Fly Ash in Spillway Concrete: Should It Be Used in the Wearing Layer?

The spillway wearing layer is the most exposed and abused concrete surface on any dam. It must withstand decades of high-velocity water carrying sand, gravel, and silt at speeds that can exceed 30 m/s during flood discharge. Unlike structural mass concrete buried within the dam body, the wearing layer has no margin for underperformance. It is either hard enough to resist erosion, or it is not.

This makes the question of supplementary cementitious materials (SCMs) in the wearing layer mix a high-stakes decision. Silica fume has an established track record of improving abrasion resistance. But fly ash, the most widely available and cost-effective SCM in South Asia, carries a reputation problem: the perception that it weakens the concrete surface. For projects where alkali-aggregate reaction (AAR) is a known risk, this creates a genuine engineering dilemma. Excluding fly ash may improve abrasion resistance but leave the concrete vulnerable to expansive reactions that cause far more damage over a 50 to 100-year service life.

This technical brief examines the research evidence, field data, and current standards to answer a practical question: should fly ash be included in spillway wearing layer concrete, and if so, how much?

The Three Erosion Mechanisms: A Quick Recap

Before discussing SCM strategies, it is worth recalling the three distinct mechanisms that damage spillway surfaces. Each one responds differently to concrete composition.

Abrasion occurs when sediment-laden water grinds against the concrete surface. It is the dominant erosion mechanism on most spillways, especially those in rivers carrying significant bed load. Resistance depends primarily on aggregate hardness, paste density, and compressive strength.

Cavitation occurs at high velocities (typically above 12 m/s) when water pressure drops below vapour pressure, forming and collapsing bubbles that pit the surface. Cavitation resistance depends more on tensile strength and surface finish than on SCM content.

Chemical dissolution is a slower process where soft water or acidic runoff gradually leaches calcium from the cement paste. SCMs that reduce permeability generally improve dissolution resistance.

For a detailed treatment of all three mechanisms, see our companion article on spillway concrete abrasion and cavitation resistance. This article focuses specifically on abrasion resistance and how fly ash affects it.

ASTM C1138: What the Test Actually Tells Us

ASTM C1138, “Standard Test Method for Abrasion Resistance of Concrete (Underwater Method),” is the primary laboratory test for evaluating spillway concrete. Understanding its limitations is essential for interpreting the research literature.

The test protocol places a concrete disc (approximately 300 mm diameter, 100 mm thick) in a cylindrical container with water and steel grinding balls. An agitating paddle circulates the balls across the concrete surface for 72 hours. Mass loss is measured at 12-hour intervals.

Key points engineers should understand:

• No pass/fail benchmark. ASTM C1138 is purely comparative. There is no threshold value that defines “acceptable” abrasion resistance. Results are meaningful only when comparing multiple mixes tested under identical conditions.
• Typical ranges. Conventional concrete (30 to 40 MPa) typically shows 5 to 10% mass loss after 72 hours. High-performance concrete with silica fume (60 to 70 MPa) typically shows 2 to 5% mass loss. These are general ranges, not specification limits.
• Aggregate dominates. Once the paste wears away, aggregate hardness becomes the controlling factor. Hard igneous aggregates (granite, basalt) consistently outperform softer sedimentary aggregates regardless of SCM content.
• Test variability. Coefficient of variation between laboratories can exceed 15%, making small differences between mixes statistically insignificant. A 1 to 2% difference in mass loss between two mixes tested in different labs is essentially noise.

The practical implication: when evaluating fly ash versus non-fly-ash mixes, only test them side by side in the same laboratory, on the same equipment, using identical aggregates. Comparing published values across different studies is unreliable.

Fly Ash and Abrasion Resistance: The Evidence

The research record on fly ash and abrasion spans four decades. The findings are more consistent than the industry debate suggests.

The Threshold Effect

The most important finding across multiple studies is that fly ash affects abrasion resistance in a dose-dependent manner, with a clear threshold around 15% replacement.

Yen et al. (2007) conducted one of the most comprehensive studies, testing concrete with Class F fly ash at 0%, 15%, 25%, and 40% replacement levels. Their ASTM C1138 results showed:

The 15% threshold has been confirmed by other researchers and aligns with the practical observation that small fly ash additions densify the paste without significantly reducing the calcium hydroxide needed for paste hardness.

Foundational Work by Liu (1981)

Liu’s early research at the U.S. Bureau of Reclamation established the baseline understanding that concrete compressive strength and aggregate hardness are the two dominant factors in abrasion resistance. His work showed that for a given aggregate, increasing strength from 30 MPa to 60 MPa could reduce abrasion losses by 40 to 60%. This finding is critical because it means that if fly ash is used at a level that does not compromise strength (or if the mix is designed to a higher grade to compensate), the abrasion penalty can be effectively neutralized.

Recent Confirmation: Lv et al. (2025)

A recent study by Lv et al. (2025) on dam concrete incorporating 10% fly ash with fibre reinforcement found that the fly ash concrete actually showed 12% lower abrasion loss than the control. The authors attributed this to improved paste density through pozzolanic reaction and better aggregate-paste bond. This result is consistent with the threshold model: at low dosages, fly ash fills micropores and improves the interfacial transition zone without significantly weakening the paste matrix.

The Early-Age Concern

The legitimate concern with fly ash in wearing layer concrete is early-age strength development. Fly ash concrete gains strength more slowly than OPC concrete, particularly in the first 7 to 14 days. If the spillway must be placed in service before the fly ash has fully reacted (for example, due to monsoon scheduling), the surface may be more vulnerable during its initial exposure.

This concern is manageable through two approaches:

1. Design for 56 or 90-day strength rather than 28-day, giving the fly ash time to develop its full contribution.
2. Keep fly ash below 15% in the wearing layer, even if higher dosages are used in the structural mass concrete beneath it.

Silica Fume: The Proven Performer

Silica fume’s role in improving abrasion resistance is well established and not controversial. The mechanism is straightforward: its extreme fineness (average particle size of 0.1 to 0.2 microns, roughly 100 times finer than cement) fills micropores in the paste, reacts with calcium hydroxide to form additional C-S-H gel, and dramatically reduces permeability.

The Kinzua Dam Case Study

The landmark study by Holland et al. (1986) at Kinzua Dam in Pennsylvania demonstrated silica fume’s potential for spillway repair. Concrete with 10% silica fume replacement showed approximately 50% less abrasion loss than conventional concrete of the same grade in ASTM C1138 testing. This study, referenced in ACI 210R-93, “Erosion of Concrete in Hydraulic Structures,” became the basis for widespread adoption of silica fume in spillway concrete.

Indian Project Data: Tala HEP

Data from the Tala Hydroelectric Project in Bhutan provides direct evidence from a South Asian hydroelectric context. ASTM C1138 testing of wearing layer concrete produced the following results:

The data confirms two things: silica fume substantially improves abrasion resistance, and higher compressive strength provides additional benefit. The M70 silica fume mix showed roughly half the abrasion loss of the conventional M40 mix.

ACI Recommendation

ACI 210R-93 specifically recommends silica fume concrete for hydraulic structures subject to abrasion-erosion. The committee report notes that silica fume at 8 to 12% of cementitious content, combined with a low water-cementitious ratio (0.35 or below), produces concrete with “excellent resistance to abrasion-erosion by waterborne particles.”

Binary vs. Ternary Blends: The Real Comparison

The debate about fly ash in spillway concrete is often framed as “fly ash vs. no fly ash.” The more useful comparison is between binary blends (cement plus silica fume) and ternary blends (cement plus silica fume plus fly ash). This reframing changes the analysis significantly.

Binary Blends: Maximum Abrasion, No AAR Protection

A binary blend of OPC plus 8 to 10% silica fume, designed to M50 or M70 grade, delivers the highest abrasion resistance achievable with SCMs. It is the right choice when:

• Aggregates are confirmed non-reactive through ASTM C1260 and ASTM C1293 testing
• Alkali loading from cement is low (Na2Oeq below 0.60%)
• The sole priority is maximizing surface hardness

The limitation is that silica fume alone, at the 8 to 10% dosages used in wearing layer mixes, provides limited protection against AAR. While silica fume does consume some alkalis, its effectiveness against slowly reactive aggregates over long timeframes is not reliable.

Ternary Blends: Near-Equal Abrasion Plus Multi-Threat Protection

Ternary blends add 10 to 25% Class F fly ash to the silica fume system. The research data on this combination is compelling.

Cai et al. (2016) tested ternary systems for hydraulic concrete and found that cement plus 7% silica fume plus 15% Class F fly ash achieved abrasion resistance within 8% of the binary (cement plus 10% silica fume) control, while providing measurably better resistance to chloride penetration and sulfate attack.

PMC 2024 study on high-performance dam concrete showed that ternary blends with 20% fly ash and 8% silica fume achieved 90% of the abrasion resistance of binary silica fume mixes at 90 days, with the gap narrowing at later ages as fly ash pozzolanic reaction continued.

2026 long-term durability study results confirmed that ternary blend specimens maintained their abrasion resistance advantage over conventional concrete after 5 years of accelerated aging, while binary specimens without fly ash showed early signs of ASR-related microcracking in mixes with moderately reactive aggregates.

The comparison in practical terms:

For most hydroelectric projects, the ternary blend delivers the better risk-adjusted outcome.

The AAR Factor: Why This Changes Everything

The discussion above treats abrasion and AAR as separate concerns. In practice, for many hydroelectric projects in South and Southeast Asia, they are deeply connected. This is because the same aggregates used in the spillway wearing layer are typically sourced from local river deposits or quarries, and many of these aggregates are reactive.

Himalayan Aggregates and ASR Reactivity

The geological reality of the Himalayan region is that many common aggregate sources contain reactive silica minerals: strained quartz, chert, chalcedony, volcanic glass, and phyllosilicates. Well-documented cases of ASR damage in South Asian dams include:

• Tarbela Dam (Pakistan): Severe ASR cracking in the spillway structure
• Warsak Dam (Pakistan): ASR-related deterioration documented within 20 years
• Hirakud Dam (India): ASR diagnosed in concrete over 50 years old

The FHWA (Federal Highway Administration) has documented that in field structures across the United States, there have been zero confirmed ASR cases in concrete containing 25% or more Class F fly ash. This is a remarkable statistic and reflects fly ash’s dual mechanism of alkali binding and pore solution chemistry modification.

Why Silica Fume Alone May Not Be Enough

ACI 201.2R-16 and ASTM C1778 provide prescriptive levels of SCMs required to prevent ASR based on aggregate reactivity class and exposure conditions. For moderately to highly reactive aggregates in moist environments (which describes virtually every spillway), the guidance recommends:

• Silica fume alone: 10 to 15% (higher than typically used for wearing layer, and this much SF creates workability problems)
• Class F fly ash alone: 25 to 40%
• Ternary (SF + FA): 7 to 8% SF plus 15 to 25% FA

The ternary approach achieves adequate ASR prevention at SCM levels that remain compatible with abrasion resistance requirements.

The 10 to 30-Year Latency Problem

ASR is not an immediate failure. Slowly reactive aggregates may take 10 to 30 years to develop visible damage. A spillway that performs beautifully for its first decade may begin showing map cracking, gel exudation, and expansion-related structural distress in its second or third decade. By that point, the cost of repair is orders of magnitude higher than the cost of proper SCM selection during original construction.

This latency means that the absence of ASR in young concrete is not evidence that the aggregates are non-reactive. It means the reaction has not yet had enough time, moisture exposure, and alkali accumulation to manifest. For a spillway designed for 75 to 100 years of service, a mix that sacrifices AAR protection for marginally better abrasion resistance is a poor trade.

Recommended Specifications for Spillway Wearing Layer

Based on the evidence reviewed above, the following specifications represent current best practice for spillway wearing layer concrete. Three options are presented to match different aggregate reactivity scenarios.

Option 1: Binary Blend (Non-Reactive Aggregates Confirmed)

Use only when aggregate non-reactivity is confirmed through both ASTM C1260 (14-day mortar bar) and ASTM C1293 (1-year concrete prism) testing with project-specific materials.

Option 2: Ternary Blend, Conservative (Moderate Reactivity Risk)

Appropriate when ASTM C1260 results show borderline expansion (0.10 to 0.20% at 14 days) or when aggregate testing is incomplete.

Option 3: Ternary Blend, Full AAR Protection (Reactive Aggregates)

Required when aggregates are confirmed reactive (ASTM C1260 expansion exceeding 0.20%) or when aggregate sources are variable and reactivity cannot be guaranteed batch to batch.

The Practical Decision Framework

For engineers and project managers making this decision, the flowchart is straightforward.

Step 1: Characterize the aggregates. Conduct ASTM C1260 and ASTM C1293 testing on all proposed aggregate sources. Do not rely solely on petrographic examination, which can miss slowly reactive minerals. If the project timeline does not allow for the 12-month ASTM C1293 test, use ASTM C1260 as a screening tool and default to the conservative ternary approach.

Step 2: Assess the risk tolerance. Consider the service life (typically 75 to 100 years for a dam spillway), the difficulty and cost of future repairs, and the consequence of AAR-related deterioration. For major hydroelectric projects where spillway failure would compromise the dam safety programme, the cost of extra fly ash is trivial relative to the risk.

Step 3: Select the blend.

• Aggregates confirmed non-reactive (both ASTM C1260 and C1293 pass): Binary SF blend is optimal for maximum abrasion resistance.
• Aggregates reactive, borderline, or untested: Ternary blend with 15 to 25% Class F fly ash plus 7 to 8% silica fume. The exact fly ash percentage should be determined through ASTM C1567 testing with the proposed SCM combination.
• Aggregates highly reactive or sources variable: Ternary blend at the upper end (25% FA) with mandatory ASTM C1293 confirmation testing of the proposed mix.

Step 4: Validate with ASTM C1138. Regardless of which option is selected, conduct ASTM C1138 comparative testing on at least three candidate mixes using the actual project aggregates and cementitious materials. This test is the final check that the selected mix meets abrasion performance expectations relative to the other options.

Lessons from South Asian Hydroelectric Projects

PCCI’s experience across six landmark hydroelectric projects, totalling over 4,000 MW, has provided direct insight into the fly ash question in the context of Himalayan and sub-Himalayan geology.

At the Tala HEP, the ASTM C1138 data presented earlier in this article demonstrated that silica fume concrete at M50 and M70 grades achieved excellent abrasion resistance for the spillway wearing layer. The aggregate sources at Tala, drawn from Bhutanese river deposits, were characterized through comprehensive testing to establish the appropriate mix design strategy.

At the Karchham Wangtoo HEP in Himachal Pradesh, mix designs spanning concrete, shotcrete, and grout were developed with careful attention to the locally available aggregates and their long-term durability characteristics.

The consistent lesson across these projects is that the decision about fly ash in the wearing layer cannot be made in isolation. It must be integrated with a comprehensive understanding of aggregate reactivity, exposure conditions, required service life, and the overall SCM strategy for the project.

Conclusion

The question “Should fly ash be used in the spillway wearing layer?” does not have a universal answer. The evidence supports three clear conclusions:

First, fly ash at 10 to 15% of cementitious content does not significantly impair abrasion resistance. The historical reluctance to use any fly ash in wearing layers is not supported by the current research base.

Second, for projects where aggregate reactivity is a concern (and in the Himalayan region, it almost always should be), excluding fly ash from the wearing layer creates a long-term durability risk that far outweighs the marginal abrasion benefit of a pure silica fume system.

Third, the optimal approach for most hydroelectric spillways is a ternary blend combining silica fume (for abrasion resistance and impermeability) with Class F fly ash (for AAR protection and workability), at dosages calibrated to the specific aggregate reactivity and verified through both ASTM C1138 and ASTM C1567/C1293 testing.

The wearing layer is too important and too expensive to repair for mix design decisions to be based on convention rather than evidence. Test the aggregates. Test the mixes. Let the data drive the decision.

If your project requires technical guidance on spillway concrete mix design, aggregate reactivity assessment, or SCM optimization, contact our team to discuss your specific requirements.