Abstract
Fukuyama inner harbor is an inlet located in the city area of Fukuyama, Hiroshima, Japan, where the offensive odor of the hydrogen sulfide released from organically enriched, anoxic sediments has become a social issue. Since hydrogen sulfide is also highly toxic and reactive with oxygen, its presence may cause further oxygen depletion in the bottom waters and ultimately, the total decay of local aquatic ecosystems. To evaluate its effects on dissolved sulfide formed in the sediments, steelmaking slag was applied in two different experimental treatments, capping the sediments and mixing with the sediments. The results revealed that both uses of the steelmaking slag significantly suppressed hydrogen sulfide gas by reducing the dissolved sulfide in the sediment interstitial water. This was clearly much more effective than the results obtained by capping with natural stones that was conducted as a control method. The examination with a SEM-EDX suggested that iron sulfide may form on the slag surface after immersing in the sulfide-containing Fukuyama harbor sediments. It was concluded that applying steelmaking slag would be an effective method to suppress the annoying odor of hydrogen sulfide gas generated from the sediments of coastal areas due to its chemical reaction with sulfide ions.
1. Introduction
In enclosed bays and inlets with stagnant water, organic matter tends to accumulate on the bottom and then decomposes, resulting in an oxygen-deficient condition, thereby creating a reducing environment suitable for activation of sulfate-reducing bacteria. It is known that sulfate-reducing bacteria produce hydrogen sulfide from sulfate ions.1,2) Hydrogen sulfide causes a “blue tide”, and its high toxicity may damage organisms living nearby. As an example, in Tokyo Bay, there are a number of dredged pits where the dissolved oxygen concentration decreases dramatically in summer, and blue tides occur every year.3,4)
Fukuyama inner harbor is an artificially dug canal-like inlet located in the head of Fukuyama Port, Hiroshima, Japan, and is 100 m in width, 2.2 km in length, and 2–4 m in depth. The bottom sediments are silty, containing high organic matter.5) Although there is a combined sewage treatment plant at the head of the canal, untreated sewage is often released when heavy rainfall exceeds the facility’s holding capacity. This causes organic matter loading in the inlet. As a result, a reductive environment is formed in the bottom layer of the head of the inlet from spring through summer, and the odor of the hydrogen sulfide that is released is considered a public nuisance. Thus, remediation of the environment is highly desired from citizens living around the inlet.
In Japan, 45 million tons/year of iron slag is generated. Blast furnace slag, which comprises 2/3 of the total slag generation, has been used as “blast-furnace slag cement”,6) a substitute for ordinary cement, which is regarded as a high value-added product in a recycling-oriented society. On the other hand, steelmaking slag, 1/3 of the total generation, has not been properly utilized. Recent researches have focused on the high iron content of steelmaking slag, sometimes up to 20%. Steelmaking slag that contains high amounts of iron can be used to recover rocky shore desertification and to improve bottom sediment quality in enclosed seas by Fe ion elution.7,8,9)
In a laboratory study using bottom sediment, it was reported that the dissolved sulfide concentration was decreased by the addition of steelmaking slag.10) In our small-scale on-site experiment, the authors observed that dissolved sulfide in the interstitial water of the bottom sediments was reduced by placing stone-size steelmaking slag on the bottom sediment.8) More recently, we also clarified the mechanisms of hydrogen sulfide reduction by steelmaking slag in laboratory experiments.11,12,13) To date, other industrial by-products such as granulated coal ash14,15) and oyster shells16,17) have also been applied to remediate sediment quality. However, a number of local harbors and inlets suffer from the odor of hydrogen sulfide gas. We need to seek a material being produced in such large amounts that it can suppress the odor of hydrogen sulfide wherever it arises.
In the present study, we conducted laboratory experiments using steelmaking slag to remediate sediment quality in terms of hydrogen sulfide removal in different application modes: placing the steelmaking slag over the bottom sediments or mixing it with them. The laboratory experiments were conducted as preliminary tests towards field experiments planned for Fukuyama inner harbor.
2. Materials and Methods
2.1. Materials Used in the Experiments
The bottom sediment (hereafter, simply “sediment”) for the present experiments was collected from Fukuyama inner harbor in September 2010 using a grab sampler. The sediment was silty and rich in organic matter (Fig. 2), and of poor environmental quality (Table 1); the total sulfide concentration of the sediment was more than 10 times the Water Quality Standard for Fishery Use (0.2 mg/g dry mud),18) and the COD was higher than the standard (20 mg/g). The sediment had a strong hydrogen sulfide gas odor.
Table 1. Properties of the Organically enriched sediments collected from Fukuyama inner harbor.
| Water content ratio (%) | Total-sulfide (mg/g) | COD (mg/g) | Total-N (mg/g) | Total-P (mg/g) | Ig/loss (%) |
|---|
| 340 | 2.33 | 29.0 | 3.22 | 1.23 | 13.4 |
The steelmaking slag (hereafter, simply “slag”) generated at West Japan Works, JFE Steel Corporation was sieved to 5–10 mm in diameter for use in the present experiments (Fig. 3). As summarized in Table 2, major components of the slag are Fe, CaO, SiO2, and Al2O3, of which Fe and Ca contents were particularly high compared to those in the natural stones (granite) of the same size (5–10 mm), which were used for comparison with the slag.
Table 2. Chemical compositions of steelmaking slag and granite (mass%).
| M. Fe | FeO | Fe2O3 | SiO2 | CaO | Al2O3 | MnO | MgO | P2O5 | TiO2 | S | Na2O | K2O |
|---|
| Slag | 3.4 | 8.4 | 7.9 | 29.3 | 33.0 | 6.0 | 8.7 | 4.9 | 3.8 | 1.2 | 0.1 | – | – |
| Granite | – | 0.67 | 0.44 | 76.1 | 0.80 | 13.6 | 0.011 | 0.024 | 0.001 | 0.033 | – | 3.1 | 4.4 |
Experiments were carried out from October 5, 2011 to March 27, 2012 at Takehara Marine Science Station, Hiroshima University, Takehara City, Hiroshima Prefecture. The slag was placed on top of the sediment (hereafter we call this “Slag Capping”) or was mixed with the sediment (hereafter we call this “Slag Mixing”), as illustrated in Fig. 4. These two different treatments were performed by assuming practical constructions to which steelmaking slag is applied. Black polyethylene containers with 30-l volume (346 mm in inside diameter and 320 mm in height) were placed in a large-volume (500 l) black polyethylene container, and natural seawater offshore of the experimental station was introduced into the large container to minimize the fluctuation of the water temperature between each experiment container. For the Slag Capping treatment, 3.0 liters of slag were placed on top of 15.0 liters of sediment. Although the slag partially sank into the sediment due to its high specific gravity, a slag layer approximately 30 mm thick was visually observed to have remained on top of the sediment. For the Slag Mixing treatment, 3.0 liters of slag was mixed with 15.0 liters of sediment. For the sake of comparison, a third treatment was performed with 3.0 liters of natural stones of the same grain-size range (5–10 mm) as the slag placed on top of 15.0 liters of sediment (hereafter, “Natural Stone Capping”). In addition to these three treatments, a treatment consisting of only 15.0 liters of sediment (approximately 160 mm high) was placed in the container as a “Control”.
Each 30-l container was capped with an air-tight lid, and a Tygon® tube (inner diameter: 4 mm) was connected to the lid. Sand-filtered seawater pumped up from the coast in front of Takehara Science Station (Table 3)19) was introduced to all the containers at the same flow rate of 3.0 l/day. The flow rate was determined based on an estimation of the mean exchange rate of seawater in Fukuyama inner harbor.20) The lid was convex in shape to maintain a 10-ml headspace to allow us to collect gas samples. Each experimental treatment was performed in triplicate.
Table 3. Temporal changes of Salinity and temperature of sand-filtered seawater pumped up from the coast in front of Takehara Science Station.
| Oct. 12, 2011 | Nov. 16 | Dec. 20 | Jan. 25, 2012 | Feb. 15 | Mar. 5 |
|---|
| Salinity (psu) | 31.48 | 31.70 | 32.80 | 32.42 | 32.90 | 32.92 |
| Temperature (°C) | 24.60 | 21.20 | 18.30 | 12.05 | 10.59 | 10.23 |
Samples were collected at 15, 27, 40, 68, 122 and 172 days. Gas in the headspace was collected using a syringe, and the hydrogen sulfide gas content was measured with detection tubes (4L, 4LK, and 4LT; manufactured by Gastec). Then, the lid was removed and overlying water samples were collected at 80 mm above the sediment surface using a 50-ml syringe (Fig. 5). Sediment samples were collected at 10 mm and 80 mm below the sediment surface, and were centrifuged (3000 rpm, 15 min) to collect interstitial waters. Water temperature, pH, oxidation-reduction potential (ORP) and dissolved sulfide concentrations were also measured for the overlying water and the interstitial waters before centrifugation. The pH and ORP were measured using HM-30P and RM-30P meters (both manufactured by DKK-Toa Corp). Dissolved sulfide concentrations were determined by Kitagawa-type detection tubes (200SA and 200SB; manufactured by Komyo Rikagaku Kogyo).
2.3. Observation of Slag Particles Immersed in Sediments
Slags of the same quality as used in section 2.2 were placed on the sediments at Fukuyama inner harbor in August 2011, and collected after four months (December 2011). Sediments adhering to the slag surface were carefully removed using paper towels. Elemental mapping analysis of Ca, Fe, and S on the surface of the slag particles was performed by scanning electron microscope-energy dispersion type X-ray spectroscopy (SEM-EDX; manufactured by Japan Electron Optics Laboratory). Linear analysis was performed for the elements Fe, S, C, O, Na, Mg, Al, Si, P. Cl, Ca and Mn at a 0.5 mm interval × 20 points. For comparison, elemental mapping analyses of Ca, Fe and S of sediment and slag that was not mixed with the sediment were also performed.
3. Results
3.1. Hydrogen Sulfide Reduction Experiments from Sediments Using Steelmaking Slag
3.1.1. Change in pH
The temporal changes of temperature and the pH values in the interstitial water and overlying water are shown in Figs. 6 and 7, respectively. Although the water temperature changed during the experimental period due to the influence of seasonal variations, few temperature differences were observed among the different experimental settings.
In Slag Capping, the pH ranges were 7.7–8.3, 7.3–8.0 and 7.6–8.2, respectively, in the upper and the lower interstitial water and the overlying water. For Slag Mixing, they were 7.5–8.3, 7.7–8.5 and 7.5–8.1, respectively. In Natural Stone Capping, the pH showed a slightly lower values with small fluctuations; 6.8–7.3 and 7.0–7.7 in both the upper and lower interstitial waters and in the overlying water, respectively. In the Control treatment, the pH range of the upper/lower interstitial waters was 6.8–7.3, while that of the overlying water was 7.0–7.4.
3.1.2. Dissolved Sulfide
Temporal changes of dissolved sulfide in the interstitial water and overlying water in each experiment condition are shown in Fig. 8. In Slag Capping, the dissolved sulfide concentrations in the lower interstitial water showed a range of 0.6–7 mg-S/L, while it was low, being at the detection limit (0.5) to 1.0 mg S/L (except immediate after installation) in the upper interstitial water. In contrast, in Slag Mixing, no dissolved sulfide was detected in all the position even in the lower interstitial water, except immediately after installation.
In Natural Stone Capping, the dissolved sulfide concentrations were quite high in the interstitial water with values of 6–21 mg S/L in the upper layer and 5–70 mg S/L in the lower layer, respectively (except immediately after installation), while it was low with the values below the detection limit in the overlying water, indicating physical suppression by stones. In the Control treatment, the dissolved sulfide concentrations were significantly high not only in the interstitial water (15–57 mg S/L in the upper layer and 16–150 mg S/L in the lower layer) but also in the overlying water (2–10 mg S/L).
In all cases, those of the overlying waters were below the detection limit, which are much different from the condition of the Control treatment. These results confirmed that the concentration of dissolved sulfides in the overlying water was controlled by capping the sediments with slag and even with natural stones, and by mixing with slag. Regarding the dissolved sulfide concentrations in the interstitial waters, it was remarkably lower in the two slag treatments than in the Natural Stone Capping and Control. Moreover, the effect of reduction of sulfides by slag application lasted for the entire period of the experiments (6 months).
3.1.3. Oxidation-reduction Potential
Temporal changes in ORP values in the interstitial waters and overlying waters under each experimental condition are shown in Fig. 9. In Slag Capping, the ORP values ranged from −130 to −60 mV in the upper interstitial water, and those of the lower interstitial water were slightly lower with the values of −180 to −110 mV, and were positive in the overlying water with values from 100 to 300 mV, except immediately after installation. In Slag Mixing, the ORP values were not much different in the upper and lower layers of the interstitial water, ranging from −240 to −50 mV and −210 to −70 mV respectively, while those in the overlying water were 85 to 300 mV.
In Natural Stone Capping, the ORPs were again not so different between the upper and lower layers of the interstitial water, ranging from −160 to −130 mV and −170 to −120 mV while those in the overlying water were high with values of 100 to 300 mV. In the Control treatment, not only were there negative values, −170 to −130 mV and −180 to −130 mV, in the upper and the lower interstitial waters, but values were also sometimes negative in the overlying water, from −50 to 80 mV.
From these results, there is a tendency of higher ORP values in the upper interstitial water in Slag Capping compared with the other three conditions, while no significant difference in ORP was observed in the lower interstitial water. Furthermore, positive ORP values were observed in the overlying water in Slag Capping, Slag Mixing and Natural Stone Capping, but not in the Control.
3.1.4. Hydrogen Sulfide Gas
Temporal changes in the concentration of gaseous hydrogen sulfide in the headspace above the seawater are shown in Fig. 10. In both Slag Capping and Slag Mixing, the H2S gas concentrations were less than the detection limit. In Natural Stone Capping, H2S gas ranged from the detection limit to 3.7 ppm, depending on the measurement dates. In the Control, H2S gas was much higher in the values ranging from the detection limit to 120 ppm. From these results, it was confirmed that both Slag Capping and Slag Mixing are effective in suppressing volatilization of H2S gas to the atmosphere.
3.2. Examination of Steelmaking Slag Surface Immersed in Sediments
A SEM image and EPMA elemental maps of the surface of the slag immersed in the Fukuyama harbor sediments and also pictures of the sediment particles are shown in Figs. 11 and 12, respectively. On the slag surface, the distributions of Fe and S showed good correspondence in their areal distributions; however, there was virtually no correspondence between Ca and S. On the other hand, on the sediment, Fe, S, and Ca were almost uniformly distributed, and there was no coincidence in the distributions of Fe and S as observed on the slag surface. For comparison, an SEM image and EPMA elemental maps of an intact slag particle that was not immersed in the sediment are shown in Fig. 13. No coincidence in the distributions of Ca, Fe and S was observed. Figure 14 shows the molar ratio of Fe and S that were obtained from the three lines in Fig. 11 (20 points each with an interval of 0.5 mm along the lines). High S values coincided with high Fe values, and the molar ratio of Fe to S was larger than 1.
4. Discussion
The results of Fig. 11, in which the EPMA elemental mapping showed good correspondence in the distributions of S and Fe at the slag surface after being immersed in the Fukuyama inner harbor sediment suggested the possibility that iron sulfide (FeS) is formed on the surface of the steelmaking slag. These results agreed with those of other reports.11,12) The molar ratio of Fe/S examined on the surface of the slag after immersion in Fukuyama harbor sediment was larger than 1 (Fig. 14), which also supports the notion that the major compound formed in the reaction of dissolved sulfide with steelmaking slag is iron sulfide (FeS).
As a result of a reflectionless XRD analysis and X-ray absorption fine structure (XAFS) analysis, it is presumed that iron sulfide and elemental sulfur11) were formed by the reactions expressed in Eqs. (1) and (2)23,24)
|
H
S
-
+F
e
2+
→FeS+H+
| (1) |
|
H
S
-
+2F
e
3+
→
S
0
+2F
e
2+
+H+
| (2) |
In addition to the formation of FeS and S0, the pH values (8.0–8.5) observed in slag applications may have a cause that changed HS− to sulfate ion (SO42−) as reported in the literature.12) In the conditions with both Slag Mixing and Slag Capping, it is supposed that the dissolved sulfides may not be the main forms even under reductive conditions.
Nevertheless, it would be important to say that the pH values were in the range similar to the ordinary seawater and did not reach the level above 9.5 at which so-called white turbidity occurs due to the generation of magnesium hydroxide.21) One of the possible reasons is the neutralization effect by mixing with the sediment of low pH (ca. 7.0). Another may be due to calcium ions eluted from the steelmaking slag, which might be partly consumed by hydration reaction with compounds such as silicate in the sediment before diffusing into the overlying water.
During the time course of the present experiments, dissolved sulfide in the interstitial water tended to increase peaking at Day 122 and decreased again to Day 172 (Fig. 8). This may imply that the successive processes in combination of transformation of dissolved sulfides and transportation of H2S gas from water to the air. In waters like Fukuyama inner harbor, where bottom sediments are in reductive condition containing rich organic matter, dissolved sulfides, H2S (aq.) and HS−(aq.), are formed from sulfate ions (SO42−) through sulfate-reduction by bacteria.22) The dissolved sulfides formed in the sediments diffuse to the overlying water. Then the H2S (gas) can be released into the air. It is considered that these successive processes may have occurred in our experiments.
Slag Mixing may be the best way in terms of remediation of reductive condition of sediments. On the other hand, Slag Capping would be the second best for remediation of sediment quality itself, but it would be practically the best way on the aspect of construction when we apply the method to the field. Furthermore, packing of slag grains on the sediment surface may play roles to change the form of sulfur from HS− to SO42− and/or FeS formation in the diffusion process between the interstitial water and the overlying water.
The hydrogen sulfide reduction effects by Natural Stone Capping is much different from those with Slag capping in terms of reduction of dissolved sulfide in the interstitial water, although release of dissolved sulfide can be prevented even in Natural Stone Capping. Furthermore, there was little difference in dissolved sulfide concentrations in the interstitial water between Natural Stone Capping and the Control. This suggests that Natural Stone Capping is a method that physically suppresses the release of dissolved matter from the sediment but not a way to remediate the sediment quality itself. On the other hand, steelmaking slag application to organically enriched sediments of reductive condition from which hydrogen sulfide is generated, either by capping or mixing is markedly more effective than the natural stone capping that was used so far.
5. Conclusion
Following findings were obtained in the present study.
(1) Reduction of dissolved sulfide by steelmaking slag was confirmed in both treatments; capping the sediments with steelmaking slag and mixing the sediments with slag. These slag applications suppressed the diffusion of hydrogen sulfide gas to the atmosphere. By the mixing method, the dissolved sulfide concentration in the interstitial water was significantly low compared to that of capping. An increase of the ORP was also observed, implying that oxidation of sulfides is another plausible cause of lower dissolved sulfide concentrations. The effects of hydrogen sulfide suppression by slag application lasted for the entire experiment period of 6 months.
(2) Compared to the remarkable effects by the slag application, capping with natural stones of similar grain size showed a limited effect; lower release of the dissolved hydrogen sulfide into the overlying water was observed to a certain degree due to physical prevention of diffusion, but the hydrogen gas was not suppressed completely. It can be said that steelmaking slag application is much more effective at lowering the amount of hydrogen sulfide generated in the organically enriched anoxic sediments than the conventional sand/stone capping that has been used so far.
(3) From the surface examinations of steelmaking slag with a scanning electron microscope, it was suggested that FeS could be major formed on the surface of the steelmaking slag after immersion of slag in the sediments containing sulfides.
Based on these results obtained in the laboratory experiments and quasi-field experiments, we can conclude that steelmaking slag will be a useful material for remediation of organically enriched sediments in coastal areas to suppress hydrogen sulfide.
Acknowledgements
The present study was conducted by Comprehensive Cooperative Research Agreements between Hiroshima University and JFE Steel Corporation. In the present study, the Hiroshima Prefectural Government, Public Works & Construction Department, Port and Harbor Division and Eastern Construction Office, Port and Harbor Division gave useful advice, and also permitted the authors to sample the sediments of the Fukuyama inner harbor. Mr. Sadaharu Iwasaki, the Takehara Marine Science Station, Hiroshima University, provided experimental apparatus as well as support for the experiments. The authors wish to take this opportunity to express their thanks to all concerned for their generous assistance.
References
- 1) L. S. Hansen and T. H. Blackburll: Mar. Ecol. Prog. Ser., 75 (1991), 283.
- 2) K. Marumo and M. Yokota: Rep. Mar. Ecol. Res. Inst., 15 (2012), 23.
- 3) J. Sasaki, M. Isobe, A. Watanabe and M. Gomei: Proc. Coastal Eng., JSCE, 43 (1996), 1111.
- 4) K. Iimura, H. Kobayashi and H. Ogura: Blue Tide Outbreak of Tokyo Bay, Annual Report of Chiba Prefectural Environmental Research Center, Chiba, (2010), 1.
- 5) R. Takahashi, M. Yamana and Y. Sato: Symp. Marine Environment, Vol. 9, School of Marine Science and Technology, Tokai University, Shizuoka, (2010), 19.
- 6) Nippon Slag Association: Environment-friendly Material, Iron and Steelmaking Slag, Nippon Slag Association, Tokyo, (2010), 15.
- 7) M. Yamamoto, N. Hamasuna, M. Fukushima, S. Okita, S. Horiya, E. Kiso, M. Shibuya, M. Sadaoka: J. Jpn. Inst. Energy, 85 (2006), 971.
- 8) Y. Miyata, Y. Sato, S. Shimizu and K. Oyamada: JFE Tech. Rep., 19 (2008), 1.
- 9) A. Hayashi, H. Tozawa, K. Shimada, K. Takahashi, R. Kaneko, F. Tsukihashi, R. Inoue and T. Ariyama: ISIJ Int., 51 (2011), 1919.
- 10) K. Ito, W. Nishijima, E. Shoto and M. Okada: J. Jpn. Soc. Water Environ., 20(1997), 670.
- 11) A. Hayashi, T. Watanabe, R. Kaneko, A. Takano, K. Takahashi, Y. Miyata, S. Matsuo, T. Yamamoto, R. Inoue and T. Ariyama: Tetsu-to-Hagané, 98 (2012), 207.
- 12) A. Hayashi, S. Asaoka, T. Watanabe, R. Kaneko, K. Takahashi, Y. Miyata, K. Kim, T. Yamamoto, R. Inoue and T. Ariyama: Tetsu-to-Hagané, 98 (2012), 624.
- 13) K. H. Kim, S. Asaoka, T. Yamamoto, S. Hayakawa, K. Takeda, M. Katayama and T. Onoue: Environ. Sci. Technol., 46 (2012), 10169.
- 14) S. Asaoka, T. Yamamoto, I. Yoshioka and H. Tanaka: J. Hazard. Mater., I72 (2009), 92.
- 15) S. Nakahara, K. Hiraoka, T. Yamamoto and H. Ueshima: J. Jpn. Soc. Water Environ., 35 (2012), 159.
- 16) S. Asaoka, T. Yamamoto, S. Kondo and S. Hayakawa: Bioresour. Technol., 100 (2009), 4127.
- 17) T. Yamamoto, S. Kondo, K. H. Kim, S. Asaoka, H. Yamamoto, M. Tokuoka and T. Hibino: Mar. Poll. Bull., 64 (2012), 2428.
- 18) Japan Fisheries Resource Conservation Association: Water Quality Standard for Fisheries Use (2000 version), Japan Fisheries Resource Conservation Association, Tokyo, (2000), 5.
- 19) Offered by Takehara Marine Science Station, Hiroshima University.
- 20) H. Tubokawa, K. Yasuda and M. Fukue: Symp. Marine Environment, Vol. 9, School of Marine Science and Technology, Tokai University, Shizuoka, (2010), 13.
- 21) M. Fujiwara: The 44th Shiroishi Memorial Seminar, ISIJ, Tokyo, (2001), 1.
- 22) M. Matsuo: Grants for Research by the Tokyu Foundation for Better Environment, 34-250, The Tokyu Foundation for Better Environment, Tokyo, (2006), 1.
- 23) B. Thamdrup, H. Fossing and B. Jorgensen: Geochim. Cosmochim. Acta, 58 (1994), 5115.
- 24) K. Nanba, T. Matsuo and Y. Nimura: Fisheries Sci., 67 (2001), 14.
