ML20034A988

Forwards Q-1 Comments on Section 6.2 of Bgv Plasar, Intruder Protection Based on Barrier Code.Comments Expand 891006 Comments by D Widmayer That Did Not Address Barrier

ML20034A988
Person / Time
Issue date:	04/18/1990
From:	Shewmaker R NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:	Surmeier J NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
References
REF-WM-3 NUDOCS 9004250185
Download: ML20034A988 (21)

Text

.

['

MEMO SURMEIER BS 6

i MEMORANDUM FOR: John J. Surmeier, Chief Technical Branch APR 181990 Division of Low-Level Waste Management and.Decomissioning, HMSS THRU:

Michael Tokar, Section Leader i

Technical Branch Division of Low-Level Waste Management and Decomissioning, NMSS -

FROM:

Robert E. Shewmaker Technical Branch Division of Low-Level Waste Management and Decomissioning, NMSS

SUBJECT:

BGV PLASAR COMMENTS ON SECTION 6.2,

" INTRUDER PROTECTION" BASED ON BARRIER-CODE Enclosed are Q-1-comments on Section 6.2 of the BGV PLASAR. These'

{

reflect the combined input from the Corps of Engineers, NIST and myself on the -i 1

application of the BARRIER, computer code to the BGV concept. These expand the l

series of comments provided by D. Widmayer on Section 6.2, dated October 6,-

1989, that did not address BARRIER. Please contact me if you have any l

questions related to these comments or on this subject as it applies to our-BGV PLASAR effort.

Or161ne.1 Si~ geed W i

Robert E. Shewmaker

-1.

Technical Branch Division of Low-Level Waste Management and Decomissioning, NMSS

Enclosure:

BGV PLASAR 6.2, Q-l's

. )

DISTRIBUTION:

tEiiRNT"TTEP214?3: RLBangart PLohaus JJSurmeier JGreeves MBell RShewmaker JKane DWidmayer l

PDR

.Yes: /

/

y PDR

-No: /x/

Reason:

Proprietary /

/

or CF Only /T/

Ij i

ACNW

.Yes: L/

No:

/

SUBJECTABQTRACT:

BGV PLASAR COMMENTS

[l/l 1

0FC :LLTB j r NAME:RShewmMerfjl:

-DATE:57/p/89

01/ /89:01/ /89 :01/ /89

01/ /89 :01/ /89 :01/ /89 9004250185 900418 WASTE OFFICIAL RECORD COPY g,j g g

t

.s

~.

BGV PLASAR, Sec. 6.2 (BARRIER Code) l TB-ES-RS3 - 6. 2.1, 3. 2. 2, E.1. 5, E. 2. 5, Vol. III - Chapter 4, Vol. IV - Chapter 2 :

(pp. 6-59, 3-32, E-28 E-80, 4-1, 2-1)

[

The BGV PLASAR utilizes the BARRIER computer code to evaluate the depth of degradation of the concrete -in response to contact with groundwater containing sulfates.

The depth of degradation is utilized in projecting life expectancy.

~

of the concrete structures.

Various parameters related to the chemical and physical composition of the groundwater and the concrete-are, utilized in the evaluation, t

j (a) The formulation of the relationships between the various parameters was developed in the United Kingdom into an equation based on' empirical results derived from laboratory testing.

The test program's original scope was to l

simulate the conditions which might be present where concrete'was buried-in clay and was in contact with groundwater containing sulfates.

The i

researcher"s projected concentration of sulfates-in a clay soil environment L

is expected to be ~ 0.02 mole / liter, 'however,- in order to accelerate the testing the concentrations were increased.by nearly a factor-of ten.

Data

~

over a five year period were used in the formulation of the governing

. c empirical equation.

The derived equation has also been verified-against a single mass of concrete buried in clay for over 40 years.

The BGV PLASAR and the cited references contain several equations with the same basic for-l mat, but adjusted for parameters such as water-cement ratio or the,dif-fusion coefficient of sulfate through. concrete.

The BGV PLASAR on page E-29 presents a formulation that adjusts the depth of degradation based on a linear function of water-cement ratio, whereas' Reference 4 of Volume III, Chapter 4 to the BARRIER User's Manual and its reference, EPRI Report NP-5365, provide a formulation which is a function of the diffusion coefficient.

1

It is not clear what basis is being used in the BGV PLASAR to determine the loss of cross-section thickness nor is the relationship between these two formulations clear, although there is no doubt some correlation between l

diffusion / porosity and water-cement ratio.

The relationship, however, is not believed to be linear and is influenced by other additional parameters.

l t

'l If the BGV PLASAR were an actual application for a license these issues would have to be addressed.

(b) The development of the empirical relationships which form the basis of the equations discussed in (a) above was based on the depth of degradation of the concrete that was visually determined.

The depth of actual degradation.

is only detectable by microscopic examination.

Consequently, the formula-tions will tend to underestimate the true depth of concrete degradation.

In an actual license application, the applicant would have to address this-issue by either providing' additional information or. quantifying the magnitude-of the potential error when only visual observations were made.

(c) The formulation for the depth of degradation that utilizes. the diffusion coefficient for sulfate requires that this parameter be determined for the-concrete material being utilized.

Based on thi formula,'the depth of degradation can then be calculated.

Table 4 c of Volume _III of the BGV PLASAR lists the value of the diffusion coefficient for sulfate in concrete as 3 x 10 7 2

cm /sec., but the source of this value is not provided.

There-is great difficulty in determining the diffusion coefficient for sulfate ions in concrete because of the sulfates inherent in. portland cement con-crete and the difficulty of trying to identify the source of the ion, whether inherent or a migrant ion from an external source.

One study that-was completed using radioactive tracers provides some values for the dif-fusion coefficient.

The values determined for diffusion in cement mortars for sulfate ranged from 6.1 x 10 10 to 3.5 x 10 10 2

cm /sec.

The diffusion

Reference:

Spinks, et al., " Tracer Studies of Diffusion in Set Portland Cement," Canadian Journal of Technology, Vol. 30, No. 1, pp. 20-28, 1952.

2

i 4

l coefficient for sulfate ions will vary nonlinearly with the water-cement ratio since the ratio has an influence on the pore structure of the hardened cement paste of the concrete.

Other parameters also have an

[

influence on the pore structure such as various additives that'may be

-i utilized with the basic portland cement.

An actual BGV application attempting to utilize BARRIER would be required to provide the basis for the selection of the value for the diffusion coefficient as well as the influencing parameters and the expected range of values.

(d) The BGV PLASAR in Volume IV indicates that two concrete mix designs were utilized in the measurement of concrete properties.

However, these materials apparently differ from the ordinary portland cement concretes that were utilized in the basic work by Atkinson and Hearne to develop the empirical formula for calculating the depth of degradation.

The mixes proposed contain either microsilica (silica fume) or pozzolanic materials as admixtures that would have an effect on the basis of the empirical rela-tionship defined in the formula which utilizes the water-cement ratio as a variable.

While there is some evidence in the concrete industry that points to the fact that these additives would only improve the capability of the concrete to-endure sulphate exposures with less material loss, no. specific support of

~

this has been provided in the BGV PLASAR.

In an actual application it would be necessary to justify the applicability of the BARRIER Code formu-lations for the' loss of concrete cross-section as the result of sulphate attack.

(e) The BGV PLASAR apparently utilizes the formula on page E-29 to compute the-depth of concrete degradation that will occur over the design life of the specific class of BGV system and then provides for additional reinforcing steel cover to offset the loss of section.

As noted in the basic refer-enced research work by Atkinson, et al., the variability of the predictions was expected to range over + 30 percent for the specific conditions which formed the boundary of their testing.

While not addressing the specific 3

E conditions and boundary conditions or limits of applicability for the formulations, the BGV PLASAR does not address this 1 30 percent band.

i r

An actual application for a BGV would necessarily be expected to address a 130 percent limit of variability that was noted by the original authors of the formulation.

(f) The BGV PLASAR proposes the use of a Type II cement modified by a 15 percent by weight addition of silica fume (microsilica).

The basis for the use of the microsilica is stated to be the fact that a number of studies have indicated that concretes formulated with silica fume have resulted.in con-cretes with improved (lower) values for permeability.

The BGV PLASAR does not provide a simi'lar basis for use of the material from the standpoint of the impact of silica fume on the durability of concrete.

Information cur-rently available indicates that concretes formulated with microsilica exhibit an impaired resistance to magnesium sulfate attack when compared to normal concrete formulations.

This particular work was undertaken since there was a recognition of.the limited work done on the durability.

of concretes formulated with microsilica material.

The st'udy utilized Types I and V portland cements with 15 percent by' weight of silica fume acted upon by solutions of sodium and magnesium sulfate.

Data were observed at various intervals of time up to 140 days after the exposures.

Strain, strength and volume changes were each evaluated and x-ray differ-action analyses were conducted to detect the presence of the products of sulfate attack (ettringite, monosulfate and gypsum).

Based on the research the conclusion was that the use of silica fume increased the strength and mass loss by a factor of 5 to 10 when subjected to the magnesium sulfate r

solution.

In an actual BGV application it would be necessary to address the effect various additives used to enhance certain properties of concrete way have on other necessary properties.

This is especially true when the time scale involved in the desired service life is to be extended beyond the known experience.

4

i

Reference:

Cohen, Menashi D. and Bentur, Arnon, " Durability of Portland Cement-Silica Fume Pastes in Magnesium Sulfate and Sodium Sulfate Solutions," ACI Materials Journal, May/ June 1988, pp.148-157.

'(g)

In considering the Mg++ and SO --concentrations there appears to be a 4

difference within the BGV PLASAR document.

Appendix E, Section E.1.5 on page E-28 indicates that the MG++ concentration is taken as 20 ppm whereas in Volume I, Chapter 6, Section 6.2.1, Table 6.2-3 Volume III, Chapter in Table 4.3 on page 4.3 the Mg++ concentration is noted as 5.8 ppm.

c In an actual BGV application this apparent discrepancy would need.to be corrected or the reasons for the difference explained.

In an actual application it would be necessary to provide the basis for the use of any material which has.not been shown by experience to be non-deleterious for the characteristics which are desired in a reinforced I

concrete BGV.

TB-ES-RS4 - 6.2.1, E.1.5, E.2.5, (pp. 6-59, E-30, E-81) r The BGV PLASAR presents and uses an empirical formula in-the BARRIER code to compute the onset of reinforcing steel corrosion resulting from chloride attack.

The variables are the concrete cover, the water-cement ratio and the chloride ion concentration.

I (a) The formulation used in the BARRIER code as provided on page E-30 does not consider the interaction or influence of the extent of carbonation which can decrease the time to initiation of corrosion by lowering the pH of the i

pore water solution.

Neither is a discussion provided that would attempt l

to qualitatively address this aspect of the steel corrosion process-as necessary for a BGV concept, i

i l

5 l

In an actual BGV application there should be consideration of the effects of carbonation in any model used to predict the onset of chloride induced corrosion of the reinforcing steel.

(b) The basic formulation used in the BARRIER code was developed from the i

testing programs conducted in the fielt of highway engineering in California.

These were programs to attempt to define the critical parameters leading to corrosion and failure of bridge decks and highways i

in general.

Another program to define and evaluate the important parameters in the chloride attack mechanism on reinforced concrete was conducted by the Federal Highway Administration'(FHWA).

As a' result of the FHWA studies the empirical formulation from the California program was modified and it is that formulation which is used in the BARRIER code.

No explanation is provided for the basis of the formula in order to allow an evaluation to be made of its applicability for the engineered barrier concept.

For an actual BGV application where a numerical or a quantified determination of the service life is to be made, the applicant would have to clearly describe and define the limiting conditions of any formulations being used.

The boundary conditions and assumptions would have to be clearly defined.

(c) The formulation being utilized in the BARRIER code considers the time to initiation of corrosion to be a linear _ function of the water-cement' ratio.

The water-cement ratio is not considered to.be a substitute parameter for the diffu: ion coefficient for chloride ions in concrete; consequently, the model used to define the initiation of corrosion is not considered to be sufficient to attempt to predict that point in time since diffusion will be a mechanism for chloride transport.

In an actual BGV application further development or testing would be necessary to refine the model if it were going to be used in the projection of the service life of reinforced concrete.

1 (d) The BGV PLASAR document does not provide the formulation that-is used in the BARRIER code for prediction of the rate of corrosion of the reinforcing 6

+

I L

steel.

This is found in a fourth tier: document via written _and verbal L

references obtained during the review p'rocess as noted in the following.

The BGV PLASAR, Vol. III, Chapter 4, page 4-1 refers to a document desig-nated Reference 4 which is by R. Shuman, et al., " BARRIER:

A User's Guide L

and Documentation," Rogers and Associates Engineering Corporation, Draft Report RAE-8504/3-4, December 1987.

This document has been replaced by an EPRI document, EPRI NP-6218-CCML, "The BARRIER Code:

A Tool for Esti-mating the Long-Term Performance of Low-Leve1 Radioactive Waste Disposal

~

l Facilities - User's Manual," February 1989.

That document in Chapter 2, page 2-1 states that the detailed conceptual and mathematical models-upon which BARRIER are based-are given in two companion reports noted as Ref-erences 3 and 4.

Reference 3 is an EPRI document by R. Shuman, et al.,

l

" Performance Assessment of Low-Level Waste Disposal Facilities," EPRI I

NP-5745M, April 1988'and Reference 4.is another EPRI' document by R. D.'Baird, et al., " Design and Cost Methodologies for Low-Level Waste Disposal Facilities," EPRI NP-5365M, August 1987.

Both of these reports are supported by reports that are considered to be the supplemental reports which provide the detailed information.

These are designated with an "S" suffix to the report number and carry the same title.

The EPRI report, NP-5365S, identified the formulation being used in'the BARRIER code to predict the corrosion rate of the reinforcing steel.

The formulation is based on several assumptions: -(1) the growth of the iron oxide product layer on the reinforcing steel must be slow compared to the rate of oxygen diffusion through the layer:

(2) the rate of oxygen diffusion through the product layer is fast compared to the diffusion rates of oxygen through the concrete matrix containing the cement paste; and (3) the rate of oxygen consumed by the corrosion reaction is greater than the rate of diffusion of oxygen to the reaction interface, so that the corrosion rate is-limited by the flux of oxygen.

The BGV PLASAR provides no information on the validity of these assumptions.

If the BGV PLASAR were an actual application it would be necessary to p) ovide supporting information and background for these assumptions.

l 7

TB-ES-RSS - 6.2.1 (p. 6-591 l

The BGV PLASAR in utilizing the BARRIER computer code considers a model for the l

degradation of concrete that can arise from the leaching of calcium hydroxide.

l This degradation process can result in the loss of strength of the concrete since a portion of the matrix is lost.

l The formulation used in BARRIER is based on diffusion leaching which assumes that the loss of calcium hydroxide is controlled by the diffusion into the L

surrounding soil material; (that is, the diffusion is from a fixed concentra-tion into a serti-infinite domain).

The constant in the formulation is the expo-nent that arises frorr the originators of the formulation who indicate that when t

33% of the calcium hydroxide has been depleted, the concrete strength has-been cut to one-half its original value.

Based on the work of the originators of.

l_

the formulation, diffusion would dominate the leaching process only when the conditions in the concrete matrix are such as to provide a fluid (in this case the groundwater) flux density of less than 10 9cm/sec.

i The formulation used in BARRIER does not consider the possibility of adverse l

chemical content of the fluid on leaching of the calcium hydroxide.

An acidic fluid at the outset of the process will move the time scale radically forward, giving unconservative results with regard to lifespan predictions.

In an actual BGV application, the use of the formulation would have to be justified.

Other formulations for leaching may have to be utilized in order to represent actual field conditions such as cracking and local porosity and voids which may be present in an actual structure.

TB-ES-RS6-6.2.1, A2.1 (pp. 6-59,A-11J The BARRIER computer code as utilized in the BGV PLASAR does not specifically address the phenomenon of alkali-aggregate reaction that may arise from the reaction of the constituents of cement and the aggregate.

At the current time l

no known models exist to predict the failure time of concretes that may be reac-l-

.tive as a result of various quantities of Na20 and K 0 (alkalies) in the cement 2

and certain siliceous and carbonate constituents of the aggregate.

Consequently, 8

while BARRIER does not address this' mechanism that can disrupt the integrity of concrete and shorten its service life, it represents the current state-of-the-art (with respect to in that no known models exist for life span prediction relative to alkali-aggregate reactions.

One of the secondary reference documents l

to the BGV PLASAR, EPRI Report NP-5365S in Chapter 7, provides a verbal description of the degradation which can result from alkali-aggregate reactions and the l

major influencing parameters.

It concludes by stating that the degradation l

mechanism can be described similar to that used for calcium hydroxide leaching l

that results in a loss of strength of concrete as a function of time.

Such a formulation is, however, apparently not used in the actual BARRIER computer code.

The BGV PLASAR does not clearly utilize all the knowledge that does exist on preventing, minimizing or specifically addressing the alkali-aggregate reaction <

for siliceous materials.

Certain ASTM testing standardt are available for l

utilization in identifying potentially reactive aggregates and for testing aggregate-cement combinations.

The BGV PLASAR in Section A.2.1 does not specifically include these standard tests as being required for the concrete materials to be used in the BGV concept.

The specific standards'that should be utilized are as follows:

ASTM C227 - Standard Test Method for Potential Alkali Reactivity of Cement-Aggregate Combinations (Motor-Bar i

l Method)

L ASTM C289 - Standard Test Method for Potential Reactivity of Aggregates (Chemical Method)

ASTM C295 - Standard Practice for Petrographic Examination of Aggregates for Concrete.

For the effects of alkali-carbonate reactions that may arise from the reaction between the alkali in the cement and/or an external source and certain reactive carbonate rocks, other tests can be utilized to provide data on which to judge acceptability.

In addition to ASTM C295 noted above, ASTM C586, " Standard Test Method for Potential Reactivity of Carbonate Rocks for Concrete Aggregate (Rock Cylinder Method))" and a Canadian standard, CSA A23.2-14A should be used.

ASTM 9

is also developing a new standard test method that will basically measure the change of length of a standard form due to alkali-carbonate rock reactions.

Data obtained from the application of these standards and practices will aid in.

identifying potential reactive aggregates; however, even these may not identify all reactive combinations since some aggregates react very slowly or there may be unknown sources of_ alkalis and moisture transport and transportation can' concentrate alkalis which can then attack the concrete matrix.

In an actual BGV application it would be expected that the alkali-aggregate reaction phenomenon would be more clearly addressed for its impact on the service life of the BGV structures.

TB-ES-RS7-6.2.1, A5.2.2 (p. 6-59, A-5,8_]

The BGV PLASAR in predicting the failure time of the BGV vaults utilizes the BARRIER computer code.

Within the BARRIER code-logic is the capability to calculate the loss of concrete thickness as a result of the freeze-thaw attack that may occur.

The calculation is controlled in part, by the parameter "ICYC" that is defined as the number of significant freeze-thaw cycles per year.

(a) The BGV PLASAR does not provide a listing of the input for the specific studies using the BARRIER code, but instead apparently provides a selective listing of input parameters.

Reference document EPRI Report NP-6218-CCML on page 4-19 indicates that the parameter "ICYC" is input on Record #37 which is used only in the case where the parameter "KFLAG" is equal to the value of "2", meaning an above ground vault.

On this basis it appears that for a BGV where "KFLAG" is equal to "3", no computations are performed for free-thaw degradation.

There, is, however, exposure of various portions-of the vault to weather during and after construction and perhaps before the final cover is placed.

In Vol. II, Appendix A of the BGV PLASAR Section AS.2.2 presents the Design Criteria that are summarized in Table A5-1, page A-58.

It-is indicated in l

the table that the reinforced concrete properties with regard to the weather resistance of aggregate are such as to be acceptable for at least 300 10

t freeze-thaw cycles.

It-is not clear whether this is intended to ind'icate that the aggregate alone has been " tested" for this life span or whether this refers to the entire concerete matrix.

It is also not clear how this 300 year value is related to any analysis performed for the BGV using BARRIER.

In an actual BGV application it would be necessary'to clearly indicate what considerations for freeze-thaw effects were made. This information would be necessary to make an assessment of the specific conditions governing the BGV site and conditions.

(b) For conditions where the BARRIER code is utilized to compute the freeze-thaw effects on the rate of concrete degradation, a computation is performed to determine the average thickness-of the annual concrete loss from a section. The bases for the equation utilized is a series of test data',

assumptions and theoretical development that are not necessarily compati-ble or consistent.

First, the number of freeze-thaw cycles to cause con-

' crete failure _ based on test data resulting from a specific test procedure in order to reduce the dynamic modulus of elasticity by one-half.

Second, the number of freeze-thaw cycles to reach this level of reduction in the dynamic modulus of elasticity is further defined to be a function of the entrained air content and the water-cement ratio.

A portion of this rela-tionship is defined by some test data.

Another portion of the relationship is derived from a generalized handbook curve which was intended to illustrate the relationship of air entrainment and water-cement ratio to. durability.

Test data, test parameter ran0e etc. were not identified.

From an assumed number of significant annual freeze-thaw cycles per year the total service i.

life in years could be obtained.

At this point in the formulation, another computer technique known as UNSAT-H is introduced as a means to compute the depth of water penetration-into the structural concrete.

The parameters defined as the major influ-ences were the amount of residual moisture in the concrete and the porosity of the concrete.

These were used to define the depth of concrete initially affected by freeze-thaw conditions.

From this an annual rate of degradation is computed.

1 I

l 11

.,j

t In an actual license application it would be necessary to provide a correlation between the depth and degree of saturation that occurred in the test specimens used to develp the 50% reduction in the dynamic modulus and the values assumed in the UNSAT-H calculations.

Additionally, the validity of using UNSAT-H for the concrete would need to be demonstrated and the range'of concrete parameters that were utilized would have to be.

[

provided.

TB-ES-RS8-6.2.1 (p. 6-59)

The BGV PLASAR and the BARRIER computer code do not address the mechanism of reinforced concrete degradation that is associated with carbonation..The major effect from this process is a lowering of the pH of the matrix and pore water as well as decreasing _the protection provided to the reinforcing steel.

The process of the reaction between carbon dioxide and cement components such as Ca(OH)2 and water to produce carbonic acid and carbonates such as calcite also produces physical changes in volumes in the process.

The rate at which-carbonation proceeds is primarily a function of the supply of carbon dioxide,-

whether from the air or the decomposition of organics in soils or waste mater-ial, the' diffusion rate of the carbon dioxide through the concrete, the permea-bility of the concrete, the degree of saturation of the concrete, and-the effec-tive pH of the concrete in the basic range.

The depth of_ carbonation is also a function of time, progressing roughly proportional to the square root of time based on some formulations.

l While the concrete of a BGV may not result in identifying carbonation _as a l

dominant mechanism leading to the degradation of reinforced concrete, the exist-ence of the mechanism should be recognized and considered in any analysis for i

longevity of the reinforced concrete structure.

The BARRIER computer code does not at this time consider this as a mechanism p

that leads to the degradation of reinforced concrete either as a separate l

mechanism or in combination with chlorides in the attack on the reinforcing

(

steel.

Development work to define a formulation describing the depth of carbonation with respect to time was recently provided in a research paper.

Some laboratory data are available to support the equations and the associated l

l L

12 L

constants, however, at this time work in this area is not considered to be sufficiently genera'ized to enable the development of a governing equation to define the depth and rate of carbonation for a given set of material and environmental conditions.

Reference:

Ying yu, L and Qui-dong, W, "The Mechanism of Carbonation of Mortars and_the Dependence of Carbonation on Pore Structure," Concrete Durability, American Concrete Institute, SP 100-98, Vol. 2, p. 1915, 1987.

In an actual BGV application and in any use of the BARRIER computer code for predicting concrete degradation, it would be expected that the effects of carbonation would be considered at least qualitatively, if not _ quantitatively.

TB-ES-RS9-6.2.1 (p. 6-59)

The BGV PLASAR and the BARRIER computer code do not address the mechanism of reinforced concrete degradation that is associated with acid attack.

The acids of particular concern as those of sulfur and nitrogen. -It should be noted that the degradation which.can result from the effects of sulfuric acid include'the attack mechanism considered in the sulfate attack on~ concrete as well as those effects associated directly with sulfuric acid.

It is these-aspects which have not been addressed by the BARRIER' computer code; however, in general, models for such degradation methods have not as yet been developed in any detail and only some of the variables have been identified.

A very

- simplistic relationship has been developed by Raju and Dayaratnam in order to estimate minimum cover for reinforcing steel.

The degree of deterioration of the cover was assumed to be linearly related to the depth of penetration of sulfuric acid, with full sulfuric-cement reaction occurring at the exposed surface and zero reaction at the maximum depth of acid penetration.

Some behavior has been reported that appears to be contrary to what would be expected, yet the empirical data from a significant series of tests do exist to support this fact.

The data appear to indicate that an increase in the cement factor (the amount of cement utilized per cubic yard) would increase the amount of degradation for certain values of the cement factor with the-pH level in the acidic range.

13

7 The acid attacks can originate from various sources depending on the location of the reinforced concrete member.

Airborne acids can be deposited via contami-nated natural precipitation as in acid rain.

The attack may arise from a gaseous state in the form of sulfur dioxide.

For above ground structures there it, a condition of repeated wetting with acidic water possible.

For structures that-are buried and are intended for a serice life of hundreds of years the acidity of the earth and contained groundwater may increase with time.

The acid source may emanate from the dumping of acids or chemical wastes that can react and form acids or the source may be the leaking of acids and chemicals from an industrial use.

An actual BGV application and any use of the BARRIER computer code for predicting concrete degradation, would be expected that the effects of acid attack would be considered.

References:

1 l

i Raju, P.S.N. and Dayaratnam, P., " Durability of Concrete Exposed to Dilute Sulfuric Acid," Building and Environment, V. 19, No. 2, 1984, pp. 75-79.

Fattubi, Nijad I. and Hughes, Barry P., " Ordinary Portland Cement Mixes with Selected Admixtures Subjected to Sulfuric Acid Attack," Materials Journal of the American Concrete Institute, Nov, and Dec. of 1988, Title No. 85-M50,'p. 512.

Ahiogbe, Emmanual and Rizkalla, Sami, " Response of Concrete to Sulfuric Acid Attack," Materials Journal of the American Concrete Institute, Nov and Dec.1988, Title No. 85-M46, p. 481.

TB-ES-RS10-6.2.1 (p. 6-59)

The BGV PLASAR and the BARRIER computer code do not address the degradation process of reinforced concrete that can. be initiated and sustained by biological

. conditions.

These effects can attack the concrete or the reinforcing steel', or both.

Historical evidence indicates that of the four general groups of micro-biological organisms (bacterial, fungi, algae and yeasts) the most-likely to be.

14

I f

of concern are bacteria.

Specifically, bacteria that are_ sulfate _ reducing bac-teria are those of most concern.

This in general means that the energy for the life processes of these bacteria are derived from the oxidation of.some element other than carbon, and those of particular concern thrive under certain conditions on sulfur.~

The bacteria that have been found to be most destructive to reinforced concrete are those of the genus Thiobacillus.

In general, the sulfur bacteria are likely to be found wherever warmth, moisture and reduccd compounds of sulfur are present.

The sulfate reducers are found all over the world in soils and waters and represent some of the most ancient of the living organisms on earth.

Conditions to sustain the degradation process by the-Thiobacillus bacteria are sufficient moisture to prevent the dessication of the bacteria and there must be sufficient supplies of carbon dioxide, oxygen, nitrogen compounds and hydrogen sulfide.

Soluble compounds of phosphorus, iron and other trace elements are also needed in the moisture film.

A source of hydrogen sulfide necessary for the conditions to support the Thiobacillus bacteria is another genus of-sulfur bacteria; however, these are anaerobic as opposed to the aerobic Thiobacillus genus.

These are known as sulfate reducers also and are those of the genus Desulfotomaculum.

They have the ability to I

reduce sulfates that are present in natural waters or the surrounding environ-ment and produce hydrogen sulfide as a waste product.

An environment where reinforced concrete exists with these conditions and a pH of about 9 will allow for the growth of Thiobacillus thioparus that utilizes the hydrogen l

sulfide and generates thiosulfuric and polythionic acid.

As the growth of these bacteria continue the pH will continue to decrease which, if it reaches a value of approximately 5 should create conditions for the growth of another of the genus Thiobacillus, namely Thiobacillus concretivorous.

These bacteria l.

produce high concentrations of sulfuric acid and can drop the pH to a level of l

2 or less.

This then creates the sulfuric acid degradation conditions where the decomposition of the calcium silicates and aluminates in the cement l

matrix is destroyed.

One of the unique features of the microorganisms is their small size so that they cen exist in small areas and grow.

If localized effects can be important to the functioning of the materials, then these microorganisms can have signi-ficant impacts.

Consequently, what are generally believed to be the prevailing environmental conditions on the global scale may not be a good predictor of 15

\\

what may exist on the smaller scale where bacteria growth may be' fostered.

For example, the anaerobic sulfate reducers can exist under conditions that normally would be considered obviously aerobic.

Sulfate reducers have been found under slime deposits and iron bacteria colonies in aerated water systems such as I

those served by cooling towers.

The slime or other colonies serve to shield the sulfate reducers from air, thus allowing the anaerobic bacteria to thrive.

Additionally, one cannot easily justify that an environment for microbiological attack is not possible because there is no moisture.

Trace amounts of moisture are sufficient to qualify a system as aqueous when consideration must be made_on the microorganism level.

Jet engine fuel, for example, qualifies as an aqueous system since trace amounts of water can be present in the material.

While the instances of microbiological attack on reinforced concrete is not documented as a wide-spread problem, there are also many times an inappropriate investigation' of the distress, degradation or failure since the usual response by an owner is to get the problem corrected.

The finences are expended on that effort, not on the past to define what went wrong.

Consequently, historical' j

data may not reflect the past accurately.

For an actual BGV application and the use of BARRIER where there is an emphasis on long-term durability, it would be important to address the biological mecha-nism when attempting to define the projected life of engineered reinforced con-crete barriers.

Parallels exist with respect to the corrosion of steel by sul-i fate reducing bacteria that should be addressed also in an actual BGV application or use of BARRIER.

f TB-ES-RS11-6.2.1 (p. 6-59)

The BGV PLASAR and the BARRIER computer code do not specifically address the considerations that must be made concerning a range of chemicals and materials that can be detrimental to concrete.

An extensive list of these materials is available in publications of the Portland Cement Association and the American Concrete Institute.

Sources for such materials should be precluded from the environment of a BGV or the degradation effects of the chemical or material should be considered in any prediction of the service life of the reinforced concrete that is made.

Each detrimental chemical or material should be addressed i

i

.=

s.

in the text of the BGV PLASAR or considered as detrimental.

Consideration should also be given to conditions under which one of the chemicals or materials might be introduced to the environment of the BGV af ter construction whether by dumping or by formation from constituent materials and chemicals in the natural environment of the BGV site.

In an actual BGV PLASAR application, the consideration given to the various chemicals and materials listed in the literature that can degrade reinforced concrete should be explained.

The BARRIER computer code should contain a method to address these chemicals or materials or provide clear documentation of the limiting conditions for application of BARRIER.

1 i

References:

[

Portland Cement Association, " Effects of Substances on Concrete and Guide to Protective Treatments, PCA 15001, 1989.

American Concrete Institute, ACI Committee 201, Durability of Concrete Inservice, ACI 201.2R-77.

l TB-ES-RS12-6.2.1 (p. 6-59)

The BGV PLASAR and the BARRIER computer code do not address the mechanism of salt crystallization that has been suggested recently as a possible means of l

internal stresses being increased in the concrete matrix This concept is i

based on the precipitation _of dissolved soluble salts in the pores of the i

concrete matrix as a result of water evaporation.

Once the pores have been filled with the salts crystals additional crystal growth may result in devel-opment of internal stresses that may lead to cracking of the concrete if the tensile strength of the concrete matrix is exceeded.

This mechanism has been noted as a major cause of stone deterioration.

In a BGV application if i

such an attack mechanism were to occur the salt build-up would occur on the inne surface e' the vault walls as water passed through to a surface exposed to drier conditions than the outside of the vault.

l 17

~..

1 In an actual BGV application it would be. expected that this new mechanism that may be.a contributor to concrete degradation in certain conditions should be investigated and considered appropriately.

t

References:

(

Winkler, E.M., " Stone"- Properties, Durability in Man's Environment,"

Springer-Verlag, Vienna, 1973.

Sayward, J. M., " Salt Action on Concrete," Special Report 84-25, U.S.~ Army

. I Cold Regions Research and Engineering Laboratory, Hanover, N.H.,1984.

TB-ES-RS 13-6.2.1 (p. 6-59)

The BGV PLASAR that utilizes the BARRIER computer code to evaluate the potential.

j 1

life of the reinforced concrate vaults _does not address the potential for.the attack by.any constituents of the stored ' low-level wastes within the BGV.

j Probably the most detrimental substance would be the ~ release-of any acids l

l that would be = capable of degrading the reinforced concrete vault ' structure.

l These could be in the form of acids within the low-level waste that could be j

released or compounds or elements that could combine and upon-hydrolysis or j

oxidation, produce aggressive acids.

The aggressive acids could be either I

1 inorganic or organic acids.

l i

i In an actual BGV application it would be expected that the possibility of acid attacks on the vault structure from within the vault would be discussed and appropriately considered in the BGV safety evaluation.

j i

1 TB-ES-RS14-6.2.1 (p. 6-59) l

/

The BGV PLASAR and the use of the BARRIER computer code appear to address the l

determination or projection of the service life of the' reinforced concrete on the basis of what can be considered to be nearly perfect conditions.

No pro-visions appear to have been provided for integrating the use of post-construction l

18 r

~

testing to assess to what degree the design assumptions related to the construc-tion materials, construction methods and quality assurance / quality control, etc.

have been met in the field in producing a BGV.

Until such an assessment is per-formed on each completed BGV, the structure could not be assumed to have met the design requirements and be ready to receive wastes.

The first level of assessment would include the evaluation of the in process test data and observations made

-during such activities as concrete production, transport and placement.

~

The second level of assessment would include a visual survey of the concrete surfaces once the forms are removed.

All defects that could cause the reduc-tion in a material physical or chemical property ~ assumed in the design phase would be recorded so that the information could be utilized in an analysis of I

the resulting BGV to perform as designed.

The third assessment would include another complete visual survey once the prescribed curing has been completed.

Since the quantitative values of the parameters necessary to evaluate the important characteristics for the satisfactory performance of a BGV, whether based on a computer code such as BARRIER or not, would need to be evaluated, various testing methods would need to be introduced to determine'the actual s

as-built conditions.

Such tests should, in, most cases, only be performed after a BGV has stabiiized from the standpoint of curing, moisture loss, shrinkage etc.

This will require that a passage of time be allowed after construction

-before such testing can be performed.

The testing for this fourth stage should consist of a preplanned program that has been formulated during the design phase so that all important parameters that may influence the behavior of the BGV are verified to an adequate degree.

This program may involve the use of impact echo or pulse echo techniques, impact hammers, penetration devices, testing of cast-in place specimens, coring and petrography.

Such an as-built program should address issues that would, for example, include the in place compressive strength, the degree and uniformity of consolidation, the air content, the location, size and orientation of cracks, information on any cold joints and the construction joints.

An evaluation of the permeability of the as-built BGV should be determined in order to assess the ability of the structure to retard any groundwater flow into the volume.

l l

19

4-b a

In an actual BGV application it would be necessary to address the procedures that would be used to determine as-built conditions of a BGV in order to pro-vide relevant information to assess its capability to perform intended design functions.

If a computer code such as' BARRIER had been utilized in the design phase, then it would be reused in conjunction with the as-built conditions t

imposed on the original model used in the analysis.

If another form or type of analysis was utilized that too would be reevaluated based on the as-built conditions.

A final step in this process is the declaration or certification by the designer and builder that the BGV meets the design requirements and is ready to receive waste.

TB-ES-RS15-6.2.1 (p. 6-59)

The BGV PLASAR that uses the BARRIER code to predict structural lifetimes does not provide for repair methods that may have to be' utilized in the field prior to the structure receiving its certification and acceptance.

The program should encompass the details of a repair program consistent with the types of repairs that may be necessary.

3 In an actual BGV application it would be necessary to provide a repair program addressing the specific materials etc. being used in the BGV.

20