Calcium-Rich Steel Slag as a Reactive Capping Material: Effects on Hydraulic Conductivity and Nutrient Attenuation in Cohesive Intertidal Sediments

Abstract

Fine-grained intertidal sediments are typically characterized by low hydraulic conductivity and high nutrient loads, conditions that hinder biogeochemical recovery and exacerbate eutrophication. This study examined the feasibility of calcium-rich steel slag (SS) as a multifunctional capping material for improving both physical and chemical properties of cohesive sediments. Short-term (24 h) column experiments with two slag dosages (25 g and 50 g) revealed that the higher dosage (SS50) increased sediment hydraulic conductivity by 113.2%, likely through Ca²⁺-mediated flocculation and enhanced pore connectivity. Phosphate (PO₄-P) in pore water decreased by up to 64.1%, and effluent dissolved inorganic nitrogen (DIN) declined by 62.8%, indicating combined effects of Ca-driven precipitation, adsorption, and enhanced flushing. However, SS addition also raised pore water pH (to 11.8) and lowered redox potential, leading to transient phosphate release at the effluent boundary under reducing conditions. Cation analysis confirmed Ca²⁺ stability and Na⁺ reduction, suggesting improved sediment structural integrity. The results suggest that steel slag is a promising reactive capping material capable of enhancing permeability and controlling nutrient release in cohesive coastal sediments, yet further investigation into long-term ecological effects and dosage optimization is necessary.

1. Introduction

Intertidal sediments are ecologically valuable coastal environments that support biodiversity, buffer wave energy, and facilitate nutrient cycling. However, these habitats are increasingly threatened by anthropogenic influences such as terrestrial runoff, industrial discharge, and the accumulation of fine-grained sediments. Such inputs promote the enrichment of organic matter and nutrients, leading to eutrophication and the development of hypoxic zones [1,2]. These effects are particularly pronounced in cohesive muddy sediments, where low hydraulic conductivity restricts the vertical exchange of oxygen and solutes [3,4]. Improving the hydraulic characteristics of such sediments, alongside their chemical quality, is therefore essential for effective remediation.

To address these challenges, the application of calcium-based reactive capping materials has gained attention as a means to stabilize contaminated sediments and mitigate pollutant release [5,6,7]. Such materials have demonstrated the ability to immobilize phosphate, ammonium, and heavy metals through various geochemical processes such as adsorption, precipitation, and ion exchange [3,4].

Steel production has increased globally in recent years, with an average of over 1.6 billion tons produced annually and nearly 2 billion tons in 2019, indicating a continuous upward trend. In South Korea alone, the annual generation of steel slag (SS) reaches approximately 10 million tons [8]. While efforts have been made to reuse SS in road construction, civil engineering, and cement production [9], nearly 60% of the total global output remains landfilled. To reduce environmental burdens and improve resource efficiency, it is imperative to identify new sustainable reuse pathways.

Steel slag (SS), a calcium-rich byproduct of the steelmaking industry, has been studied as a reactive amendment for soil and sediment remediation due to its high alkalinity and content of iron, calcium, and silicate compounds. Previous studies have demonstrated its effectiveness in reducing nutrient and heavy metal mobility in various environmental settings [10,11,12]. However, the majority of existing research has focused on its chemical reactivity, while its influence on the hydraulic properties of cohesive sediments remains insufficiently explored. Studies on the hydraulic conductivity of steel slag have primarily focused on its use as a construction material in combination with other additives such as cement or bentonite [8,13,14].

Given its coarse texture and porous structure [8,15], SS may also contribute to improving sediment hydraulic conductivity. Enhancing pore water flow is particularly important in fine-grained, oxygen-deprived sediments where natural flushing is limited. Moreover, the abundant calcium content of SS can promote flocculation of fine particles, improving pore connectivity. The dual function of SS—both chemical stabilization and physical enhancement—suggests that it may serve as an effective capping material for long-term sediment recovery. Nevertheless, SS contains trace levels of heavy metals such as Cr, Ni, and Zn, whose potential release has raised concerns about ecological risks including bioaccumulation and aquatic toxicity [16]. However, both recent ecotoxicological investigations of steelmaking slag leachates and ecological risk assessments of contaminated sediments have shown that heavy metals such as Cr, Ni, Pb, and Zn are generally below regulatory thresholds [17,18]. This is consistent with previous studies reporting that these elements are largely present in the form of stable oxides, which considerably limit their leaching potential under environmental conditions [19,20].

Accordingly, this study evaluates the feasibility of SS as a multifunctional capping material for cohesive intertidal sediments. Short-term column experiments were conducted to assess changes in hydraulic conductivity, nutrient concentrations, and associated cation behavior, thereby advancing the understanding of the dual role of SS in improving both physical and chemical conditions of cohesive intertidal sediments.

2. Materials and Methods

2.1. Capping Material and Sediment

The SS in this study was provided by Changshin Industrial Development Co., Ltd. (Pohang, Republic of Korea) and was crushed and sieved through a 600 μm mesh. The composition of SS was determined using X-ray fluorescence (XRF; XRF-1800, Shimadzu, Kyoto, Japan) and X-ray diffraction (XRD; X’Pert-MPD System, Philips, Amsterdam, The Netherland), using Cu Kα radiation, 2α range of 5–80°, with a 0.026° step size.

The chemical composition of the steel slag is shown in

Table 1. Composition (>1%) of the steel slag used in the experiment.

CaO (%)	Fe2O3 (%)	SiO2 (%)	Al2O3 (%)	MgO (%)	P2O5 (%)
42.65	25.21	15.85	4.54	3.99	2.44

. The major oxides were CaO, SiO₂, Fe₂O₃, and Al₂O₃, which together accounted for more than 90% of the total composition. The XRD pattern of the slag (Figure 1) identified crystalline phases dominated by Fe– and Ca–bearing minerals. The main phases were magnetite (Fe₃O₄), calcite (CaCO₃), calcium diferrate (CaFe₂O₄), and wüstite (FeO). These results indicate that Ca is present not only as carbonate but also in complex Ca–Fe oxides, while Fe occurs in both oxide and ferrite phases.

Although XRF analysis also indicated the presence of P₂O₅, Al₂O₃, and abundant Ca and Mg, no distinct peaks corresponding to P– or Al–bearing minerals, free lime (CaO), portlandite (Ca(OH)₂), MgO, or brucite (Mg(OH)₂) were observed in the XRD pattern. This is likely because these components occur in amorphous or poor crystalline forms, or at concentrations below the detection limit. In particular, the absence of CaO and Ca(OH)₂ peaks suggests that Ca exists mainly as carbonate (CaCO₃) or Ca–Fe oxide phases, consistent with carbonation and hydration processes commonly reported in steel slag [21].

The contaminated sediment was collected from Nulcha Bay, Busan, Republic of Korea (35°03′13.9″ N, 128°50′18.6″ E) using a Van Veen Grab (CL-VG 351, HY science, Pohang, Republic of Korea). After removing impurities including macro-fauna and shell debris, the collected sediment was stored in 13 L high-density polyethylene (HDPE) buckets. The samples were then transported to the laboratory and stored at 4 °C until use. The initial properties of the sediment are summarized in

Table 2. Initial properties of the sediment and its pore water used in the experiment. pH, oxidation–reduction potential (ORP), water content, and loss on ignition were measured directly in sediment samples, while nutrients (PO₄-P, NH₃-N, NO₂-N, NO₃-N) were determined from pore water extracted at the start of the experiment.

Parameter	Value	Parameter	Value
pH	7.50	ORP (mV)	−212.5
PO4-P (mg/L)	0.57	NH3-N (mg/L)	8.60
NO2-N (mg/L)	0.05	NO3-N (mg/L)	0.03
Water content (%)	71.6	Loss on Ignition (%)	6.8

.

2.2. Variable Head Permeability Test

The sediment hydraulic conductivity in this experiment was measured based on the ASTM standard method (ASTM D5856-95; [22]). Sediment was put into a cylinder-shaped case with a diameter and height of 10 cm and 12.7 cm (Figure 2) and consolidated using a shaker (SSB-30; SciLab, Seoul, Republic of Korea) at 100 rpm for 12 h. The experiment consisted of three treatment conditions: sediment only (Control), and sediment capped with 25 g (0.13 cm thickness, SS25) or 50 g (0.25 cm thickness, SS50) of steel slag. The hydraulic conductivity of sediment was determined using Equation (1), following the ASTM standard [22].

$$K=2.3{\displaystyle \frac{aL}{A(t_2-t_1)}}{\mathit{log}}_{10}\left({\displaystyle \frac{h_1}{h_2}}\right)$$

(1)

where K is the hydraulic conductivity (mm/s), a is the standpipe area (mm²), and L and A are the sediment height (mm) and area (mm²), respectively. t₂ − t₁ is the time for the water head to decrease from the initial head (h₁) to the final head (h₂). The test was conducted for 24 h using deionized water as the permeating solution.

2.3. Analysis

The pH and oxidation–reduction potential (ORP) of the sediments were measured directly in the container after the hydraulic conductivity test using a pH/ion meter (LAQUA D-53, HORIBA, Kyoto, Japan) equipped with a platinum electrode and an Ag/AgCl reference electrode. To determine water content (WC), sediment was collected from 1 cm below the surface, and the sample was dried in an oven at 100 °C for 6 h. WC was calculated using Equation (2).

$$WC \left(\%\right)={\displaystyle \frac{Moist sediment \left(g\right)-Dried sediment \left(g\right)}{Dried sediment \left(g\right)}}\times 100\%$$

(2)

The sediment was centrifuged (VARISPIN 4, CRYSTE, Bucheon, Republic of Korea) at 3800 rpm for 10 min to extract pore water. The effluent water was collected from the effluent beaker after 24 h. The nutrient concentrations, including PO₄-P, NH₃-N, NO₂-N, and NO₃-N, in the pore water and effluent water were measured using a spectrophotometer (DR3900, Hach, Loveland, CO, USA). All analyses were conducted in triplicate, and the values are presented on average with standard deviation. Effluent loads (mg) were calculated as the product of effluent concentration (mg/L) and effluent volume (L). Because deionized water was used as the influent, inflow nutrient loads were negligible and only effluent loads were considered. Areal fluxes (mg/m²/d) were then derived by dividing the load by the column cross-sectional area (m²) and the incubation time (d).

The cation concentrations (Ca²⁺, Na⁺, K⁺, and Mg²⁺) in pore water were determined using Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES; Avio 220 MAX, PerkinElmer, Shelton, CT, USA). The obtained concentrations were used for the calculation of the sodium adsorption ratio (SAR) using Equation (3)

$$SAR{({mmol}_c/L)}^{0.5}={\displaystyle \frac{Na}{\sqrt{Ca+Mg}}}$$

(3)

where Na, Ca, and Mg are the molar concentrations of the cations (mmol_c/L).

Saturation indices (SI) for Ca-phosphates were computed using PHREEQC version 3.8.6 with the minteq.v4 database at 25 °C. Because major anions/alkalinity were not measured, electroneutrality was enforced by specifying chloride as a computational balancing ion.

3. Results

3.1. Changes in Sediment Hydraulic Conductivity, Water Volume, and Water Content

The results of the variable head hydraulic conductivity measurement are presented in Figure 3a. In all cases, hydraulic conductivity decreased over the first 7 h, ranging from 3.98 to 5.07 μm/s. After that point, the hydraulic conductivity of SS50 increased to 6.16 μm/s, while SS25 and Control both decreased to 2.89 μm/s. The higher hydraulic conductivity in SS50 can be attributed to the release of Ca²⁺ ions from the steel slag. According to Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory, these divalent ions displace exchangeable Na⁺ on clay surfaces, compress the diffuse double layer, and lower the zeta potential, thereby weakening electrostatic repulsion and allowing van der Waals attraction to dominate. This process promotes particle flocculation, resulting in larger aggregates and increased porosity, which in turn increases hydraulic conductivity [3,23,24]. In contrast, the hydraulic conductivity in Control remained low due to sediment swelling, which restricted water movement [25]. The SS25 treatment showed no significant improvement over Control, suggesting that a smaller amount of SS is insufficient to enhance hydraulic conductivity. This may be attributed to the consumption of Ca²⁺ through precipitation reactions, which reduced the availability of free Ca²⁺ for cation exchange [26,27].

It should be noted that this study was limited to a 24 h experimental period, which primarily reflects short-term effects such as Ca²⁺-mediated flocculation. In long-term field applications, however the behavior of steel slag capping layers may be more complex. Over extended timescales, pozzolanic-like reactions between Ca, Si, and Al phases could lead to gradual solidification and enhanced stability of the sediment bed, a mechanism comparable to slag-based geopolymer stabilization reported in geotechnical studies [28,29,30]. Future studies should therefore examine the long-term physical stability of slag capping layers, as well as the associated mineralogical evolution and ecological consequences, before field-scale implementation.

The influent water volumes of Control, SS25, and SS50 were 102 mL, 99 mL, and 144 mL, respectively, whereas the effluent volumes were 92 mL, 94 mL, and 138 mL, respectively, (Figure 3b). SS50 has a 1.4 and 1.5-times higher influent and effluent water volume compared to Control, respectively. As discussed previously, higher Ca²⁺ concentration in SS50 resulted in higher hydraulic conductivity and thus a high-water volume can pass through the sediment.

The WC after the experiment in Control, SS25, and SS50 are 74.9%, 69.9%, and 62.2%, respectively, (Figure 3c). WC decreased as the amount of SS increased, primarily due to greater physical consolidation of the sediment under the heavier load of the slag material [31]. The high hydraulic conductivity observed in SS50 promoted water release from the sediment, resulting in the lowest WC. Meanwhile, WC in Control increased slightly compared to the initial condition, likely due to swelling behavior that retains water within the sediment matrix [32]. It should be noted that WC was measured only for the underlying sediment (beneath the slag layer) and does not include water retained within the slag layer itself. This explains the discrepancy observed in SS50, where the retained water estimated from influent–effluent balance (6 mL) did not match the large decrease in WC (−9.4 mL), likely due to additional water held within the porous slag material. Furthermore, WC is a composite measure that includes pore water, crystalline water within minerals, and surface-adsorbed water. Therefore, in this study WC is interpreted as a relative indicator of treatment differences rather than an absolute quantification of specific water types.

3.2. Effect of SS on pH and ORP Changes in Sediment

As shown in Figure 4a, the pH in Control remained comparable to that of the initial sediment (7.49), while SS25 and SS50 increased significantly to 10.06 and 11.80, respectively, (p < 0.05). The pH increased as the increment of the SS portion in sediment owing to the hydration of Calcium Oxide (CaO) and Magnesium Oxide (MgO) from the SS, which produces hydroxyl ions that are responsible for the pH increase [33]. This rapid increase in alkalinity is an ecological concern because high pH can adversely affect native microbial and benthic communities [34]. Previous studies on slag filters reported that hydraulic residence time (HRT) strongly influences effluent pH. In our sediment system, HRT was not directly controlled; instead, slag dose and hydraulic conductivity governed water–slag contact and thereby affected pH. The two doses tested here (25 g and 50 g) illustrate this dose effect but do not define an optimal range. Future studies should refine dosing levels and incorporate biological assessments to ensure that alkalinity-related risks are minimized while treatment performance is maintained.

The ORP of initial sediment, Control, SS25, and SS50 are −212.5 mV, −137.8 mV, −248.1 mV, and −290.1 mV, respectively, as presented (Figure 4b). The ORP in Control increased compared to initial sediment ORP. This could be due to the introduction of the permeating solution that supplied dissolved oxygen into the sediment [35]. In contrast, the ORP decreased in SS25 and SS50, with the most negative value observed in SS50. The decrease in ORP may show partial consistency with the pH–ORP relationship described by the Nernst equation [36]. However, the ORP in sediments is an integrated signal influenced by multiple processes. The strong alkalinity induced by slag dissolution may have shifted microbial activity toward reductive pathways (e.g., nitrate or sulfate reduction), while the release of reduced species such as Fe²⁺ and Mn²⁺ could also have contributed to lowering the ORP [37]. Moreover, although higher hydraulic conductivity can enhance oxygen transfer, particle flocculation may have restricted oxygen diffusion at the microscale [38,39]. These factors suggest that the ORP response reflects a complex interplay of geochemical and microbial mechanisms rather than a single electrochemical equilibrium.

3.3. Effect of SS on Nutrient Release in Pore Water and Effluent

3.3.1. Pore Water and Effluent Concentrations

PO₄-P concentrations in pore water and effluent water are shown in Figure 5. PO₄-P concentrations in pore water were significantly reduced in SS25 (1.07 mg/L, 30.1% decrease) and SS50 (0.55 mg/L, 64.1% decrease) compared to the Control (p < 0.05). This reduction is attributed to phosphate precipitation with Ca²⁺ and adsorption onto the slag surface [26,35]. Moreover, the lower PO₄-P concentration in SS50 was additionally due to release through the effluent water.

Table 3. Pore water Ca–phosphate saturation indices in Control and SS50.

Phase	Formula	Saturation Index
		Control	SS50
hydroxyapatite (HAP)	Ca10(PO4)6(OH)2	7.02	16.94
β-tricalcium phosphate (β-TCP)	Ca3(PO4)2	0.01	3.64
brushite (DCPD)	CaHPO4·2H2O	−1.53	−4.19

presents the calculated pore water saturation indices (SI) for Ca–phosphate phases in Control and SS50. Control (pH: 7.32) was oversaturated with hydroxylapatite (HAP) and at equilibrium for β-tricalcium phosphate (β-TCP), whereas brushite (DCPD) was undersaturated (HAP: 7.02; β-TCP: 0.01; DCPD: −1.53). SS50 (pH 11.80) was strongly oversaturated with HAP and β-TCP, and DCPD remained undersaturated (HAP: 16.94; β-TCP: 3.64; DCPD: −4.19). These values indicate a strong thermodynamic drive for Ca-phosphate formation—dominated by HAP under alkaline conditions. DCPD is a transient, less stable phase that tends to transform to HAP, commonly through octacalcium phosphate [40]. SI is an equilibrium metric; it indicates thermodynamic driving force, not reaction kinetics or removal rate. Actual phosphate removal is determined by nucleation and growth kinetics and process conditions such as high Mg²⁺ to suppress HAP formation, short hydraulic retention time, limited solid–liquid separation. Consequently, effluent PO₄-P can remain elevated even when the pore water is thermodynamically supersaturated.

In effluent water (Figure 5b), PO₄-P concentrations of Control, SS25, and SS50 were 0.43 mg/L, 0.25 mg/L, and 0.53 mg/L, respectively. Considering the similar hydraulic conductivity observed in Control and SS25 (Figure 3a), the lower PO₄-P concentration in SS25 is likely attributed to Ca²⁺ adsorption and phosphate precipitation rather than hydraulic conductivity effects. In contrast, SS50 showed a higher PO₄-P concentration (0.53 mg/L) in the effluent than both Control and SS25. The lower ORP and elevated pH likely weakened the Fe–P associations, leading to phosphate release despite the availability of Ca for precipitations [41,42]. Additionally, the high hydraulic conductivity in SS50 may have accelerated phosphate discharge before sufficient time was available for calcium-mediated precipitation. Thus, while slag addition clearly lowered pore water phosphate through Ca– and Fe–driven immobilization, excessive hydraulic conductivity in SS50 offset this benefit at the effluent boundary. These results show that slag dosage and residence time jointly determine the balance between Ca–P formation, P release, and transport [43]. These findings indicate that future applications should consider adjusting slag layer thickness or combining slag with sorbents effective under low-ORP conditions to mitigate such transient phosphate release [7,42].

As shown in Figure 6a, DIN concentrations in pore water decreased slightly from 8.6 mg/L in the initial sediment to 8.0 mg/L in Control, 7.8 mg/L in SS25, and 7.6 mg/L in SS50. The decrease was primarily associated with reductions in NH₃-N, which constituted the dominant fraction of DIN, while NO₂-N and NO₃-N contributed only marginally. In effluent water (Figure 6b), DIN concentrations were much lower than in pore water, with values of 5.8 mg/L in Control, 3.9 mg/L in SS25, and 2.2 mg/L in SS50. Compared to Control, effluent NH₃-N declined by 31.9% in SS25 and 62.8% in SS50, while NO₃-N decreased by 54.2% and 75.0%, respectively.

In SS25, the limited improvement over Control reflects minimal enhancement in vertical flushing, as influent and effluent volumes were nearly identical. Consequently, nitrogen was largely retained in pore water, with only modest decreases in NH₃-N and NO₃-N. By contrast, SS50 exhibited both higher hydraulic conductivity and greater effluent volume, which promoted enhanced DIN transport and removal.

Beyond these physical effects, geochemical and microbial pathways are also likely to have contributed. The elevated pH induced by slag dissolution favors the conversion of ammonium (NH₄⁺) to ammonia (NH₃), enhancing volatilization [44]. Released Ca²⁺ can further promote ammonium adsorption onto sediment surfaces [45], while alkaline and reductive conditions may stimulate microbial denitrification, reducing nitrate and nitrite to gaseous nitrogen species [46]. These processes together explain the concurrent declines in NH₃-N and NO₃-N observed in SS50. Although denitrification is unlikely to dominate during the short 24 h experiment, it may play a greater role over longer timescales in natural settings.

3.3.2. Effluent Loads and Fluxes

Effluent loads and areal fluxes of nutrients are summarized in

Table 4. Effluent loads and areal fluxes of nutrients for SS treatment.

Nutrients	Treatment	Effluent Load (mg)	Flux (mg/m2/d)
PO4-P	Control	0.039	4.98
	SS25	0.024	2.99
	SS50	0.073	9.22
NH3-N	Control	0.520	66.18
	SS25	0.362	46.08
	SS50	0.290	36.90
NO2-N	Control	0.004	0.50
	SS25	0.003	0.42
	SS50	0.005	0.62
NO3-N	Control	0.011	1.41
	SS25	0.005	0.66
	SS50	0.004	0.53

. For PO₄-P, the control released 0.039 mg (4.98 mg/m²/d), whereas SS25 decreased the load to 0.024 mg (2.99 mg/m²/d; 40% reduction). In contrast, SS50 showed the highest PO₄-P release (0.073 mg; 9.22 mg/m²/d; 85% increase relative to the control). This result suggests that a moderate slag dosage suppressed phosphorus mobility, while excessive slag addition (SS50) enhanced P release, possibly due to increased effluent volume and changes in P-binding conditions.

Among the nitrogen species, NH₃-N was the dominant form in the effluent. The control load was 0.520 mg (66.18 mg/m²/d), while SS25 and SS50 decreased to 0.362 mg (46.08 mg/m²/d) and 0.2898 mg (36.90 mg/m²/d), respectively, representing reductions of ~30% and ~44%. NO₂-N and NO₃-N loads were minor (≤0.011 mg), but both also decreased under slag treatments, with NO₃-N showing reductions > 50%.

These results indicate that slag addition influenced N and P release in different ways, suggesting a potential imbalance between nitrogen reduction and phosphorus release under higher slag dosage.

3.4. Effect of SS on Cation Concentrations and SAR Changes in Sediment

The concentrations of cations (Na⁺, Ca²⁺, K⁺ and Mg²⁺) in Control and SS50 are presented in Figure 7. The Na⁺ concentration of SS50 was 7131.8 mg/L, which was lower than the control (7726.2 mg/L), and the Ca²⁺ concentration was 433.9 mg/L, which was slightly higher than the control (410.9 mg/L). The control and SS50 concentrations of Mg²⁺ were 1089.0 mg/L and 974.0 mg/L, respectively. In SS50, higher hydraulic conductivity increased dilution and effluent export, whereas Ca²⁺ released from slag was partly consumed by flocculation/particle binding and by Ca–phosphate precipitation (Table 3) [23,26]. Sodium likely underwent exchange with Ca²⁺ and was subsequently flushed under the greater throughput [21]. Magnesium behavior at pH 11.8 aligns with brucite supersaturation (Table S1), which can sequester Mg²⁺ from solution.

Based on the ionic concentration data presented in Figure 7, the sodium adsorption ratio (SAR) was calculated to be 45.3 (mmol/L)^0.5 in Control and 43.5 (mmol/L)^0.5 in SS50. This slight decrease may reflect the increased availability of divalent cations (Ca²⁺ and Mg²⁺) relative to Na⁺ in SS50. Although the absolute change was small, the reduction in SAR is meaningful because SAR is widely used in soil science as an indicator of sodicity and dispersion risk. As a widely accepted indicator from soil science, a lower SAR mitigates the risk of clay particle dispersion and promotes structural stability [47]. Thus, even a modest decrease in SAR in SS50 suggests a potential improvement in sediment stability and reduced risk of sodium-induced dispersion, complementing the observed increase in hydraulic conductivity.

4. Conclusions

This study investigated the feasibility of steel slag (SS) as a multifunctional capping material for cohesive intertidal sediments, with emphasis on changes in hydraulic conductivity and nutrient dynamics. In particular, the higher slag dosage (SS50) increased hydraulic conductivity through Ca²⁺-induced flocculation, leading to improved pore water flushing and reductions in dissolved inorganic nitrogen loads. However, excessive slag addition elevated sediment pH and lowered redox potential, conditions that promoted phosphate release despite strong thermodynamic driving forces for Ca–P precipitation. Moderate slag application (SS25) effectively reduced pore water phosphate concentrations and effluent P fluxes, suggesting that dosage optimization is critical to balance nutrient immobilization with ecological safety.

This short-term study indicates that steel slag offers dual functionality for potential sediment remediation. Future research including detailed mineralogical characterization and chemical analyses—such as anion measurements for mass balance and nutrient retention/release processes—will allow more precise geochemical interpretations. In addition, longer-term experiments combined with ecological assessments will provide valuable opportunities to deepen mechanistic understanding of slag–sediment interactions and to guide the practical application of steel slag in coastal restoration.