Steel Slag Aggregate Cuts Pavement Heat While Meeting Strength Targets

Concrete pavements could be rebuilt with an industrial by-product that keeps their strength intact while trimming how hot they get under the sun. A study published on 7 October 2026 in Springer Nature’s Discover Concrete and Cement reports that basic oxygen furnace steel slag, a dense and angular waste material from steelmaking, can replace natural sand and crushed stone in Portland Composite Cement concrete at levels from 25 percent up to complete substitution, and the resulting slabs still meet the flexural strength required for rigid road pavements while conducting heat markedly better than conventional mixes.

The research team, led by Abdul Fattah Bima Rizqi Purwanto at Universitas Gadjah Mada with collaborators including Taqia Rahman and Elfiranahla Chandra Dewi, cast five concrete mixtures in which steel slag simultaneously replaced both the fine and coarse natural aggregate fractions at 0, 25, 50, 75 and 100 percent by volume. The binder was Portland Composite Cement supplied by PT Semen Indonesia, the natural aggregates were Progo sand and Clereng crushed stone from Yogyakarta, and the slag came from the Indonesian steelmaker PT Krakatau Posco. All mixtures shared a constant water-to-cement ratio of 0.38 and a fixed superplasticizer dosage of 0.2 percent by weight of cement, so that differences in performance could be attributed to the aggregate substitution itself.

The design target was a flexural strength of 4.5 MPa, the minimum specified by the Indonesian standard SNI 8457:2017 for medium- to heavy-traffic rigid pavements carrying more than one million equivalent standard axle loads. Every mixture cleared that bar. Mean flexural strengths ranged from 4.64 MPa at 25 percent substitution to 5.11 MPa at full replacement, with the control mix at 4.73 MPa. Compressive strength, measured on cylindrical specimens at 28 days, stayed within a narrow band of 34.58 to 37.27 MPa but responded non-monotonically: the peak of 37.27 MPa occurred at 50 percent slag content, just above the control’s 37.15 MPa, while the 25 percent mix was weakest. More slag therefore did not translate into steadily stronger concrete.

The physical character of the slag helps explain the mechanical results. Scanning electron microscopy showed a rougher, more irregular and more porous surface on the slag particles than on the natural stone, a texture that can promote mechanical interlocking with the surrounding cement paste. Energy dispersive X-ray spectroscopy revealed a chemically heterogeneous surface, with iron-rich regions reaching 34.6 percent iron at one measurement point and calcium-rich zones reaching 27.3 percent calcium at another. The slag also absorbed far more water than the natural aggregates, 3.25 percent for the fine fraction and 4.77 percent for the coarse fraction, against 0.98 and 2.62 percent respectively, a property the authors flag as relevant to long-term durability questions they did not directly test.

The thermal story is where the study becomes most striking. Thermal conductivity, measured with a transient hot-wire probe on three specimens per mixture, generally rose with slag content, climbing 28.2 percent to a maximum of 1.510 W/m·K at full substitution, though the 50 percent mix bucked the trend with the lowest mean value of 1.117 W/m·K. Density climbed in parallel, increasing roughly 18 to 19 percent over the control at full replacement, a consequence of the slag’s higher specific gravity that raises self-weight and must be weighed against the structural and handling implications the authors say need further study.

To see what that conductivity means in practice, the team ran a 24-hour heating and cooling simulation in an insulated heating box. Five slab specimens, one per mixture, sat under a 1000 W halogen lamp delivering a measured 719.5 W/m² of thermal radiation to the surface for 12 hours, followed by 12 hours of natural cooling, with thermocouples embedded 1.5 centimetres below the top and bottom faces logging temperatures every five minutes. During heating, the fully substituted mix peaked at 66.8 °C on its top surface against 70.3 °C for the control, a reduction of up to 4.3 °C across the slag mixes, while their bottom surfaces ran up to 1.1 °C warmer, evidence that heat was being pushed downward into the slab rather than accumulating at the surface.

The vertical temperature gradient shrank even more dramatically. In the control mix, the difference between top and bottom temperatures reached 17.8 °C during heating; at 75 percent slag it peaked at only 9.5 °C. That matters structurally, because large through-depth gradients in rigid pavements drive non-uniform expansion, thermal stresses and cracking. The authors note that conductive aggregates help keep temperature distribution more uniform, potentially reducing temperature-induced distress over daily cycles. During cooling, however, the trade-off appeared: slag concrete surfaces remained up to 0.7 °C warmer than the control at the end of the cooling phase, meaning the daytime cooling benefit is partly offset by slightly greater heat release at night.

Skid resistance emerged as the clearest limitation. Measured with the British Pendulum Tester on smooth, untreated, largely mortar-covered surfaces, mean values fell from 94.7 BPN for the control to 63.7 BPN at full substitution, a non-monotonic decline in which the 75 percent mix scored 85.1 BPN. All mixtures still exceeded the thresholds in the Transport Research Laboratory’s Overseas Road Note 18, but the fully substituted mix satisfied only the Category B requirement, making it less suitable for demanding sites such as roundabouts and sharp curves. The authors attribute the drop to surface mortar covering the slag rather than to the slag’s intrinsic texture, and recommend surface texturing such as tining, grooving or shot-blasting where high friction is needed.

The sustainability case is substantial on the materials side. Full substitution eliminated natural fine and coarse aggregate use entirely, replacing roughly 1745 kg/m³ of natural aggregate with about 2211 kg/m³ of steelmaking by-product. That matters because steelmaking generates 130 to 200 kilograms of slag per tonne of crude steel, and diverting it into pavements eases both waste accumulation and quarrying pressure. The authors are careful, however, to note that no life-cycle assessment was performed, so potential CO₂ reductions are not quantified, and because albedo and thermal emissivity were not measured, their urban heat island interpretation rests on conductive heat transfer and the measured temperature response alone.

Several caveats bound the conclusions. The heating-cooling test used a single specimen per mixture without replication and was evaluated descriptively; the laboratory box did not reproduce wind, varying solar angles, ambient conditions or real pavement-subgrade boundaries, so results are comparative rather than field predictions. Volumetric expansion from residual free lime and periclase in the slag, a known dimensional-stability risk, was not evaluated, nor were shrinkage or long-term durability. The experimental design also cannot separate the effects of fine and coarse slag replacement. What the study does establish, within one experimental framework, is a set of trade-offs: 50 percent slag gave the highest compressive strength, 100 percent gave the highest flexural strength and conductivity, 75 percent gave the flattest temperature gradient, and skid resistance generally fell as slag rose. No single substitution level optimised everything, and the authors argue that slag pavement mixes should be selected by balancing structural adequacy, thermal behaviour, surface safety and density rather than by maximising replacement.

Subject of Research: Using basic oxygen furnace steel slag as aggregate in Portland Composite Cement concrete for heat-reducing rigid pavements

Article Title: Mechanical and physical properties of portland composite cement concrete with basic oxygen furnace slag aggregates for heat reducing rigid pavements

Article References: Purwanto, A. F. B. R., Saleh, G. R. J., Setiawan, A. F., Siswosukarto, S., Satyarno, I., Rahman, T., Ratnawati, Susanto, T. E., & Dewi, E. C. (2026). Mechanical and physical properties of portland composite cement concrete with basic oxygen furnace slag aggregates for heat reducing rigid pavements. Discover Concrete and Cement, 2(1), Article 81. https://doi.org/10.1007/s44416-026-00125-w

Image Credits: AI Generated

DOI: 10.1007/s44416-026-00125-w

Keywords: steel slag, rigid pavement, Portland Composite Cement, thermal conductivity, urban heat island, flexural strength, skid resistance, recycled aggregates, concrete pavements, temperature gradient, sustainable construction, Indonesia