (4).
Liners in categories (1) and (2) with their assumed environments
'(pH > 5) would be expected to experience corrosion rates on the order of 10 mpy.
In this pH ranga, however, the formation cf ferrous hydroxide is expected and.
l this will both slow th diffusion of oxygen and act to produce an alkaline envir-s onment at the corroding surface. These factors will minimize both the pitting rate and general corrosion.
The potential for stress corrosion cracking to occur as well as hydro-gen embrittlement were also considered. These corrosion mechanisms, however, were dismissed at this time as the current environmental, material and stress parameters did not indicate those mechanisms would occur. As additional data is available to further define the internal conditions, these mechanisms will be reconsidered as necessary.
In discussions with the coating manufacturers, it was learned that neither coating utilized is recommended for low pH environments; however, this recommendation, in general, is based on concentrated acid solutions and not dilute solutions. The fact -is that few lataratory tests were conducted by the coating manufacturers in dilute acfds and no test data exists for a 1% or less hcl solution. Tests in 1.5% oxalic acid at room temperature however, showed no degradation in eight months on the Plasite 7155. Phenoline 368 tested in 1*
citric acid at 1300F and 1500F showed both acceptable and unacceptable results in different tests.
It is the opinion of the coating vendors, however, that the dilute acid at ambient temperature will not be significantly more agressive than demineralized water in terms of its effect on-the coating but rather its effect
~
on the carbon steel substrate due to exposure through coating defects or long f
term permeation will be critical. Initial results from on-going GPU Laboratory r
Discussion (continued) tests on the effects of the proposed worst case environment on the liner coatings has shown no coating degradation in two months.
It is our assessment that the coatings will be unaf fected in the near term by the aqueous environment and can be expected to have a lifetime similar to that of coatings exposed ho deminera-lized water (approx. 10 years). This assessment, hov2ver, may be altered based on the results of tests of resin beads in direct contact with tha coating in a non-aqueous environment. Although initially the coating most likely will contribute to the overall liner integrity, in the long term its ef fectiveness in preventing corrosion will be nil.
Recom=endations Liners in categories (3) and (4) will require near term remedies to 1.
Discussion should be commenced to identify preclude leakage to the environment.
Potential solutions a satisf actory method for restoring long term integrity.
would include methods. to modify the internal environment chemically by raising pH or scavenging oxygen.
Ef forts should be made to sample the liquid remaining in liners in 2.
categories (3) sac (4) initially followed by categories (1) and (2) to confirm This data would afford a or disprove the current hypothetical environments.
more definitive analysis. If possible, gas samples should also be taken to indicate determine if hydrogen is being produced or other gases which might degradation of the resin beads.
3.
Contingency plans need to be developed in the event leakage can result f rom corrosion to the liner. This might be best accomplished by desig-ning a corrosion resistant container into which the current liners can be placed.
4.
Af ter the effluent sample has been analyzed as recommended in item 2, Electrochemical corrosion tests should be conducted to determine the actual corrosion rates which can be expected from the actual liner environment.
e l
\\
l -
Recommendations (continued) 5.
Once the resins have been identified which exist inside the liners, laboratory tests should be conducted exposing carbon steel coupons-to resins of the same type and condition under the influence of a radiation sou'rce producing I'
a dose rate equivalent to that inside the liners. The coupons should include stressed samples as well as welded samples.
o
e 1
ATTACIDiENT I Evaluation of Internal Envircament Fpient Liners for Storage of Spent Resins K. H. Tr edet ick Rev. 1, December 2, 1980
~
Two basic sets-of conditions must be considered in order to make a complete evaluation of environmental conditions within the resi6 liners which are employed for treatment of radioactive wastes at Three Mile Island. The first consideration is the chemistry of the free-standing liquidsand gasses which remain in the cask when it is placed in storage.
The second consideration is the changes which materials within the liner will undergo.
These changes will consist largely of resin decomposition from normal aging or from radiation effects. Both of these processes represent long-range concerns.
Chemistry data from more than sixty liners has been evaluated and the results have been employed to characterite the conditions inside stored liners. Experimental data indicates that following routine de-watering, a ma:cimum of 1.5 gallons of liquid remain in the bottom of a liner. For purposes of evaluation, it is conservatively assumed that this liquid will have the same chemical composition as the last process t
~1iquid which was measured at the effluent of the liner during service.
Tables One, Tua and Three contain tabulations of the final effluent data for the Epicor II liners used in the evaluations.
Initial examin-ation of the data suggests four general categories which are as follows-i
-1.
Liners in this category are characterized by low conductivity
]
effluent'and pH values in the range of 5 or greater. Conduct-l ivity values ara generally less than 50 umho and range down i
~
~
. to less than 1 umho.
This range of pH and conductivity values suggesus a solution of boric acid in deminerali=ed water.with minute quantities of neutral sales. Conductivity values toward the higher side of this range suggest the presence of slightly greater quantities of dissolved salts.
2.
These liners contain primarily mixtures of salts.
The pH value ranges from approx. S to 8 with the majority in the range of 7.
Conductivities range from 50 to 4000 umho with the majority falling between 200 and 2000 umho.
3.
Liners in this category contain very dilute mixtures of neutral salts and strong acids. pH values range from 3.5 to 5.0 and conductivities are less than 70 umho.
i 4.
This category also contains liners with mixtures of neutral salts and strong acids but is somewhat less dilute than those described in Category 3.
pH values range from 2.5 - 3.5 and e
conductivities from 100 - 700 umho.
For purposes of evaluating actual concentrations of the various constituents of solutions, the predominant salt was assumed to be sodium chloride and the strong acid hydrochloric acid. Where the pH values were lower than could be attributed to boric acid concentrations corresponding to measured boron concentrations, hydrochloric acid was assumed to be present and a concentration was calculated from the pH values. A conduc-
4-4
-3 1
tivity value was assigned to the calculated acid concentration and the f
' difference between this calculated cc.ductivity and the measured.conduc-civity was assumed to result from contribution by sodium chloride. When pH values were neutral, sodium chloride was assumed to be the sole contrib-utor to conductivity. Equivalent conductances are from " CRC Handbook of Chemistry and Physics, 50th ed." and values from the 0.02 gram equivalents /
i 1000 cubic cm column were used. Following is a listing of the conditions which were assumed in making the analysis of each category:
Category J
i Parameter 1
2 3
4 1
pH 5.0 7.0 3.5 2.5 conductivity, umho 50 2000 70 700 i
Boron, as ppm 3 750 1500 950 1400 Sodium, ppm 1.0 200 1.0 2.0 Accordingly, the conditions in liners from each category should approx-imate the following:
Category Parameter 1
2 3
4 Boric Acid, w!.
0.45 0.85 0.55 0.8 Hydrochloric Acid, w7.
NP(1)
NP(1) 0.001 0.01
}
0.002 0.1 NP NP (1) Not present k
e v
s
. Since there is evidence of strong acid presence in categories 3 and 4, a calculation was made to determine the mass of iron which could be dis-placed'by the 1.5 gal. max. of the solution assumed to be present (from the GPU dewatering report) at the bottom of the liner.
This quantity of iron would amount to approx. 0.1 g for liners in Category 3 and approx.1 g for liners in Category 4, af ter which the acid would be spent.
A second consideration is the effect of oxygen remaining in the liner after it is sealed. Oxygen contained in free space at the top of the liner, that present in void spaces between dewatered resin beads and dissolved oxygen present in the water of hydration were considered. A free space of 18 inches at the top of each liner and a void volume of 40% were assumed.
A dissolved oxygen content of 8 ppm was assumed for the water of hydration and water of hydration volumes of 125 and 450 gal, were assumed for the 4 x 4 and 6 x 6 liners.
These assumptions would yield approximately 270 g (0.6 lb) of totsi oxygen in a 4 x 4 liner af ter closure and 800 g (1.86 lb) in a 6 x o.
The oxygen present within liners would ultimatuiy be expected to react with either materials within the liner or with exposed portions of the metal liner itself.
The conservative assumption would be that all.
of the oxygen reacts with the liner. Although a series of intermediate re-actions would be expected to take place, the final products saould be iron Initially, Fe2 3 should be the predominant species. As the oxygen-O oxides.
in the liner is depleted, formation of Fe3 4 would be f avored.
In actuality, 0
25 7
Initial decrosslinkage of the cation resins would probably not result in any pH changes, since the functional groups would remain unaffected.
More severe degradation would destroy the functional groups releasing sulfur compounds and the associated cation. Since sodium is the major cationic constituerit, neutral compounds such as sodium sulfate would result with no depression of pH values.
Since it appears that radiation results in similar decomposition products to those produced by chemical agents, it is not anticipated that the environ-mental changes to the liner interiors would ha significantly different than those which would result from normal resin decomposition due to aging.
Radiation decemposition is believed to result from the effects of peroxides which are produced by irradiation of water associated with the resin and, perhaps, to a lesser extent from direct cleavage of molecule bonds by the radiation.
According to Cohen (1), the formation of peroxides occurs when reducing and oxidizing radicals. are spatially separated and exist at high concentrations.
This condition would exist only in the presence of very high beta-ga=ma exposure rates, neutron exposures or as the result of other re-actions such as n, alpha which possess very high linear energy transfer values.
At lower densities of energy absorption, the recombination reaction will be favored and peroxide levels should be minimal. This condition is expected to exist within the stored liners.
(l') Cohen, Paul, " Water Coolant Technology of Poner Reactors", Gordon and Breach Science Publisher, New York,1969.
5-some combination of the two oxides would be expected.
Assuming no reaction with anion resin which is stored in a Tiner the oxygen could be expected to react with the quantities of iron des-I cribed in the table below.
Liner 02 g(1b)
Fe,g(ib)(Fe3 3 formation) Fe,g(1b)(Fe3 4 formation) 0 0
1 4x4 270(0.6) 62S(1.38) 706(1.56) 6x6 800(1.76) 1788(3.94) 2093(4.61)
Competition from anion resin would he expected to reduce these quanti-ties significantly, especially if the epoxy coating of the liner did not have an initial defect.
Long-term resin degradation will contribute to changes in the internal environment of the liners. Normal decomposition and radiation induced de-composition are potential contributors.
It 13 assumed that both will result in similar products of decomposition.
The anior. resin appears most vulnerable with the breakdown mechanism being loss of functionality. The cation resin is significantly more resistant and is likely to suffer loss of cross linkage as its primary decomposition mechanism.
Loss of anion functionality would result in the release of the associated anion and various amines such as crimethylamine.
Since the predominant anions are borates, which are buffers, and the amines are alkaline in nature, such decomposition sould benefit the liner environment by buffering strong acids and/or raising the pH.
4
-=*v
.y
TABLE ONE
)
i PREFILTER SUMHARY Final Final-Final ID-pH Cond, umho Na, pp=
B, ppm
- Ca te go ry 1
PF-1 5.15 3.22
<1 672 PF-2 6.44 14.5 12 728 e 1
PF-3 7.33 202 19 1160 2
PF-4 8.0 942 10 822 2
PF-5 8.27 3980
<1 656 2
d PF-6 7.57 1220 150 517 2
/<
PF-7 7.09 365 3
1984 2
PF-8 7.36 235 24 1109 2
j PF-9 7.58 342.
27 1298 2
PF-10 7.92 24.6 1.8 76 1
PF-11 8.05 0.45
<1 4 10 1
PF-12 7.87 1100 104 1568 2
f PF-13 7.7 1220 180 1807 2
" 14 8.08 1900 420 1460 2
i PF-15 7.75 1300 115 1552 2
PF-16 2.79 700
<1 1392 4
Y PF-17 3.52 140
<1 1320 4
i PF-18 3.39 180 6.6 1298 4
PF-19
-3.13 300
<1 1353 4-PF-20 4.89 7.2 2
259 1
1 PF-21 6.3 84.5 4.3 801 2
PF-22 5.28 3.15
<1 498 1
i PF-23 7.56 2100 180 2770 2
PF-24 4.95 7.76.
<1 801 1
'PF-25 5.07 10.86
<1 686-1
4
. Final Final Final ID pH Cond, umho Na, ppm B. ppm Category F
.PF-26
-4.96 13.1 1.23 757 1
PF-27 4.82-6.31 (1
779 1
'PF-28 7.19 440 35
-1514
.,-2 PF-29 5.55 18 4
757 1
PF-30 5.0 6.68
<1 763 1
PF-31 5.12 4.6
<1 965 1
4 PF-32 5.19 4.12
<1
.963 1
PF-33 5.66 1.86
<1 595 1
PF-34 4.70 9.25 41 920 1
i PF-35 5.34 3.40
<1 693 1
PF-36 5.43 3.15 1.5 985 1
1 PF-37 5.08 4.6 2.1 96?
1 PF-38 5.13 4.0
<1 1039 1
PF-39 5.53 8.5
<1 909 1
PF-40 4.0 41.5
<1-930 3
PF-42 3.8 55
<1 985 3
j PF-42 4.7 9.65
<1 952 3
PF-43 3.67 69.0
<1 1017 3
PF-44 7.6 370 100-433 2
PF-45 6.89 230 88~
757 2'
PF-46 6.05 140 38 822 2
1, 6
O k
i 3
I i
i j
o TABLE TWO DEMINERALIZER #1 (DF)
SUMMARY
Final Final Final ID pH Cond, ucho Na, ppm B, ppm Category DF-1 5.27 2.25
<1 600 1
D F-2 5.87 1.74 1.9 486 1
DF-3 5.56 55.7
<1 577 2
DF-4 8.62 58 9.8 75 2
DF-5 7.11 77 6.0 901 2
DF-6 6.37 0.7
<1 22 1
DF-7 7.21 0.405
<1
< 10 1
DF-8 7.56 935 120 2066 2
DF-9 7.66 1025 110 1926 2
DF-10 5.56 4.62
<1 1428 1
DF-11
- 6. 9.s 730 75 3354 2
DF-12 5.24 3.15
<1 995 1
DF-13 5.53 3.0
<1 995 1
TABLE THREE DEMINERALIZER #2 (DS)
SUMMARY
~
Final Final Final
. ID pH Cond, umho Na, opm_
B, pom Category DS-1 5.58 1.82
<1 420 1
DS-2 5.18 2.91
<1 893 1
DS-3 6.21 0.53
<1 15.1 1
DS-4 6.92 0.56
<1 11 1
DS-5 5.28 3.95 41 1385 1
DS-6 5.34 3.10
<1 736 1
a s
I e
e e
9 6
ATTACIDINT II
s ATTACRMENT II and (2) forrosion of Carbon Steel Liners in Categories (1)
The effect of pH on the corrosion rate of carbon steel in aerated demineralized water is significant, as can be seen in Figure 1.
It,is readily that between pH 4 to 10 where the pH is normally expected.to be inside apparent the liners, that the corrosion rate is relatively uniform and on the order of 10 mils per year.
For the purpose of defining the theoretical maximum amount of iron which can be corroded by a puddle of water on the tank floor under the above conditions the following assumptions were made:
The amount of oxygen which exists inside the liner is equal to 20%
1.
of the volume of air which would be present inside the liner assuming the resin already occupies 40% of the volume and that no in-leakage of air will occur once the liner is sealed.
2.
For solutions in categories (1) and (2) the total amount of corrosion will be dependent only on the amount of oxygen available.
Stoichiometric calculations indicate approximately 4.6 pounds of tren can be converted to Fe3 4 in a 6 I 6 liner at the time oxygen is depleted. With 0
iron having a density of.28 lbs/cu. in. approximately 15 ctbic inches of metal could be dissolved. This dissolution aan occur over a variety of areas and will depend on the size of the cortbsion initiation site and the progression of the
~
Other corrosion 1aterally under t'ae coating increasing the area of attack.
f actors affecting the corrosion rate would be the gradual depletion of oxygen in the system and the build up of corrosion products sloeing the diffusion of oxygen to the corroding surface. Both - these phonomena will tend to slow the
. corrosion reaction below the estimated 10 mpy.
perforation of the liners in categories (1)
Based on this a:sessment, This is unlikely, however, as and (2) could occur in approximately 25 years.
Also as the corrosion rates will be decreasing with time as oxygen depletes.
.-.-~
- s
.- s 1
< Corrosion of _ Carbon Steel Liners in Categories (1) and (21 (continued) the corrosion spreads out to larger areas, there will not be sufficient oxygen l
to produce complete penetration. In addition, this analysis assume,s no adverse changes in the environment with time. The liner bottom, however, may have a
been weakened in the corroded area depending on how long the solution has been in contact with the steel and precautions may have to be taken to prevent accidental perforation if the liner were ever lifted. Hewever, at a 10 mil per year corrosion rate this is not a near term concern.
i I
t REFERENCES 1
1.
W. G. Whitman e' R. P. Russel. The Acid Corrosion of Metals. Industrial and Engineering Chemistry,- Vol.17, No. 4,1925.
2.
M. G. Fontana and N. D. Green. Corrosion Engineering, 2nd Ed., McGraw Hill i
3.
H. H. Uhlig. Corrosion Handbook, John Wiley -&
Sons 4.
H. H. Uhlig. Corrosion and Corrosion Control, John Wiley & Sons
)
i 5.
G. W. Whi man, R. P. Russell and V. J. Altien. Effect of Hvdrogen Ion Concentration in the Submerged Corrosion of Steel. Industrial and Engineering Chemistry, Vol. 16, No. 7, 192 4.
6.
F. N. - Speller. Corrosion. Causes and Prevention, McGraw Hill i
4
,r 4
s er
,m<'
s r-
-*-,v--
- .m....
- me.e O.O'i l
7(
8 r
h
- 0.03 i
9 i
9 9
i
~
9
_.p N
_.._ _ ~ -
_. __ w O
V' d
O.01 I
C O
' V'.
O G
L C
% se e.N c ^
.c,. :
v
\\.
\\.
0.01
/
l
/
(
/
/
J-I
/
.0 l
l
?ic. 1 Effect of pH v. correstan of iren in aerated eat: ra tt: r. :e tectarature G.~ai t::an. Ausseil. anc Altieris.
i e
~
k