Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports (2019)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 As described in Chapter 1, ASR, at various levels and severities, has been confirmed in the 48 contiguous states (Thomas et al. 2013c) and in many concrete pavement facilities worldwide, as well as in bridges and other concrete structures. Although the development of the distress can be moderately slow, it is progressive in nature and can produce significant concrete deterioration that leads to reduced serviceability and premature failures. To help provide a foundation for this project, a literature search was conducted to document the ASR problem and determine the experiences of agencies in mitigating the distress on pavement facilities. Based on that search, this chapter provides a brief history of ASR, looks at the mechanisms of ASR development, describes how ASR is identified in the field, and reports on measures for preventing ASR in new concrete and mitigating ASR in existing concrete pavement structures. ASR History Highlights The identification of ASR and its associated mechanisms is commonly ascribed to work done in the late 1930s and early 1940s by Thomas Stanton, an engineer with the California Division of Highways. Prior to that time, most aggregates used in concrete construction were generally considered to be inert, but Stanton identified a deleterious reaction that occurred between reactive silica components of certain aggregates and the alkalis in the cement paste (Stanton 1940a,b, 1942). Stanton observed that the reaction created expansive forces that often led to cracking in the concrete and possible ultimate failure of the structure (Stanton 1940b). In a 1938 survey, Stanton documented one of the first concrete pavement projects with ASR as occurring on a section of roadway north of Bradley, CA, and was able to trace the problem to the use of a local fine aggregate as the source of the problem (Stanton 1940b). At about the same time, researchers from the U.S. Bureau of Reclamation confirmed the presence of ASR as a contributor to cracking on several dams (Meissner 1941). As part of his work, Stanton concluded that if portland cement did not contain in excess of 0.60% alkalis (expressed in terms of Na2O equivalent), no detrimental expansion would result (Stanton 1940b). He also documented the potential use of pozzolans to help control ASR, and his early work led to the development of the first standard laboratory test method for identifying the potential reactivity of cement-aggregate combinations, issued in 1950 as ASTM C227, Tentative Method of Test for Potential Alkali Reactivity of Cement-Aggregate Combinations. Building on Stanton’s pioneering efforts, many agencies soon adopted the 0.60% limit as the definitive means of preventing unwanted expansion as the result of ASR, with reasonably good success. For example, a 1955 interim report by the state of California showed no evidence of C H A P T E R 2 Literature Review Summary

10 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports harmful ASR on 42 concrete bridges, 2 pavements, 1 tunnel, and 56 minor structures that were built between 1940 and 1954 after the adoption of the limit (Tremper and Kovanda 1955). Although the limits on equivalent alkali contents were helpful, over the next several decades, the inability to definitively determine the potential reactivity of aggregates exposed to a range of environmental conditions hampered efforts in controlling or minimizing ASR distress. Part of this was related to some characteristics of the tests that were used (e.g., test duration, definition of expansion threshold limits), but it was also recognized that many of the tests were not suitable for use with more slowly reactive aggregates. In addition, changes in portland cement manufac- turing and the complexity of modern mix designs also contributed to difficulties in managing ASR. In response, the Strategic Highway Research Program (SHRP) initiated a major program in the late 1980s to help address some of those inadequacies and to provide improved guidance on addressing ASR (Stark et al. 1993). At about the same time, the Canadian Standards Associa- tion released a set of revised guidelines in 1994 for preventing the risk of damage due to ASR in new concrete construction, featuring recommendations related to aggregate testing, low alkali cements, and the use of SCMs (Thomas et al. 1997). In the United States, the SHRP work served as a springboard to some initiatives seeking to provide increased guidance on managing ASR. For example, the Federal Highway Administra- tion (FHWA) published a series of reports in 2002 providing guidelines for the field surveys and laboratory identification of MRD on concrete pavements, including ASR (Van Dam et al. 2002a,b; Sutter et al. 2002). Similarly, the Innovative Pavement Research Foundation (IPRF) funded several studies looking at mitigation treatments for ASR on airfield pavements (Rangaraju 2007; Whitmore 2009; Zollinger et al. 2009), as well as a study on establishing a field protocol for evaluating airfield pavements with MRD, including ASR (Van Dam et al. 2009). In addition, the DOD prepared a report in 2006 that presented an overall assessment of the condition of the DOD facilities and infrastructure with respect to ASR (DOD 2006). This was accompanied at about the same time by an Engineering Technical Letter (ETL) issued by the Air Force Civil Engineer Support Agency (AFCESA) that provided comprehensive guidance to Air Force civil engineering units on dealing with ASR (AFCESA 2006). FHWA continued its efforts in providing guidance and direction to agencies in its ASR Development and Deployment Program, a multiyear program initiated in 2006 that was designed to increase the durability and performance of concrete pavements and structures; that program produced some major guidance documents related to ASR (Thomas et al. 2007; Thomas et al. 2011; Thomas et al. 2013a; Thomas et al. 2013b; Thomas et al. 2013c; Thomas et al. 2013d). This work led to the development of the current American Association of State Highway and Transportation Officials (AASHTO) protocol for dealing with ASR (AASHTO R 80) as well as the ASTM guide for reducing the risk of ASR (ASTM C1778). In 2012, the FAA updated their ASR specification contained in their Advisory Circular on Airport Construction (current version 150/5370-10G) (FAA 2014). The major change was the assessment of the 0.10% expansion limit at 28 days (instead of 14 days) for testing done using both ASTM C1260 (for coarse and fine aggregates tested separately) and ASTM C1567 (modified for testing of coarse and fine aggregate combined in the job mixture proportions). Similarly, DOD released a new specification in 2015 with updated ASR guidance, requiring that both fine and coarse aggregate for all concrete be tested for alkali reactivity and specifying low alkali cements and SCMs (DOD 2015). As a result of these efforts and initiatives, it is generally accepted that ASR distress can be pre- vented in new concrete construction (although only through long-term monitoring of recent projects can this be confirmed). However, the ability to address and mitigate ASR in existing structures remains an ongoing challenge.

Literature Review Summary 11 Mechanisms of ASR Development As defined previously, ASR is a deleterious reaction that occurs in concrete between reactive silica present in the aggregate and the alkalis present in the pore solution of the hydrated cementi- tious paste. The reaction entails the dissolution of the silica, the formation of an alkali-silica gel, and interaction with calcium. As the gel takes up water, it expands significantly, destroying the integrity of the weakened aggregate particle and the surrounding cement paste. The reaction most often leads to the development of a map-cracking pattern on the pavement surface, often starting at the joints and then working across the entire slab. On continued development, the cracks can begin to spall and deteriorate, and the pavement may also exhibit blowups, shoving, and other pressure-related damage (Van Dam 2002a). In some cases, a gel exudate may appear in or surrounding the cracks. ASR may appear as early as 5 years after construction, although 10 to 15 years may be more typical. In some cases, ASR may not appear for 20 years or even later. There are three requirements for deleterious alkali-silica reactivity to occur, as described below and illustrated in Figure 2 (Thomas et al. 2011): 1. A sufficient concentration of alkali hydroxides (sodium hydroxide, NaOH, and potassium hydroxide, KOH) in the pore solution of the concrete. The main source of alkalis in concrete is the portland cement, but additional alkalis may come from other concrete components (e.g., aggregates, admixtures) or from external sources (e.g., deicing salts, seawater). 2. A sufficient quantity of unstable silica in the aggregate. 3. A sufficient supply of moisture in the concrete. The ASR reaction ceases below a relative humidity of 80%, but increases in intensity as the relative humidity within the concrete increases from 80 to 100%. The elimination of any one of these requirements will prevent the occurrence of damaging alkali-silica reaction. Figure 2. Sequence of ASR development (Thomas et al. 2011).

12 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports As explained by Thomas and colleagues (2013a), ASR is initiated by a reaction between the hydroxyl ions in the pore solution and certain types of silica in the aggregate. In the highly alkaline environment, certain forms of “unstable” silica tend to dissolve out, making them available to react with the alkalis. As the alkalinity of the pore solution increases, the potential for alkali-silica reaction increases as more stable forms of silica become susceptible to attack (Van Dam et al. 2002a). The result is an alkali-silicate gel that absorbs water and swells if in a moist environment. It is believed that to cause significant damage, the alkali-silica gel must contain cal- cium, although the role of calcium is not fully understood (ACI 2016). The gel takes on water and swells, fracturing affected aggregate particles and surrounding paste (see Figure 3). The gel that is produced through the reaction appears as a glassy-clear or white powdery deposit within reacted aggregate particles (Van Dam et al. 2002a) and may not always be visible to the naked eye. A more detailed description of the ASR reaction process, including the chemical reactions that occur and the mechanisms of expansion, is provided by others (Thomas et al. 2013a; Poole 2017). Some different aggregate types are susceptible to ASR. Thomas et al. (2013a) indicate that the most important deleteriously alkali-reactive rocks are opaline cherts, chalcedonic cherts, quartzose cherts, siliceous limestones, siliceous dolomites, rhyolites and tuffs, dacites and tuffs, andesites and tuffs, siliceous shales, phylites, opaline concretions, and limestone-filled quartz and quartzites. Common susceptible aggregate types worldwide by country/region are summarized by Poole (2017). The rate of the ASR reaction can vary considerably among the various aggregates, but is also influenced by other factors such as the environmental conditions and exposure to certain deicing chemicals. In that regard, several studies have suggested that the application of certain deicing and anti- icing chemicals used on airfield pavements (e.g., potassium acetate, sodium acetate, sodium formate, and potassium formate) may contribute to the expansion of ASR-susceptible aggregate and may also provoke the reactions in aggregates that did not previously exhibit ASR tendencies (Rangaraju and Olek 2007; Rangaraju 2007). Among the different deicers evaluated, potassium acetate caused the most severe attack in test specimens with reactive aggregate, and high-alkali cements showed a more rapid rate of deterioration (Rangaraju and Olek 2007). An evaluation of mitigation methods indicated that low- and intermediate-calcium fly ashes, slag, and lithium admixtures are able to considerably reduce expansions in mortar bar specimens exposed to 2.5 mm Figure 3. Aggregate particle fracturing due to ASR (courtesy of Karl Peterson).

Literature Review Summary 13 potassium acetate deicer solutions when used at adequate dosage levels (Rangaraju 2007). Interestingly, while laboratory studies conducted at high temperatures (176 oF [80 oC]) and under high alkaline conditions have shown that high concentration potassium acetate solu- tions are very capable of inducing and promoting ASR distress in concrete, such evidence in field conditions was not readily evident in a follow-up study on concrete airfield pavements (Rangaraju and Olek 2011; Balachandran et al. 2011). This is thought to be because the potassium acetate is able to penetrate only a few millimeters into the concrete. Identifying ASR in Pavements Field Surveys The initial identification of ASR typically begins with the observation of telling distress manifestations and features. Thomas et al. (2011) indicate that the classic symptom of ASR is map cracking, which takes the form of randomly oriented cracks on the surface of concrete elements that are relatively free (unrestrained) to move in all directions. Along with that, dis- coloration or staining around the cracks is also often observed, and a gel exudate may appear in the cracks (Van Dam et al. 2002a). Other visible signs of ASR can include spalling, extrusion of joint sealant material, surface popouts, and expansion that causes displacement and damage to in-pavement fixtures and adjacent structures. Figure 4 presents some of the various signs and symptoms of ASR distress on airfield pavements. Map cracking, spalling, and pressure-related damage and characteristics can also be caused by some other factors, so their presence does not necessarily mean that ASR distress is occur- ring. Moreover, the similar distress manifestations exhibited by many other MRD types make it extremely difficult to positively identify ASR and its associated mechanisms based on visible distress indications alone, and it is only through a petrographic analysis (ASTM C856) that determination of causation can be established. PCI Surveys Most airfield pavement condition surveys are conducted using the Pavement Condition Index (PCI) procedure, as documented by ASTM D5340. This procedure was originally developed for the Air Force in the 1970s and is used at airports throughout the world, provid- ing a systematic methodology for assessing pavement condition. Beginning in 2012, the PCI procedure presented in ASTM D5340 added ASR as a specific distress on its airfield pavement protocol (but, interestingly, not on its roadway pavement protocol), with the following severity categories defined: • Low: Minimal to no foreign object debris potential (i.e., loose pieces) from cracks, joints or ASR-related popouts; cracks at the surface are tight (predominantly 0.04 inches [1.0 mm] or less). Little to no evidence of movement in pavement or surrounding structures or elements. • Moderate: Some FOD potential, but increased sweeping or other FOD removal methods may be required. May be evidence of slab movement or some damage (or both) to adjacent structures or elements. Medium ASR distress is differentiated from low by having one or more of the following: increased FOD potential, crack density increases, some fragments present along cracks or at crack intersections, surface popouts of concrete may occur, pattern of wider cracks (predominantly 0.04 inches [1.0 mm] or wider) that may be subdivided by tighter cracks. • High: One or both of the following exist: (1) loose or missing concrete fragments that pose high FOD potential, and/or (2) slab surface integrity and function significantly degraded and pavement requires immediate repairs; may also require repairs to adjacent structures or elements.

14 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports a) ASR map cracking. c) Severe ASR cracking and spalling. e) Compressed expansion joint due to ASR movement. b) Close-up of cracking and exudate. d) ASR spalling at joint. f) Slab shifting due to ASR movement. Figure 4. ASR signs and symptoms (courtesy of Applied Pavement Technology, Inc.).

Literature Review Summary 15 The PCI procedure does recognize that ASR is material dependent and states that coring and concrete petrographic analysis is the only definitive means of confirming ASR. Nevertheless, the procedure makes a diagnosis based on visible conditions only, and with the absence of a “map cracking” distress call in the procedure, the result is that PCI surveys performed on airport pavements may report an ASR condition that may or may not be legitimate. The FAA formerly had an Advisory Circular (AC) for the identification of alkali-silica reaction in concrete airfield pavements, but that AC was cancelled in 2011. Materials-Related Distress Rating (MRDR) Procedure Because of the issues associated with the PCI procedure in identifying ASR (and other MRD), the Innovative Pavement Research Foundation (IPRF) developed an MRD survey tool in 2009 that is complementary to the PCI procedure and is useful in noting the presence and progres- sion of MRD (Van Dam et al. 2009). Furthermore, the MRD survey tool can be used to support decisions regarding the maintenance, rehabilitation, and replacement of pavements in much the same manner that the PCI currently supports those decisions as part of the pavement manage- ment process. The MRD survey process includes an examination of several signs or indicators of MRD, including staining, pattern cracking, perpendicular cracking, parallel cracking, popouts, sliver spalling, exudate in cracks, signs of expansion, and joint disintegration and scaling (Van Dam et al. 2009). The output of the survey, the MRDR, is a single value that combines the various MRD distresses for a section of pavement. That resulting MRDR value provides insight into the prevalence of MRD and whether maintenance activities are required to address conditions and minimize the risk of FOD. The MRD survey procedure has seen some limited use at selected airports with known MRD issues. Recent FHWA Guidance Thomas et al. (2013a) describe general guidelines for performing a field survey of concrete structures that may be affected by ASR. The general concept is to identify defects that are spe- cific or unique enough that the distress can be attributed to ASR. For pavements, the most common distress manifestations associated with ASR are noted to include map cracking, joint deficiencies/deterioration, and, in some cases, popouts. A supporting document is available with numerous photographs depicting common ASR manifestations and severity levels (Thomas et al. 2011). Thomas et al. (2013a) acknowledge the difficulty in determining from field surveys whether ASR is the primary (or only) factor that is leading to the observed distresses and deterioration. Based on distress data alone, Table 1 was developed to provide guidance on the probability that ASR may be contributing to the observed deterioration. Field and Laboratory Testing Although the visible features of ASR distress can be strongly suggestive, only through additional field and laboratory testing is it possible for the specific distress mechanisms (and hence the type of distress) to be confirmed. These additional tests are described below: • Coring. Concrete core samples can be retrieved for visible inspection to determine the depth of deterioration, the depth and tortuosity of the cracking through the concrete (and either around or through the aggregate), and the presence of “reaction rims” around the coarse aggregate. The core samples should be collected following established guidelines (e.g., ASTM C42) and proper identification and packaging/storage protocols to prevent drying and contamination prior to and during transport to the laboratory (Van Dam et al. 2002b).

16 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports • Field Identification Tests. Several test procedures are available that have been used in the field identification of pavements suspected of suffering from ASR. These tests use chemicals that are applied to a freshly exposed concrete surface in the field to indicate whether ASR is present. – Application of uranyl acetate solution. This test uses a uranyl acetate compound that stains the products of the ASR reaction to make them visible under ultraviolet light. The test requires experienced technicians and proper interpretation and does not distinguish between the harmless presence of ASR gel and destructive damage developing from ASR reactions; furthermore, the fluorescence may be caused by sources other than ASR gel (AFCESA 2006). The uranyl acetate solution contains a low-dosage radioactive compound and may be subject to safety and disposal restrictions. – Application of sodium cobaltinitrite and rhodamine compounds. These tests apply non- radioactive chemicals (sodium cobaltinitrite and rhodamine solutions) to the prepared area. The sodium cobaltinitrite reacts with soluble potassium to produce a yellow precipi- tate, while the rhodamine reacts with calcium-rich ASR gel to form a pinkish precipitate (Guthrie and Carey 1999). In essence, the yellow stain indicates potassium-rich silica gels thought to represent active ASR and the pink stain shows calcium-rich silica gels thought to represent older ASR (Gore et al. 2014). • Laboratory Testing. As mentioned previously, laboratory testing is essential for the iden- tification and confirmation of ASR. This testing is normally performed on samples from the field coring program and should consist of a petrographic examination that is performed in accordance with ASTM C856 by an experienced petrographer. Detailed information on conducting petrographic analyses of concrete samples for ASR confirmation are provided by Van Dam et al. (2002a,b); Thomas et al. (2013a); and Godart and de Rooij (2017). One quantitative output that can be produced based on a petrographic analysis is the Damage Rating Index, or DRI. The DRI is a method for evaluating the condition of concrete by counting the number of typical features of ASR (e.g., cracks, reaction rims, ASR gel in void spaces, etc.) on polished concrete sections (16x magnification) over an area of at least 31 in2 (200 cm2) (Godart and de Rooij 2017). In essence, it gives a measure of the amount of deterioration due to ASR and can be useful in providing benchmark condition levels and in monitoring the progression of the distress. As part of the detailed evaluation, relevant project documentation (e.g., design and construc- tion reports, mix design information, materials testing reports, etc.) should be gathered and reviewed, as they can provide useful insight into potential causes or contributing factors. Feature Low Probability for ASR Medium Probability for ASR High Probability for ASR Expansion and/or displacement of elements None Some evidence (e.g., closure of joints in pavements, jersey barriers, spalls, misalignments between structural members) Fair to extensive signs of volume increase leading to spalling at joints, displacement and/or misalignment of structural members Cracking and crack pattern None Some cracking pattern typical of ASR (e.g., map cracking or cracks aligned with major reinforcement or stress) Extensive map cracking or cracking aligned with major stress or reinforcement Surface discoloration None Slight surface discoloration associated with some cracks Many cracks with dark discoloration and adjacent zone of light-colored concrete Exudations None White exudations around some cracks; possibility of colorless, jelly-like exudations Colorless, jelly-like exudations readily identifiable as ASR gel associated with several cracks Table 1. Classification system for condition survey results (Thomas et al. 2013a).

Literature Review Summary 17 Preventing ASR in New Concrete Construction General Approach Tremendous advancements have been made in the past decade or so with regard to the prevention of ASR in new concrete construction. Basically four strategies are available to mitigate ASR in new concrete mixtures (Thomas et al. 2013a): 1. Avoid the use of susceptible aggregates. Aggregate reactivity can be assessed using a thorough testing program featuring a combination of petrographic analysis of aggregates (ASTM C295) and expansion testing of mortar (ASTM C1260) or concrete (ASTM C1293). This information, when coupled with an evaluation of the field performance of pavements containing those aggregates, can help identify the potential risk for ASR. However, this option may not always be the best for preventing ASR in new construc- tion because in many cases non-reactive aggregates simply may not be available in certain locations, or perhaps local reactive aggregates with otherwise suitable properties for use in concrete are available at low cost. Moreover, some uncertainties can be associated with the laboratory test results. 2. Use supplementary cementitious materials. This is a widely used and very efficient approach to addressing ASR in new concrete construction. Although possibly contributing a small amount of alkalis, the addition of SCMs is effective because it will combine with alkali hydroxides in the concrete and produce additional calcium silicate hydrate, thereby reducing the alkali hydroxides in the concrete (Taylor 2015). ASTM C618 Class F fly ash is generally considered most effective in addressing ASR because of its low calcium content, although slag cement (ASTM C989), silica fume (ASTM C1240), and natural pozzolans (ASTM C618 Class N) can also be effective, depending on the nature of the SCMs and the quantities used. Table 2 provides general recommendations regarding typical SCM levels, expressed as a percentage of the total cementitious materials. It is emphasized that SCMs have varying degrees of effectiveness and should be thoroughly tested to ensure their impacts on control- ling ASR. ASTM C1567 can be used to evaluate the effectiveness of SCMs in mitigating or controlling ASR. The availability of fly ash has become an issue for many agencies in the past decade due to regulatory directives, changes in the energy sector, and economic conditions. A 2016 survey of state highway agencies, the FAA, and DOD indicated that all 52 respondents use fly ash, with more than 80% of the respondents reporting supply issues (AASHTO 2016). Agencies also expressed concern about the availability of Class F ash to address ASR and the overall consistency in the fly ash they received. As fly ash use is anticipated to grow, Type of SCM Level Required(as % of total cementitious materials) Low-calcium fly ash (<8% CaO; typically Class F fly ash) 20 to 30 Moderate-calcium fly ash (8–20% CaO; can be Class F or Class C fly ash) 25 to 35 High-calcium fly ash (>20% CaO; typically Class C fly ash) 40 to 60 Silica Fume 8 to 15 Slag Cement 35 to 65 Metakaolin (calcined kaolin clay) 10 to 20 Table 2. Required levels of SCMs to control ASR (Thomas et al. 2013a).

18 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports the need for alternative sources and new methods of using existing fly ash was identified (AASHTO 2016). 3. Minimize the total alkalis in the concrete mixture. Although many specifications allow the use of potentially reactive aggregates, provided that the cement alkali content does not exceed 0.6% sodium oxide equivalent (Na2Oe, as defined in ASTM C150), some studies have shown that alkali contents between 0.45 and 0.6% may still be sufficient for ASR to occur whereas contents below 0.40% rarely are (AFCESA 2006). This is because the 0.6% limit does not take into account the cement content of the concrete, with higher cement contents contributing more alkalis to the mixture. Thus, it is now recognized that a more accurate index of the risk of expansion is to consider the total alkalis in the concrete mixture (considering both the cement alkalinity and the cement content). The limit on the total alkalis in the concrete can vary with the level of prevention required; for highway pavements, the ASTM C1778/ AASHTO R 80 guide documents suggest a maximum alkali limit of 3.0 lb/yd3 (1.8 kg/m3) Na2Oe for a high level of protection. It is also important to recognize that some SCMs also contain significant quantities of alkali. If these are not soluble, they do not need to be included in the calculation of the concrete alkali content. On the other hand, an SCM with a high soluble alkali content may make ASR worse. For this reason, the prescriptive requirements in the ASTM C1778/AASHTO R 80 guide documents put limits on the alkali content of fly ash, slag cement, and silica fume that are to be used to mitigate ASR for highway pavements. 4. Use lithium-based admixtures. Lithium compounds have been shown to be effective in miti- gating ASR, but their effectiveness is highly dependent on the type of aggregate, the amount of alkali in the concrete, and the form of lithium (Thomas et al. 2007). Thus, it is not possible to prescribe a universal dosage level of lithium to control ASR, with the required levels determined by testing various amounts of lithium with the specific aggregate being consid- ered for use. Lithium nitrate is slightly more effective than other forms of lithium, is safe to handle, and does not exhibit “pessimum” behavior as do lithium hydroxide or lithium carbonate compounds (Malvar et al. 2001). The pessimum effect refers to the greater amounts of expansion being exhibited at lower doses of lithium, instead of less. Federal Aviation Administration For new concrete airfield pavement construction, the FAA has a reactivity testing require- ment in which aggregate and mixture proportion reactivity are tested using modified versions of both ASTM C1260 and ASTM C1567 (FAA 2014): • First, the coarse and fine aggregates are tested separately in accordance with ASTM C1260 test conditions, with the aggregate considered innocuous if the expansion of the test speci- mens does not exceed 0.10% at 28 days. The FAA modified version extends the duration of the test from 14 days of immersion to 28 days of immersion. However, regardless of the result, the FAA still requires that ASTM C1567 be run on the combined materials. • The combined coarse and fine aggregate are tested under ASTM C1567 conditions, with the test modified using the proposed mixture design proportions of aggregates, cementitious materials, and/or specific reactivity reducing chemicals. The FAA modified version extends the duration of the test from 14’ days immersion to 28’ days immersion. If the expansion of the test specimens made with the proposed materials does not exceed 0.10% at 28 days, the proposed combined materials are accepted; if the expansion is greater than 0.10% at 28 days, the aggregates will not be accepted unless adjustments to the combined materials mixture can reduce the expansion. The FAA modification to ASTM C1567 also includes the testing of the proposed mixture proportions. The FAA requires low alkali cements when no other mitigating measures are added. Any fly ash that is used is required to conform to ASTM C618 fly ash with a calcium oxide content of

Literature Review Summary 19 less than 13%, a maximum loss on ignition (a measure of unburned carbon) of 6%, and a total alkali content of less than 3% (FAA 2014). Although calcium content is not currently specified for fly ash classification in ASTM C618, the calcium requirement almost certainly will require a Class F fly ash. It is reported that future versions of ASTM C618 will directly consider calcium oxide content for classification. DOD The construction of concrete pavement for Army, Navy, and Air Force airfields is covered under a United Facilities Guide Specification Section 32 13 11 (DOD 2015). Both fine and coarse aggregate for all concrete are required to be tested for alkali reactivity as follows: • Evaluate the fine and coarse aggregates separately using conditions as dictated in ASTM C1260. If the test results indicate an expansion of greater than 0.08% after 28 days’ immersion, either reject the aggregate or perform additional testing utilizing the proposed low alkali portland cement, blended cement, SCM, or lithium nitrite in combination with each individual aggre- gate. If only SCMs are being evaluated, test under ASTM C1567 conditions, and if only lithium nitrate is being evaluated, test in accordance with the USACE CRD C662. The testing is performed to determine the quantity that lowers the expansion equal to or less than 0.08% at 28 days’ immersion. • If none of the above options lowers the expansion to less than 0.08% after 28 days, the aggregate should be rejected. The use of low alkali cement is also required for all mixtures, and acceptable pozzolans include Class F fly ash (Class C is not permitted for paving concrete), Class N natural pozzolan, ultra- fine fly ash (UFFA) and ultra-fine pozzolan (UFP), slag cement, and silica fume (DOD 2015). Recommended ranges for the contents of these various SCMs are provided in Table 3. ASTM and AASHTO Guide Documents An approach separate from that used by the FAA and DOD is available from ASTM and AASHTO. ASTM first published ASTM C1778, Standard Guide for Reducing the Risk of Deleteri- ous Alkali-Aggregate Reaction in Concrete, in 2014, with the most recent version being approved in 2016. ASTM C1778 is very similar to AASHTO R 80, Practice for Determining the Reactivity of Concrete Aggregates and Selecting Appropriate Measures for Preventing Deleterious Expansion in New Concrete Construction, which came out in 2017, but was first published as a provisional practice (AASHTO PP 65) in 2011. These documents provide guidance to highway agencies for the determination of the reactivity of concrete aggregates and in the selection of preventive Supplementary Cementitious Material Minimum Content (percentage) Maximum Content (percentage) Class N Pozzolan and Class F Fly Ash SiO2 + AL2O3 + Fe2O3 > 70 percent 25 35 SiO2 + AL2O3 + Fe2O3 > 80 percent 20 35 SiO2 + AL2O3 + Fe2O3 > 90 percent 15 35 UFFA and UFP 7 16 Slag Cement 40 50 [Silica Fume]* [7] [10] *Silica fume must only be used for projects outside of the continental U.S. (OCONUS) where Class F fly ash and slag cement are not available and when approved. Table 3. Recommended ranges of SCM contents (DOD 2015).

20 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports measures and are based largely on the significant work performed by FHWA (Thomas et al. 2006; Thomas et al. 2008; Thomas et al. 2012). Based on ASTM C1778 (AASHTO R 80), the potential for aggregate reactivity is judged by: • Consideration of field performance, taking into account any differences in materials and mixture design that may have occurred. • Petrographic analysis of the aggregate (ASTM C295) to determine the presence of potentially reactive minerals. • Mortar bar expansion (ASTM C1260/AASHTO T303) not greater than 0.10% after 14 days’ immersion in 1 M NaOH solution at 176 °F (80 °C), and expansion of concrete prisms (ASTM C1293) of not greater than 0.040% at 1 year. Note that a test period of 2 years is required for ASTM C1293 if the effectiveness of SCMs is being investigated. If the aggregate is identified as being potentially ASR reactive, ASTM C1778 requires that it be rejected or used with appropriate preventive measures using either a prescriptive approach or a performance approach. The prescriptive approach in ASTM C1778 considers the reactivity of the aggregate, the type and size of the structure, the exposure conditions, and the composition of materials being used (Thomas et al. 2012). It uses the accelerated mortar bar test (ASTM C1260/ AASHTO T 303) to classify the reactivity of the aggregate from non-reactive to very highly reactive, and then from those results a level of ASR risk is defined considering the exposure conditions. The required level of prevention is determined by considering the level of ASR risk and the classification of the structure based on the consequences of having an ASR problem. Pavements are commonly considered as being Class SC3, meaning that the consequence of major ASR damage can cause significant safety, economic, or environmental consequences, yet minor risk of ASR is acceptable. It could be argued that certain critical airfield pavements, especially those at military bases, should be classified as SC4, meaning that there are serious safety, economic, or environmental consequences if there is minor damage and thus ASR cannot be tolerated. The final step within the prescriptive approach to mitigation in the ASTM C1778 is to identify preventive measures based on the level of prevention required, with options including using SCMs, limiting the alkali content of the cementitious materials, or a combination of the two. The performance approach in ASTM C1778 recommends that the concrete prism test (ASTM C1293) be used to evaluate the efficacy of ASR mitigation strategies such as SCMs or lithium compounds (Thomas et al. 2012). As the test will require 2 years to conduct (requiring that expansion of concrete not exceed 0.040% at 2 years), it is recommended that a range of SCM or lithium types and dosages be tested to ensure an effective strategy can be determined. A method is also provided for the accelerated mortar bar test (ASTM C1567/AASHTO T303) to determine SCM dosage once the longer duration concrete prism test has been conducted for the specific aggregates under consideration. Although not included in ASTM C1778, AASHTO R 80 provides a method to assess the effectiveness of lithium nitrate for mitigating ASR based on Thomas et al. (2008). Common Laboratory Tests As conveyed in the preceding discussions, some laboratory tests have been available for the assessment and evaluation of potential ASR in aggregates and in mixtures. A summary of some of the more prominent tests is presented in Table 4. Current guidance on preventing ASR in new concrete construction generally focuses on the use of ASTM C1260, ASTM C1567, and ASTM C1293, with additional details on these tests provided below (Van Dam 2016a,b): • ASTM C1260, often referred to as the accelerated mortar bar test, provides results in a relatively short time of 16 days. This makes it useful as not only a screening test, but also a

Literature Review Summary 21 Test Method Comments ASTM C227, Standard Test Method for Potential Alkali Reactivity of Cement- Aggregate Combinations (Mortar Bar Method) • Mortar bar (2.25 parts aggregate to 1 part cement) test intended to study cement-aggregate combinations. • Several reported problems with test, including excessive leaching of alkalis from specimens. ASTM C289, Standard Test Method for Potential Alkali-Silica Reactivity of Aggregates (Chemical Method) • Chemical test in which crushed aggregate is immersed in 1M NaOH solution for 24 hours—solution is then analyzed for amount of dissolved silica and alkalinity. • Poor reliability; problems with test include: – Other phases present in aggregate may affect dissolution of silica. – Test is overly severe, and aggregates with good field performance often fail the test. – Some reactive phases may be lost during pretest processing. ASTM C295, Standard Guide for Petrographic Examination of Aggregates for Concrete • Useful evaluation to identify many (but not all) potentially reactive components in aggregates. • Reliability of examination depends on experience and skill of individual petrographer. • Results should not be used exclusively to accept or reject aggregate source—findings best used in conjunction with other laboratory tests (e.g., ASTM C1260 and/or ASTM C1293). ASTM C441, Standard Test Method for Effectiveness of Mineral Admixtures or Ground Blast-Furnace Slag in Preventing Excessive Expansion of Concrete Due to Alkali-Silica Reaction • Mortar bar test, intended to assess effectiveness of SCMs in reducing ASR expansion. • Test uses high-alkali cement and Pyrex glass. • Test not very reliable because of the use of Pyrex glass, which is sensitive to test conditions and contains alkalis that may be released during the test. Test does not correlate well with data from concrete mixtures containing natural aggregates. ASTM C856, Practice for Petrographic Analysis of Hardened Concrete • Useful for analyzing concrete (from laboratory or field) and for identifying presence of reactive aggregates or reaction products. • Reliability of examination depends on experience and skill of individual petrographer. • Essential for relating aggregate reactivity to field performance. ASTM C1260, Standard Test Method for Potential Alkali Reactivity of Aggregates (Mortar-Bar Method) (AASHTO T303)* • Mortar bar test, with specimen submerged in a 1 N sodium hydroxide solution at 176oF (80ºC). • 16-day test originally designed to assess aggregate reactivity. • Accelerated test suitable as screening test, but because of the severity of the test, it should not be used by itself to reject a given aggregate. • If aggregate is tested using both ASTM C1260 and ASTM C1293, the results of ASTM C1293 should govern. ASTM C1293, Standard Test Method for Concrete Aggregates by Determination of Length Change of Concrete Due to Alkali-Silica Reaction • Concrete prism test, generally regarded as best indicator of field performance. • Developed as an aggregate test (using non-reactive fine aggregate to test reactivity of coarse aggregate, and vice versa). • Test requires 1 year for completion. • Also can be used to test effectiveness of SCMs and lithium compounds, but test is then typically run for 2 years. • Widely accepted test method; however, long duration of test is major drawback. ASTM C1567, Standard Test Method for Determining the Potential Alkali- Silica Reactivity of Combinations of Cementitious Materials and Aggregate (Accelerated Mortar Bar Method)* • Mortar bar test, essentially identical to ASTM C1260, but used to investigate the potential for deleterious ASR in mortar mixtures containing SCMs. • Should only be used for aggregates for which a reasonable correlation between ASTM C1260/AASHTO T303 and ASTM C1293 has been established. • ASTM C1567 can be modified as per AASHTO R80 to evaluate lithium nitrate-based admixtures (although ASTM C1293 is deemed to be a more effective test method). *Note that the FAA and DOD use modified versions of these standardized tests, increasing immersion from 14 days to 28 days as well as some other subtle differences, including testing of aggregate combined in mixture proportions in the FAA and DOD variants of ASTM C1567. Table 4. Test methods for assessing ASR (based on Thomas et al. 2013a).

22 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports test that can be conducted during the mixture design process or even during construction if one of the relevant concrete constituents changes. The test consists of making mortar beams containing the aggregate of interest (either fine aggregate or crushed coarse aggregate) and after 2 days, soaking them in a 176 °F (80 °C) 1N NaOH solution for 14 days. Length change measurements are made periodically and the total expansion after the 14 days of soaking is typically used as the criteria to classify the aggregates as potentially reactive or not. The FAA and DOD variants of this test extend the time of immersion from 14 days to 28 days. Different agencies have different criteria on the expansion limit. In ASTM C1778, the limit is 0.10% at 14 days’ immersion for an aggregate to be considered non-reactive (R0). An important caveat in ASTM C1778 is that the results of the ASTM C1293 are considered to be a better predictor of aggregate reactivity than those from ASTM C1260. • ASTM C1567 is conducted under the same testing conditions as ASTM C1260, with the dif- ference being that the efficacy of SCMs as mitigation strategies are investigated. The same sample preparation, soak solution, temperature, and duration of immersion (14 days) are used. Both the FAA and DOD variants again extend the time of immersion from 14 days to 28 days. In addition, the FAA requires that the combined gradation be tested together instead of testing each aggregate size separately. AASHTO R 80 has an ASTM C1567 variant for testing the effectiveness of lithium nitrate. • ASTM C1293 is commonly referred to as the concrete prism test (CPT) as it employs con- crete prisms made with the aggregates under evaluation. In the CPT, a standard concrete mixture with an alkali loading of 1.25% by mass of cement (equivalent to an alkali loading of 8.85 lbs/yd3 [5.25 kg/m3]) is made and cast into prisms. After an initial 24-hour curing, the concrete prisms are stored over water at 100 °F (38 °C), typically for 1 year when screen- ing aggregate for use in concrete containing only pure portland cement. The expansion limit is 0.04%. The test duration is extended to 2 years when evaluating cementitious systems containing SCMs. The CPT is considered the best available test for assessing the potential field performance of aggregates (Thomas et al. 2013a). The major limitation of ASTM C1293 is the duration of testing (1 to 2 years), which is feasible for aggregate source screening, but makes it highly impractical for project-specific evaluation. Another problem is that alkalis are known to leach from the concrete prisms during testing, an issue partially addressed by increasing the initial alkali loading beyond what would normally occur to compensate for the loss in alkalis over time. But this approach only partially addresses the issue because alkali leaching can have a profound effect in practice, and thus it is not recommended that ASTM C1293 be used to establish the alkali threshold for an aggregate source or aggregate-binder combination (Thomas et al. 2013a). With any laboratory testing regime or protocol, there is some risk of accepting an aggregate source that may actually be reactive (false negative) or of rejecting an aggregate source that may actually be acceptable (false positive). Both are undesirable, but false negatives can carry signi- ficant performance and long-term cost impacts. Thus, it is desired that appropriate tests and limits be selected to help minimize these false readings. Treatment of ASR in Existing Concrete Pavements To date, no definitive methods have been identified for addressing the ASR mechanism in existing concrete pavements, although some different measures have been tried. For example, some agencies have used conventional methods to reduce the amount of moisture in the con- crete pavement, but these generally show only marginal benefits, if any. Similarly, most agencies have experience with measures used to correct deficiencies associated with the ASR reaction (e.g., spalling, cracking), but these are generally viewed as interim measures that provide only

Literature Review Summary 23 temporary relief. More recently, topical surface treatments to reduce moisture ingress or to slow the ASR reaction have been used by a few agencies, but their long-term effectiveness is not known. Ultimately, the use of structural overlays or reconstruction can also be consid- ered treatment options, depending on the severity of the distress. The common ASR treatment (including preventive, corrective, and mitigative) strategies that have been used on existing concrete pavements are described in this section. Joint and Crack Sealing The sealing of joints and cracks in an existing pavement structure (see Figure 5) may reduce the rate of moisture infiltration in the short term, but the long-term effectiveness of those materials can vary depending on some factors, including the climate, the joint design, the sealant material itself, and the progression of the ASR distress, among others. Typical performance of a joint or crack sealant material is about 3 to 10 years. Moreover, it must be recognized that moisture enters into a pavement structure from other locations and not just the surface. Thus, at best, joint and cracking sealing may only be a short-term solution to minimizing water infiltration, but sufficient moisture is likely available from beneath the slab to support ASR development. Joint and crack sealing does nothing directly to address any deterioration that has already occurred. Retrofitted Subsurface Drainage The addition of subsurface drainage to an existing pavement is intended to remove moisture from under the pavement and at joints and cracks, which may assist in slowing or delaying ASR development. However, the effectiveness of retrofitted subsurface drainage is largely dependent on the characteristics of the base materials, which may limit their effectiveness (Smith et al. 2014). Moreover, on large expanses of pavement (e.g., on airport runways, taxiways, and aprons), the ability for retrofitted drainage systems to provide effective drainage over such large areas is questionable, and it is uncertain whether sufficient amounts of moisture can be removed to reduce humidity levels within the concrete to below the critical 80% level. Finally, Figure 5. Sealed crack and joint on ASR-affected pavement (courtesy of Applied Pavement Technology, Inc.).

24 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports the addition of the edge drains does nothing to address the deterioration that has already occurred in the pavement, which would require some corrective repairs or treatments. If an existing pavement already has a drainage system, it is important that the system be pro- actively maintained to ensure its overall effectiveness and functionality. As a minimum, this should include periodic inspections (both of the outlets and of the internal drainage system itself) and may also require regular cleaning and flushing of the system (Smith et al. 2014). Surface Treatments/Sealers A variety of topical treatments may be applied to the surface of concrete pavements in efforts to mitigate or delay the development of ASR. These are described in the following sections. Surface Sealers These are materials such as silanes and siloxanes that are applied to the pavement surface to reduce or prevent the ingress of moisture (as well as other substances, e.g., deicing chemicals) into the concrete. These materials are applied topically and penetrate the concrete pores to some degree and react with the concrete to provide water repellency (Sutter et al. 2008). Silanes and siloxanes are “breathable” sealer materials, in that they allow water vapor out of the concrete while minimizing the ingress of liquid water from the surface. However, the longevity of these sealers is often short, requiring the repeated application of the materials at regular intervals (Godart and de Rooij 2017). Non-breathable sealers, such as high-molecular weight methacrylate, do not permit water vapor to exit the concrete and thus can increase the relative humidity within the pavement concrete, especially near the surface. Silanes materials are the products most commonly evaluated with regard to addressing ASR in existing concrete pavements (Thomas et al. 2013a). These materials may be either water based or solvent based and can be produced in varying concentration levels (from 20% up to 100%) (Thomas et al. 2013a), with higher concentration levels likely more effective than lower concen- tration levels. Silanes may be applied at a rate of 100 to 200 ft2/gal (2.5 to 4.9 m2/L), with depths of penetration in the range of 0.08 to 0.25 inches (2 to 6 mm) (Thomas et al. 2013a). Silanes have been shown to be effective at reducing cracking and expansion on concrete high- way barriers (Thomas et al. 2013a; Folliard et al. 2016; Deschenes et al. 2017), but these are relatively thin structures with minimal surface area in contact with the grade, and the silane could be applied on both sides. An Arkansas study demonstrated that silane was more effective than either linseed oil and elastomeric paint in not only reducing the internal relative humidity of highway barriers, but also in reducing expansion over three damage levels (Deschenes et al. 2017). However, when applied on several spans of a reinforced concrete bridge in Alabama, silanes were considered ineffective in reducing humidity levels, cracking, and expansion (Johnson et al. 2013). In 2006, several different surface treatments were applied to an ASR-afflicted concrete apron at Riverton Regional Airport in Riverton, WY (Basham 2009). Five different products were included in the study: sodium tartarate (a low-viscosity, water-based product that seals concrete against water), lithium nitrate, siloxane, silane, and boiled linseed oil. Pre- and post-application cores were retrieved and evaluated using the Damage Rating Index and tested using pulse velocity to characterize the properties of the concrete. Although intended to be monitored over a 5-year period, the study was terminated after 2 years when the pavement started to exhibit severe FOD and had to be reconstructed. The results from the truncated study suggested that lithium nitrate, sodium tartarate, and siloxane may reduce the rate of ASR deterioration, but this could not be stated conclusively (Basham 2009).

Literature Review Summary 25 A study on the use of silanes on an ASR-afflicted concrete pavement was performed in 2012 on a 12-mile (19-km) stretch of I-530 near Pine Bluff, AR (Thomas et al. 2013c). Two sections of pavement, each approximately 1800-ft (550-m) long and of varying ASR severity, were studied and included a control segment and two test segments at two silane concentration levels (40 and 100%) (Thomas et al. 2013c). No data are available regarding the performance of these sections, but it is noted that a contract award was made in the summer of 2017 to reconstruct this portion of I-530 (Arkansas Democrat-Gazette 2017). The effects of silanes and other surface sealants to mitigate ASR in existing concrete are being studied as part of the runway reconstruction effort performed at Northwest Arkansas Regional Airport (Heymsfield et al. 2016). It was recognized that ASR existed in other pavements at the Airport, so remediation measures in the form of three surface sealers were investigated: elastomeric coatings, linseed oil, and silane (Heymsfield et al. 2016). An outdoor exposure site was constructed at the University of Arkansas Engineering Research Center and featured the evaluation of the sealers on specimens retrieved from taxiway pavements at the airport. Based on a 1-year evaluation period, it was noted that the linseed oil and silane coatings showed slight reductions in the amount of expansion, although it was noted that the strain differences were low and a longer evaluation period would be needed for more definitive results (Heymsfield et al. 2016). In 2010, the Nebraska Department of Roads initiated a field evaluation of seven silane/siloxane sealers on an 8-year old concrete pavement to determine their ability to delay ASR progression (Heyen et al. 2015). Based on the results of the field testing and accompanying laboratory testing, the research indicated (Heyen et al. 2015): • 100% and 40% silane solvent-based sealers exhibited consistently high performance through- out all the tests and offered the best protection against chloride ion intrusion. • 40% water-based silane and blend of lithium and silane/siloxane exhibited medium performance. • Waterborne silane-siloxane based and the acrylic-based polymer exhibited lowest performance. During the evaluation, it was noted that the solvent-based sealers were found to have greater visible/measured depths of penetration than the water-based sealers, and it was suggested that reapplication of the sealers would be needed every 4 years (Heyen et al. 2015). A laboratory evaluation of topical treatments was conducted at the University of Arkansas to assess their impacts in reducing expansion (Waidner 2016). After about 15 months, the silane materials were noted to be most effective in reducing the amount of expansion and worked better than both linseed oil and lithium. Lithium Compounds Lithium compounds, which change the nature and behavior of the alkali-silica gel from expansive to non-expansive, have been used in an attempt to reduce ASR expansion in pavements already suffering from ASR (Thomas et al. 2013a). Laboratory research on the use of lithium compounds on ASR-infected mortar or concrete specimens has shown a marked reduction in expansion levels, but limited field trials have been conducted on their use on existing concrete pavements (Thomas et al. 2013a). Both lithium nitrate and lithium hydroxide have been used in field applications, but lithium nitrate is more commonly used, typically at a 30% concentration level and at rates between 167 and 333 ft2/gal (4.1 and 8.2 m2/L) (Thomas et al. 2007). Rigorous monitoring of the depth of penetration has not been performed, but limited data suggest that penetration depths of only a few millimeters are achieved (Thomas et al. 2013a).

26 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports The depth of penetration of the lithium is an ongoing concern, as it must penetrate deep enough into the concrete to affect the ongoing reactivity within the mass of the slab. The pres- ence of cracking in the concrete will help facilitate the penetration of the lithium, but at the same time extensive cracking and deterioration may suggest that it is too late to treat the affected concrete (Thomas et al. 2013c). Furthermore, cracks at the surface are often open, but at depth are filled with the ASR reaction product, thus impeding penetration of the lithium compound. In an attempt to increase the depth of penetration, some field trials have been conducted using electrochemical methods or vacuum impregnation. Electrochemical methods apply an electrical field to help drive the lithium ions into the concrete, whereas vacuum impregnation uses negative pressure to help achieve penetration into the deteriorated system (Thomas et al. 2013c). On bridge columns, preliminary results from field studies indicate that vacuum impreg- nation was not effective, but better results were achieved using electrochemical methods in which the lithium nitrate was driven to a depth of 2 inches (50 mm) with dosages sufficient to reduce ASR (Thomas et al. 2013a). However, electrochemical methods require the presence of embedded steel to establish the electrical field and would not be possible on non-reinforced concrete pavements. With regard to the electrochemical method, when it was applied to structures, the depth of penetration was observed to be greater when the temperature of the electrolyte solution was increased from 86 oF to 104 oF (30 oC to 40 oC) (Ueda et al. 2016). Lithium nitrate was also observed to be more effective than other lithium salts (Ueda et al. 2016). Several pilot projects featuring the use of lithium compounds were evaluated under the SHRP (Krauss et al. 2006). One project in Nevada was treated with lithium hydroxide solution applied at a rate of 60 ft2/gal (1.5 m2/L). After 7 years of monitoring, no difference in perfor- mance was observed and the ASR continued to progress on both sections (Krauss et al. 2006). A separate project in Delaware also featured the surface application of lithium hydroxide to a pavement surface and, it was determined to have had no effect in stopping or slowing the rate of ASR deterioration (Krauss et al. 2006). A series of field trials evaluating the topical application of lithium compounds on concrete pavements was performed in the late 1990s and early 2000s (Folliard et al. 2006). A general summary of these installations is provided in Table 5. Limited follow-up data are available on these projects, but investigations on the I-84 project in Idaho indicated that the lithium penetrated only a few millimeters (Thomas et al. 2013a). A 3-year study of the effectiveness of a topical application of lithium nitrate on an ASR- distressed concrete pavement in Norfolk, NE, revealed no reduction in ASR progression (Kelly and Tuan 2006). The observed depth of penetration and lithium concentrations were limited, although pressure installation techniques demonstrated promising results on a small scale (Kelly and Tuan 2006). The use of lithium on a second pavement project in Delaware was evaluated under the FHWA’s ASR Development and Deployment Program (Thomas et al. 2013c). In 2009, a 30% solution of lithium nitrate was applied topically to 16 lane-miles (26 lane-km) of ASR-affected concrete pavement along US 113 near Georgetown, DE. Cores of the pavement revealed that significant lithium concentrations were found only in the upper 0.25 to 0.5 inches (6 to 13 mm) of the pavement, and it was generally concluded that the topical application of lithium nitrate is not an effective ASR mitigation technique (Thomas et al. 2013c). In the mid-2000s, a study on the effect of lithium compounds applied on three concrete airfield pavements exhibiting ASR was performed by the Innovative Pavement Research Foun- dation (Whitmore 2009). The study included airfield pavement facilities at Cheyenne Regional

Literature Review Summary 27 Airport, Phoenix Sky Harbor Airport, and Hartsfield-Jackson Atlanta International Airport, and featured the use of lithium applied at a rate of 220 ft2/gal (5.4 m2/L) (applied in two separate 440 ft2/gal [10.8 m2/L] doses) (Whitmore 2009). An analysis showed that the depth of penetra- tion of the lithium compounds was no more than 0.70 inches (18 mm), and overall no changes in pavement condition were observed during the short evaluation period. It was recommended that additional long-term monitoring be performed to see if any differences emerged between the control and treated sections. High-Molecular Weight Methacrylate (HMWM) HMWM is a product that is intended to strengthen an existing concrete pavement affected by ASR by filling the cracks and bonding the cracked concrete pieces together. HMWM is applied topically by brush, squeegee, or spray unit, and sand or other small aggregate can be added to help ensure surface friction (Van Dam et al. 2002a). HMWM been used on pavements with more advanced stages of ASR. In the 1990s, two concrete pavement projects affected by ASR, one in Nevada and one in California, were the subject of a study on the use of surface treatments, including HMWM (Stark et al. 1993). At the Nevada site, significant reductions in mid-panel and joint deflections were realized after the application of HMWM, while on the California project it was found that HMWM penetrated cracks up to 2 inches (50 mm) deep, which corresponded to the maximum depth of the surface cracks (Stark et al. 1993). However, a later follow-up study on these projects revealed that HMWM had no effect on the long-term performance of the Nevada project, but it was noted that the pavement may have been too severely deteriorated at the time of application (Krauss et al. 2006). The follow-up on the California project indicated that the methacrylate reduced spalling and extended the life of the pavement by about 3 to 5 years (Krauss et al. 2006). Project Date of First Application Lithium Compound Application Rate, L/m2 Number of Applications Pavement Condition Prior to Treatment Comments on Effectiveness Rt. 14, Wolsey, SD 1995 Lithium nitrate 18.36 36.72 55.08 1 Fair Little visible difference (control sections and treated areas unchanged during evaluation period) Rt. 1, Bear, DE 1998 Lithium nitrate 24.48 6 (2 per year for 3 years) Good to fair On southern end of project, treated areas faring better than control— no noticeable difference on northern end Rt. 15, New Ulm, MN 1998 Unknown 24.48 1 & 2 Fair to poor Second treatment done on portions; no noticeable improvement (controls and treated areas both continued to deteriorate about the same) I-85, High Point, NC 1999 Unknown 12.24 1 Fair Not evaluated—used as enhancement of durability of asphalt overlay I-84, Mountain Home, ID 2004 Lithium nitrate 24.48 1, 2, or 3 Fair to poor Too early to determine; ongoing monitoring includes expansion measurements and crack mapping Table 5. Summary of lithium applications to existing ASR-affected concrete pavements (after Folliard et al. 2006).

28 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports HMWM was also used on a pavement facility at Seymour-Johnson Airforce Base (AFB) in an attempt to strengthen the concrete and reduce spalling and FOD. After 8 years there was no noticeable difference in the condition of the treated and untreated sections, and consequently HMWM is not recommended for Air Force use (AFCESA 2006). Interim Corrective Repairs and Treatments Several corrective repairs and treatments are commonly used on concrete pavements afflicted with ASR, not to address the ASR mechanism itself, but rather to address the distresses and symptoms associated with the deleterious reaction (AFCESA 2006). Partial-Depth Repair (PDR) PDR is one rehabilitation method that can be used to repair localized concrete deterio- ration caused by ASR (see Figure 6). These repairs consist of the removal of concrete near the surface and replacement with an acceptable patching material, often a rapid-setting cementitious or proprietary material to reduce closure times. Partial-depth repairs are most commonly performed along transverse and longitudinal joints where ASR-induced spalling is most prevalent, and they are most effective when the distress is limited to the upper one-third to one-half of the concrete slab thickness (Smith et al. 2014). In some cases, it may be chal- lenging to identify areas of sound concrete, in which case a full-depth repair may be required (AFCESA 2006). Although capable of restoring serviceability, PDRs are not always an ideal treatment because the progressive nature of the ASR distress continues to develop beyond the boundaries of the patch (Van Dam et al. 2002a). Moreover, they may not be effective if the deterioration goes beyond about one-third to one-half of the slab thickness, or if the extent of severe deteriora- tion encompasses entire slabs; in those cases, full-depth repairs or slab replacements may be more appropriate. Nevertheless, partial-depth repairs may serve as effective short-term solutions until more substantial rehabilitation can be programmed, but their cost-effectiveness should be evaluated in comparison with other interim measures. In general, the performance of partial-depth repairs has been mixed, and it is largely recognized that their proper construction and installation is critical in order to achieve good performance (DOD 2001; Smith et al. 2014). Effective bonding between the repair material and the concrete Figure 6. Partial-depth repair on ASR pavement (courtesy of Applied Pavement Technology, Inc.).

Literature Review Summary 29 substrate is critical, and the presence and movement of joints must be accommodated. Expedient methods featuring cold milling and rapid-setting materials are available to limit closure times for partial-depth repair work (Hammons and Saeed 2010). Achieving good bonding between the existing concrete and patching material is essential, yet may be difficult depending on the presence and quantity of ASR gel in the pavement. For example, at Holloman AFB in New Mexico, ASR was present throughout the upper region of the concrete and appeared as a white haze on the repair surface that proved impossible to remove by washing, brushing, waterblasting, and sandblasting (AFCESA 2006). Failure rates of the partial-depth repairs exceeded 80%, but better performance was achieved when the repairs were placed deeper in the slab (minimum 6-inch [150 mm] depth) (AFCESA 2006). Full-Depth Repair (FDR)/Slab Replacement FDR and slab replacement (see Figure 7) are commonly used to address ASR distresses in concrete pavement (particularly those that are extensive throughout the expanse and depth of a slab). These repairs consist of the removal of isolated, deteriorated areas through the entire thickness of the existing slab and the replacement with a concrete repair material selected to meet the opening time and performance requirements. However, as with partial-depth repairs, full-depth repairs should be viewed not as a solution to the ASR issue, but rather as a means to extend the life of the pavement until more substantial rehabilitation can be performed; this is because continued deterioration is likely to occur in the original concrete adjacent to the concrete repair (Van Dam et al. 2002a). Furthermore, the installation of these repairs may not be cost-effective if a significant amount of patching or slab replacement is required. Full-depth repairs and slab replacements have an excellent track record of performance on concrete pavements. Their performance is largely driven by proper boundary selection, suitable foundation preparation and support, and effective load transfer provisions (DOD 2001; Smith et al. 2014). The use of precast repairs is emerging as a viable strategy for facilities requiring more rapid opening times. Pressure Relief Joints (PRJs) PRJs are joints installed after initial construction to relieve expansive stresses in a pavement that otherwise could lead to joint spalling, slab buckling, blowups, or pushing/shoving of in-pavement fixtures and adjacent structures (Smith et al. 1987). Pressure relief joints are Figure 7. Slab replacement on ASR pavement (courtesy of Applied Pavement Technology, Inc.).

30 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports typically 1 to 2 inches (25 to 50 mm) wide and are installed by making full-depth sawcuts across the full width of the pavement, removing the concrete, and inserting a compressible filler. The severe impacts of expansion have been noted on some military installations. At Andrews AFB, the expansive forces were severe enough to cause an entire lane of parking ramp slabs to fail and tent-up in one morning (AFCESA 2006). Cases of differential movement between adjacent slabs have also been documented (see Figure 8), as have instances where the slabs actually bowed up because of the formidable expansive forces (AFCESA 2006). Although the installation of pressure relief joints can provide temporary relief from expansive forces in the pavement, PRJs do not address the causes of the expansion. In many cases, the relief joints close very quickly, creating the need for additional joints. At Seymour-Johnson Air Force Base, 1.5-inch (38-mm) wide relief joints closed in about 2 years while 4-inch (100-mm) wide relief joints closed to 1 inch after 5 to 10 years (AFCESA 2006). The widespread use of pressure relief joints on ASR-affected highway concrete pavements in Nebraska resulted in significant movement of the slabs, which opened up joints and cracks and thereby reduced aggregate interlock load transfer (Smith et al. 1987; Hoerner et al. 2001). Structural Overlay Solutions A structural overlay, either asphalt or concrete, can be used to address ASR on existing con- crete pavements. These can have variable thicknesses, but are commonly designed to meet the future traffic loadings while accounting for the current level of deterioration in the existing pavement. Although they do not directly address the ASR issue, they can effectively restore the serviceability of the facility and eliminate FOD potential. Still, their effective service life may see significant reductions depending on the extent and severity of the ASR in the existing pavement. Of general concern is that overlays may not be an appropriate treatment in pavements under- going significant ASR-related expansion as evidenced by blowups. The potential for continued expansion after the application of an overlay should be considered. Asphalt Concrete (AC) Overlay AC overlays can be placed rapidly and in convenient construction stages, allowing significant flexibility in maintaining traffic operations. However, the performance of asphalt overlays is Figure 8. Offset of pavement joints on ASR-affected pavement (courtesy of Applied Pavement Technology, Inc.).

Literature Review Summary 31 strongly dependent on the underlying pavement conditions, and subsequently may require a significant amount of pre-overlay repair. In severe cases of ASR, rubblization of the existing concrete pavement may need to be considered to disrupt the existing slab to eliminate damage due to future expansion, but this often drives up the required thickness of the overlay. Addition- ally, concerns have been expressed that concrete pavement afflicted with ASR could potentially expand after rubblization, resulting in differential swelling in the new asphalt overlay (AFCESA 2007). A study on the use of rubblization on airfield pavements found no evidence to support that contention and documented a long history of successful use of rubblization on ASR- afflicted concrete pavements on both airfield and highway facilities (Buncher et al. 2008). At Gimpo Airport in South Korea, portions of an ASR-afflicted parallel taxiway were milled and overlaid in 2009 with various types of asphalt overlay solutions, including a conventional HMA overlay, a crumb rubber modified overlay, an HMA overlay with a pavement fiber and waterproofing interlayer, and an HMA overlay with a glass fiber grid (Kwak et al. 2014). In each case, the top 4 inches of concrete were removed by milling and replaced with the same thickness of the various asphalt mixtures. After 3 years of monitoring, potholes had formed and were affecting performance, requiring the removal and replacement of these overlays with a 4-inch (100-mm) stone matrix asphalt overlay that was expected to provide better performance (Kwak et al. 2014). Unbonded Concrete Overlays Unbonded concrete overlays can be effective rehabilitation treatments for existing pave- ments affected by MRD because their performance is less dependent on the condition of the underlying pavement. Moreover, because they eliminate the need for pavement breakup, removal, disposal, and reworking of the foundation materials, they are an attractive alternative to complete reconstruction of the pavement facility. However, their generally greater thicknesses could create issues with sideslopes and matching elevations at intersecting pavement facilities. In addition, if the ASR has resulted in severe expansion as evidenced by blowups, consideration should be given to the potential impacts of future expansion. Bonded Concrete Overlays Bonded concrete overlays are generally not recommended for use on concrete pavements with ASR because of the degradation in the existing pavement and the difficulty in achieving good bond. That being said, a bonded concrete overlay was applied on an ASR pavement at Pease Air Force Base in the 1990s (AFCESA 2006). The surface of selected slabs was milled to remove the top layer of deterioration and was followed with a 3-inch (76-mm) bonded concrete overlay. The purpose was only to correct surface FOD problems, and underlying longitudinal cracks from the underlying pavement were observed to reflect through very rapidly (AFCESA 2006). Reconstruction The most effective method of addressing ASR is through the total reconstruction of the existing pavement. This alternative is most appropriate on severely deteriorated pavements in which ongoing maintenance activities are recurrent and costly. In the new reconstructed pavement, it is imperative that a durable concrete mixture be developed in accordance with prescribed specifications in order to prevent the recurrence of the ASR. When the existing pavement is reconstructed, the material from the old pavement can be removed, crushed, and recycled for use in some different construction applications, such as fill, granular base, or subbase. The use of recycled concrete aggregate (RCA) can help reduce costs and environmental impacts associated with the reconstruction, but it must be done in accor- dance with prevailing specifications. For example, the FAA allows the use of recycled concrete

32 Practices to Mitigate Alkali-Silica Reaction (ASR) Affected Pavements at Airports aggregate as a base course under Item P-219, but notes that “concrete that has deteriorated from alkali-silica reaction (ASR) may be used for recycling base course with appropriate analysis” (FAA 2014). However, AFCESA (2006) states that “recycled concrete from concrete undergoing ASR reactions should not be recycled in any application within the airfield pavement structure, e.g., use as aggregate in PCC mixtures, as base course aggregate, or in drainage layers.” An IPRF study prepared the recommendations shown in Table 6 for the use of RCA produced from ASR-distressed concrete. Other Treatments Several other treatments have been used to address ASR in existing concrete structures, but they are not particularly relevant to pavements. These treatments include: • Application of external restraint. Physical restraint or confinement has shown some success in reducing the deleterious expansion due to ASR in the direction of restraint. This has been used on structures or concrete columns, but is generally restricted to relatively small masses of structural concrete (Thomas et al. 2013a). The major shortcoming with the use of restraint in actual pavements is that it is extremely expensive and that, in order to be most effective, the restraint must be applied early in the life of the pavement before the extent of potential damage has been observed (Van Dam et al. 2002a). • Drying the concrete. Drying refers to the removal of moisture from the concrete to a level below which expansive ASR forces can develop (typically taken as 80% humidity). However, the ability to sufficiently dry pavements (or other concrete structures in contact with grade) in the field is limited (Van Dam et al. 2002a). Summary This chapter presents an overview of ASR and its impact on concrete pavements. The over- all mechanisms of the distress are described and are noted to require three factors for ASR development: (1) a sufficient concentration of alkali hydroxides in the pore solution of the concrete; (2) a sufficient quantity of unstable silica in the aggregate; and (3) a sufficient supply of moisture in the concrete. Some different methods for identification of ASR are reviewed, but many of these are based solely on visible indicators, and it is emphasized that only through a petrographic analysis (conducted in accordance with ASTM C856) can the presence of ASR be confirmed. Construction Application Aggressive ASR, Primary Pavement Aggressive ASR, Secondary Pavement Aggressive ASR, Tertiary Pavement Mild ASR, Primary Pavement Mild ASR, Secondary Pavement Mild ASR, Tertiary Pavement Deep Fill • • • • • • Sub-Base --1 • • • • • Base --1 --1 • • • • Drainage Layer --1 --1 • • • • • = Recommended Application 1 Detailed benefit/risk analysis must be conducted Primary Pavement = Essential pavements (runway, parallel taxiway, main apron) Secondary Pavement = Occasional use pavements—engines running (ladder taxiways, hold apron) Tertiary Pavement = Other airfield pavements Table 6. Application and risk matrix for RCA from ASR concrete (after Saeed et al. 2006).

Literature Review Summary 33 Mitigation methods for new concrete mixtures are described and noted to include four general approaches: 1. Avoid the use of susceptible aggregates. 2. Use supplementary cementitious materials. 3. Minimize the total alkalis in the concrete mixture (and do not just focus on the alkali content of the cement). 4. Use lithium-based admixtures. Recent years have seen FAA and DOD specifications evolve with regard to ASR mitigation, and airport authorities and highway agencies alike adopt all or portions of these strategies to help prevent the development of ASR in new concrete pavement construction. Although significant progress has been made in eliminating the potential for ASR in new construction, no definitive methods have been identified for addressing the ASR mechanism in existing concrete pavements. Some different treatments, each with a different purpose or intent, have been tried, but these have met only limited success. These treatments can be broadly categorized in terms of: • Conventional moisture-reduction treatments, such as joint/crack sealing and subsurface drainage to reduce the amount of water that could help feed the ASR reaction. These treat- ments have not demonstrated any significantly discernible effects in controlling ASR. • Surface treatments and sealers that are applied topically to the pavement, such as silanes/ siloxanes, lithium compounds, and HMWM. While silanes/siloxanes are showing some marginal benefits in minimizing water intrusion and reducing the relative humidity of the concrete, topical applications of lithium compounds have not demonstrated any apparent effect. One issue associated with all of these materials is achieving an adequate depth of penetration because, as a topical application, they predominantly treat only the surface. • Interim corrective measures, such as partial- and full-depth repairs and pressure relief joints. These address the symptom of the distress and not the mechanisms, but can be effective in restoring serviceability and gaining some additional life while other more substantial rehabilitation can be planned and programmed. • Structural overlays, including both asphalt and concrete solutions. These could be an attractive alternative to complete reconstruction. • Reconstruction of the existing pavement, which is the only fully effective treatment that can address ASR in the existing pavement (assuming that appropriate measures are taken to prevent its recurrence in the new pavement). More detailed information on the mechanisms, identification, and mitigation of ASR can be found in some key source documents (e.g., CSA 2000; AFCESA 2006; DOD 2006; Farny and Kerkhoff 2007; Thomas et al. 2011; Thomas et al. 2013a; Sims and Poole 2017; ASTM C1778; AASHTO R 80).