Regular Article
Effect of Iron Powder on Inhibition of Carbonation Process in Cementitious Materials
2015 Volume 55 Issue 7 Pages 1522-1530
Details
2015 Volume 55 Issue 7 Pages 1522-1530
Addition of a pore blocker to cementitious materials is one of the known measures for protection from carbonation. However, the connected pore network inherent to cementitious materials can make conventional pore blockers ineffective, but these connected pores can be advantageous when iron powder is used. This study found that ionised iron powders can diffuse along the connected pores and fill the neighbouring pores that mostly range in diameter from 0.075 to 7.500 µm, which is equivalent to the range that harmful ionic species can penetrate. Under the severe condition of accelerating carbonation and corrosion-accelerating curing, the replacement of sand with 2% iron powder is found to reduce the total porosity by up to 50% compared to that of the control specimen without iron powder. An electron probe micro-analyser was used to visually confirm the corrosion, ionisation, and diffusion of the iron powder that filled the pores in the cementitious materials.
The steel embedded in the concrete has many advantages. However, the embedded steel can be the major contributor to the deterioration of concrete durability, when chemical reactions (such as carbonation and chloride attack) occur.1)
A passive film formed on the surface of embedded steel protects the steel from corrosion by blocking water and oxygen, making reinforced concrete a semi-permanent durable material. However, both the carbonation process and chloride attack break down this film. The carbonation process occurs from the fact that carbon dioxide penetrates the porous network and reacts with the Ca(OH)2 dissolved in the presence of water, transforming Ca(OH)2 into calcium carbonate and resulting in the low alkalinity of concrete. The failure of maintaining the level of alkalinity (pH ≈ 8.3) causes the passive film to be broken down, allowing water and oxygen to come in contact with the embedded steel. Following this contact, anodic and cathodic electrochemical reactions occur.2) As a result, the now rusted embedded steel expands to around 3.5 times its original size, destroying the concrete and finally causing the end of the service life of the reinforced concrete structure. Furthermore, chloride attack can locally break down the passive film on the embedded steel if a sufficient amount of chloride ions, i.e., above a chloride threshold value, reaches the passive film.3) Once the threshold value has been reached the corrosive process continues, which can then also cause the end of service life of the reinforced concrete structure in question.
From review of the available literature, there are primarily three ways of protecting the embedded steel from corrosion: (1) surface treatment of the steel before it is placed in the concrete; (2) surface coating on the concrete; and (3) the addition of pore blockers. For (1), the use of an epoxy coating and galvanized zinc has been widely used. However, it is very important to note that there already exists a passive film on the surface of steel, and therefore the inhibition of the corrosion should act on this film and not on the bare steel. To overcome this issue, Tae et al.4) used steel containing Cr and found that a Cr content of over 5% by steel weight can resist the corrosion of the steel embedded in concrete in carbonated environments. In other studies, inhibitors can be added to the concrete mix to chemically compete with the chloride ion and delay the rate of corrosion or to avoid corrosion completely. In some cases, the inhibitors can re-passivate the pre-corroded steels in carbonation and chloride contaminated concretes. Calcium nitrites, amino alcohols, and sodium monofluorophosphates are the most commonly used materials to prevent or to cure the corrosion of the embedded steel in concrete exposed to deteriorative environments. Söylev and Richardson2) reported that calcium nitrite is only applicable to non-carbonated concrete that is contaminated with low chloride ion content. Alonso and Andrade5) have observed that the effect of sodium nitrite on carbonated concrete is better than that of calcium nitrite, but the effectiveness decreases when the chloride content increases. For method (2), the inhibitors mentioned above can also be used, but their effectiveness is found to be limited, depending on the level of penetrability. In addition, the surface-applied inhibitors may not be permanent because of their evaporation, for example, as witnessed for the volatile components of amino alcohols.6)
For method (3), however, its effect can be semi-permanent if the harmful chemical factors originate from external sources. Since one of the biggest factors in the carbonation process is the permeability of concrete and, thus, its porous structure,7,8) the pore blockers can be a first line of defence without changing the chemical properties of the concrete. However, little information has been provided in the literature on this methodology, and there have been only a few reports of significant successes. This is probably because concrete is a porous material, where the pores tend to be connected to each other. Thus, the pore blockers find it difficult to fill all of the connecting pores in the concrete or to completely prevent the ingress of the harmful chemical factors arising from the atmosphere.
However, in this study, micro-sized iron powder (to be intentionally corroded) was chosen as a pore blocker, and its potential effect on the prevention of the ingress of harmful ionic species into the cementitious materials was examined. Therefore, the aim of this study is to inhibit the carbonation process and to prevent the corrosion of rebars embedded in mortars.
The hydration of Portland cement is a complex process. The exact chemical mechanisms behind the formation of hydration products and the phases that constitute the hydration products are still not clearly known.9,10) However, there is general agreement that after hydration, there are left a substantial amount of pores where gas, ion, and moisture can transport through the cementitious system over a broad range of conditions.11) In general, the interconnectivity of the pores, whether the pores are connected, is very important for determining the level of ingress of harmful ionic species from external sources.
The ink bottle effect and percolation theory provide detailed information on the interconnectivity of the pores in cementitious materials, as shown in Fig. 1. If the capillary pores are connected (Fig. 1(a)), and interfacial transition zones (ITZs) surrounding the aggregates are connected (Fig. 1(b)), percolation from top to bottom can be achieved. Bentz12) reported that unlike high strength concrete, ITZs in normal strength concrete are thick and percolated. The thicknesses of ITZs in the normal strength concrete are 20 μm.13) Scrivener14) observed that the inclusion of aggregate disrupts the local packing of cement powders, resulting in a high water to cement ratio in these ITZ regions.
Pore network.
The addition of iron powder may produce porous regions (similar to the formation of ITZs) around it. However, after the surface of the cementitious material is carbonated by the presence of carbon dioxide, it is expected that the induced porous regions can be filled by the expansion of the corroded iron powder. Further to this, it is hypothesised that the corrosion products from iron powder would also fill other neighbouring, pre-existing pores. Figure 2 shows a conceptual image of how the iron powder fills the porous regions around the iron powder. The iron powder would undergo (1) corrosion, (2) ionisation, and (3) diffusion. The detailed process is as follows:
Corrosion, ionisation, and diffusion of iron powder.
(1) First, corrosion of the iron powders situated near the surface of the cementitious material occurs after a passive film on the iron powders is destroyed by carbonation. Oxygen and moisture are prerequisites for the corrosion.
(2) This corrosion involves an electrochemical reaction (Fe → Fe2+ + 2e− and ½O2 + H2O + 2e− → 2OH−).
(3) The ionised iron powders, i.e., Fe2+, diffuse via pathways connected to neighbouring pores.
As depicted in Fig. 2(b), in stage (1) and subsequently stage (2), the iron powders would expand and fill (or densify) the pores surrounding the powders themselves. In stage (3), the ionised iron powders would fill the neighbouring pores, which is likely to include the ITZs surrounding the aggregates. Finally, the deployed iron powders at the surface of the cementitious material would act like a real barrier, preventing the ingress of harmful ionic species. This process is confirmed by an electron probe micro-analyser (EPMA), which is discussed later in Section 4.4.
Table 1 shows an outline of the experiment and Table 2 provides the mixture proportions of the mortar specimens. The cement-sand ratio by mass was 1:3, and the water to cement ratio was set at 0.65. For the partial replacement of sand with iron powder, two types of direct-reduced (DR) iron and one type of mill scale (MS) were used for comparison purposes. The DR type is known to be highly susceptible to oxidation and rusting, and the MS type is one of the by-products produced during steel processing and can be used as a raw material in direct reduction plants that produce DR iron.15) The amounts of iron powder used were 0%, 2%, and 4% by mortar volume. The content of iron powder was determined based on the fact that the critical threshold value of the maximum porosity required for preventing the corrosion of embedded steel is known to be around 10 to 15%.16) Considering that corroded iron powder expands to around 3.5 times the natural size1) and the porosity of conventional concrete is around 25%, the selected contents of iron powder (to be corroded) would be sufficient to reduce the volume of the porosity of the specimens prepared in this study to the required level. For the preparation of the specimens, steel reinforcing bars with a diameter of 10 mm were embedded with mortar cover of 20 and 40 mm, as shown in Fig. 3. The specimens were sealed and cured for 14 d, upon which they were de-moulded and air-cured in a constant temperature-humidity chamber at 20 ± 3°C and 60 ± 5% R.H. for 4 weeks, until the experimental tests were conducted.
| Iron powder | Mortar cover (mm) | Tests Conducted | |
| Type | Partial replacement (%) | Depth of carbonation Porosity Corrosion area of rebar Weight loss of rebar EPMA3) | |
| DR-A1) | 0 | 20 40 | |
| MS | 2 | ||
| DR-B2) | 4 | ||
| W/C | Water content (kg/m3) | Mix weight (kg/m3) | ||
|---|---|---|---|---|
| Cement | Sand | Iron powder | ||
| 0.65 | 335 | 516 | 1257 | 0 |
| 516 | 1206 | 157 | ||
| 516 | 1155 | 314 | ||
Section detail of a specimen.
Following the air-curing, all specimens were relocated to the chambers for testing the carbonation of the mortar specimens and the corrosion of the embedded reinforcing bars. Figure 4 shows the overall procedure of the experiments and the times for each test. All experimental tests were conducted with three replicated samples. The average values or representative images of the results obtained from the replicated samples were used for this investigation. The carbonation depth was measured after every 4 weeks curing in an accelerated carbonation chamber, including the interval of every 30 cycles in the corrosion-accelerating curing. For corrosion-accelerating curing, the one cycle consists of 1 day at 60°C and 95% R.H., and 1 day at 30°C and 60% R.H (total 60 days for 30 cycles). The porosity of the mortar and the corrosion area and weight loss of the rebars were measured after completion of the corrosion-accelerating curing of 30 cycles and 120 cycles, including the interval of every 4 weeks in the accelerated carbonation chamber. Hence, for these two tests, total curing age was 12 months (carbonation test: 112 days and corrosion test: 240 days). In addition, to visualise the corrosive procedure of the iron powder filling the pores in mortar, i.e., corrosion, expansion, ionisation, and diffusion, the EPMA was applied to the prepared samples upon completion of the 60 cycles of corrosion-accelerating curing.
Detailed procedure of experimental tests.
The physical properties of the cement and aggregate used in this study are shown in Tables 3 and 4, respectively. The cement used was ordinary Portland cement and the aggregate was river sand. Tables 5 and 6 show the chemical composition and the grading of the iron powders, respectively.
| Density (kg/m3) | Blaine fineness (m2/kg) | Stability (%) | Setting time (min.) | Compressive strength (MPa) | |||
|---|---|---|---|---|---|---|---|
| Initial setting | Final setting | 3 d | 7 d | 28 d | |||
| 3150 | 352 | 0.16 | 207 | 353 | 21 | 31 | 39 |
| Density (kg/m3) | Fineness modulus | Absorption ratio (%) | Unit weight (kg/m3) | Percentage passing 80 μm (%) |
|---|---|---|---|---|
| 2560 | 6.5 | 2.0 | 1577 | 0.3 |
| Constituent/property (%) | DR-A | MS | DR-B |
|---|---|---|---|
| Fe | 97.24 | 75.79 | 94.93 |
| Si | 0.01 | 0.47 | 0.11 |
| Mn | 0.02 | 0.66 | 0.20 |
| P | 0.79 | 1.15 | 0.73 |
| O | 2.21 | 20.86 | 4.16 |
| Cu | – | 0.66 | – |
| Grain size (μm)/property (%) | DR-A | MS | DR-B |
|---|---|---|---|
| Over 150 | 0.8 | 0.7 | 3.8 |
| 106–150 | 29.2 | 2.8 | 28.8 |
| 75–106 | 33.8 | 7.2 | 23.6 |
| 63–75 | 10.5 | 6.4 | 6.2 |
| 45–63 | 11.8 | 17.4 | 16.5 |
| Under 45 | 13.9 | 65.5 | 21.1 |
The accelerated carbonation test was conducted under the conditions of 20°C, 60% R.H, and a CO2 concentration of 5%. Prior to placing the specimens in the chamber, epoxy resin was applied to both sides of the specimens so that only their top and bottom sections were exposed to carbon dioxide. The carbonation depth was examined by spraying a specimen cross-section parallel to the direction of carbonation with a 1% phenolphthalein solution. It can be seen that a pH higher than 9 produces pink stains and a pH lower than 9 leaves no colour, indicating that carbonation occurs (typical of the phenolphthalein solution).
3.3.2. PorosityMercury intrusion porosimetry was adopted for measuring the porosity of the mortar. A sample was prepared by cutting the specimen 10 mm from the top surface, allowing data to be obtained from a carbonated zone. The size of the sample was 5 mm × 5 mm × 5 mm. The measurement was conducted for the size of 0.005–369.000 μm in the prepared sample.
3.3.3. Corrosion Area and Weight Loss Due to CorrosionA convenient method to measure the corrosion area of steel was adopted in the present study. This is because the effect of iron powder on resisting the corrosion of embedded steel in concrete is assumed to be very significant, so that the great accuracy is not required to distinguish each datum. Detailed procedure for measuring the corrosion area and weight loss of embedded steel in the concrete is described as follows.
After the completion of the corrosion-accelerating curing, an embedded reinforcing bar was taken out of a mortar specimen. The corrosion area was determined by tracing the contours of rust on a transparent sheet, blacking out the rust areas, and calculating the corrosion area using software to measure the area automatically. The weight loss of the rebar was measured according to the standard testing method for corrosion of concrete structures and corrosion evaluation method of iron materials in concrete, specified by the Japan Concrete Institute (JCI-SC1).
3.3.4. EPMAA sample of size 20 mm × 20 mm × 5 mm was prepared by cutting the mortar specimen from the surface, as shown in Fig. 5. For scanning an image of the sample, two parts (a carbonated zone and a non-carbonated zone) were pre-separated. The carbonated zone was pre-determined to be between 0 and 10 mm, and the non-carbonated zone was pre-determined to be between 10 and 20 mm. EPMA was applied to the prepared sample with an accelerated voltage of 0.2 to 30 kV and a probe spot mode.
Preparation of a sample for EPMA.
Figure 6 shows the carbonation depth of specimens measured every 4 weeks in the chamber of accelerated carbonation, including an exposure interval to the corrosion-accelerating curing of 30 cycles. In general, the longer the specimens were exposed, the deeper the carbonation depth, regardless of the mortar type. However, the specimens with iron powder significantly delayed the increasing rate of the carbonation depth. It was clear that increasing the content of iron powder decreased the depth of carbonation in all tested specimens.
Depth of carbonation.
The good results for the specimens with iron powder can be attributed to a pore-blocking effect, as previously explained in Section 2.2. The iron powders located at the surface of the specimens corrode and thereby expand, filling the pores of the specimens. This is because the fast penetration of corrosive factors, such as moisture and oxygen, causes chemical reactions with the iron powder and creates the corrosion products, i.e., α-FeOOH and γ-FeOOH. This process with iron powder delays or prevents subsequent reactivity with pre-existing calcium hydoxide from internal sources and with induced carbon dioxide from external sources. Hence, the carbonation of cementitious materials is resisted.
As for the type of iron powder, DR-B was the most effective in reducing the carbonation depth. However, it should be noted that the performance of DR-B and DR-A was very similar, whereas that of MS was rather poor. The MS type, even with a content of 4%, was less effective than either DR-A or DR-B types with 2%. This result implies that the grading of iron powder is not an important parameter (Table 6), rather it is the type of iron powder that is important. The notable results obtained from the DR types is probably because of their high levels of Fe, around 95% (Table 5). This characteristic makes the DR types more susceptible to corrosion than any other type of iron, including the MS.15)
4.2. PorosityFigure 7 shows the relative pore volume of specimens with and without iron powder upon completion of 30 cycles (Fig. 7(a)) and 120 cycles (Fig. 7(b)) in corrosion-accelerating curing conditions. From the literature, it is known that the pathway for harmful ionic species to deteriorate cementitious materials mainly arises from capillary pores with diameters that range from 0.003 to 30.000 μm. Regarding the capillary pores, the specific size upon which the harmful ionic species can penetrate ranges from 0.075 to 7.500 μm.17)
Relative pore volume of mortars.
It can be seen that the iron powder (any type) was effective in reducing the particular pore size of 0.075–7.500 μm, which is critical to prevent the ingress of harmful ionic species, and had little or no effect on reducing the smallest (under 0.010 μm) and largest (over 7.500 μm) pores. Since the size of un-corroded iron powders ranges from 45 to 150 μm (Table 6), the pores that are filled by the iron powder are far finer than this range. Hence, it can be said that the corrosion products of iron powder are diffused across a cement matrix of fine pores.
The results in Fig. 7 also show that the type has a significant effect on the magnitude of pore volume reduction. DR-A and DR-B types were the most effective in reducing the pore volume in the critical range (0.075 to 7.500 μm). This result is consistent with the data previously shown in Fig. 6 and explains the successful results observed for the specimens with DR-A and DR-B.
Figure 8 shows the relationship between the total volume of pores and the depth of carbonation tested in this study. From review of the literature, the carbonation progress of cementitious materials is dramatically delayed or ceased when the total volume of pores is around 10 to 15%. Similar results were also obtained in this study. As can be seen in Fig. 8, the depth of carbonation was significantly reduced as the total volume of pores in the mortar specimens with iron powder decreased.
Relationship between total porosity and carbonation depth.
Figure 9 shows the corrosion area of the rebars taken out of the mortar specimens with and without iron powder upon completion of 30 and 120 cycles under corrosion-accelerating curing conditions. The effect of mortar cover, i.e., 20 mm (Fig. 9(a)) and 40 mm (Fig. 9(b)), on the corrosion of the rebars is also presented. The corrosion area of the rebars, Fs, is calculated as follows:
| (1) |
Corrosion area of rebars.
Upon completion of 30 cycles, and with a mortar cover of 20 mm (Fig. 9(a)), corrosion of the rebars occurred in the control mortar specimen and the mortar specimen with MS2% and MS4%, whereas other specimens with DR-A2%, DR-A4%, DR-B2%, and DR-B4% were safe from corrosion. In particular, with iron types DR-A4% and DR-B4%, the rebars were fully protected from corrosion, even after completion of 120 cycles. For the mortar specimens with a cover of 40 mm (Fig. 9(b)), the corrosion of the rebars only occurred in the control mortar, while no corrosion was observed for the specimens with any type of iron powder.
By comparing the results after 120 cycles of corrosion-accelerating curing (Fig. 9(a)), it can be seen that there is a large difference between the specimens with and without iron powder. This means that partial replacement with iron powder significantly delays the corrosion rate of rebars embedded in mortar specimens. Further supporting evidence can be found below.
Figure 10 shows the relationship between the porosity of mortars with and without iron powder and the corrosion area of the rebars embedded in the mortars upon completion of 30 and 120 cycles under corrosion-accelerating curing conditions. The data shown is from the results of mortars with a cover of 20 mm tested in this study. It is clear that the porosity and corrosion areas are highly correlated.
Relationship between total porosity and corrosion area.
It is true that a corrosion rate is a function of the total pore volume in concrete. The higher the volume of pore, the higher is the corrosion rate. As found in this study, the addition of iron powder in concrete decreases the pore volume and decreases the corrosion area. The level of decreased corrosion rate is dependent on the type of iron powder, and it is believed that θ can give quantitative information on this level. For example, in Fig. 10, tan θ is a function of the pore volume and the corrosion area (i.e. tan θ = increased corrosion area/decreased pore volume). For a given content of iron powder, the lower the value of tan θ, the higher is the level of the decreased pore volume and the lower is the level of the increased corrosion area. The value of tan θ can be determined by the iron types with different Fe contents and different grain sizes. In Fig. 10, tan θ of all mortar specimens with iron powders was around 1 or 0 (DR-A2% = 1.07, DR-A4%=0.00, MS2%=1.16, MS4%=1.00, DR-B2%=1.10 and DR-B4%=0), whereas that of the control mortar without iron powder was 5.8. Hence, tan θ is an important parameter for developing a numerical modelling and needs to be explored to predict the effect of iron powder on the corrosion area of concrete in the future study.
It is noted that the slight reduction in the porosity in the control mortar specimen is because carbonated material is 3–5% denser than un-carbonated material, depending on the type of aggregate and on the binder to aggregate ratio.18)
4.4. Weight Loss Due to CorrosionFigure 11 shows the weight loss of the rebars due to corrosion. The weight loss of the rebars, WF, is calculated as follows:
| (2) |
Weight loss of rebars due to corrosion.
In Fig. 11(a), after the completion of 30 cycles with DR-A and DR-B types of 2 and 4%, the rebars fully retained their initial weight before being exposed to the corrosion-accelerating curing condition, indicating that no corrosion occurred. However, without iron powder and with MS2% and MS4%, the rebars showed a weight reduction of up to 0.4%. At completion of 120 cycles, however, it can be seen that for all types of iron powder, the loss rate of the rebars was significantly delayed compared to that without iron powder. The rebar without iron powder lost 1.6% of its initial weight, whereas with the other types of iron powder, the rebars only lost a maximum of 0.5%. In fact, for the 4%-types of DR-A and DR-B, the rebars achieved no loss in weight.
For the mortar cover of 40 mm (Fig. 11(b)), on the other hand, all rebars fully retained their weights at the completion of 30 cycles, but at the completion of 120 cycles, the weight loss was only observed in the rebar without iron powder.
4.5. EPMAFigure 12 shows the corrosion, ionisation, and diffusion of iron powder in a carbonated zone compared to those in a non-carbonated zone, while Fig. 13 shows the effect of iron powder on filling the pores at the same location shown in Fig. 12. In Fig. 12, the Fe in all images are mapped in colour. The colours closer to white represent a denser percentage of Fe, and a blacker colour represents a smaller percentage of Fe such that the iron that is bluish in colour (Fig. 12(b)) can be said to be corroded.
EPMA images, with mapping of Fe.
Effect of iron powder filling pores due to corrosion, Aggr. = aggregate.
In Fig. 12(a), it can be seen that the iron powder and aggregate are mixed and well distributed. This image describes the iron powder before it is corroded, i.e., that which is relatively far from the surface (between 10 and 20 mm), where the carbonation of the mortar specimen has not yet proceeded. Hence, this zone would retain a high alkalinity and the passive film on the iron powder would still be present. However, when the mortar specimen loses the high alkalinity due to carbonation, the passive film is destroyed and the iron powder starts to react with the oxygen and moisture, resulting in the corrosion of the iron powder. In Fig. 12(b), it can be seen that due to the aforementioned process arising from corrosion, the ionised iron is well diffused along the pre-located pathways that characterise a typical cementitious system, as previously discussed in Section 2.1. Consequently, this process sporadically fills the pores located in the cementitious material, as can be seen in Fig. 13. Pores of various sizes are formed around iron powder and aggregate, and also located in a matrix of cemente paste (Fig. 13(a)), but when the iron powder is corroded, most of these pores are filled by the corroded, ionised, and diffused iron powder (Fig. 13(b)).
It is believed that the corrosion, ionisation, and diffusion of iron powders would commence from the surface of the cementitious material and progress towards deeper sections of the material until the level of porosity reaches the threshold value (10 to 15%) and fully resists the ingress of harmful ionic species.
Further, it can be imagined that the passive film on embedded rebars can be destroyed by internal sources such as chloride ions that exist in the constituents of cementitious materials. However, oxygen and moisture are the prerequisite sources for corrosion. Therefore, although the passive film on the rebars is destroyed, it is still possible that partial replacement with iron powder can delay the corrosion rate of the rebars by blocking the ingress of such harmful prerequisite sources from the atmosphere. In addition, by preventing the ingress of harmful ionic species, this method is particularly useful for improving the durability of cementitious materials that contain a high water to cement ratio (without changing the chemical properties of the materials).
In this study, micro-sized iron powder was chosen as a pore blocker for cementitious materials to prevent the ingress of harmful ionic species from external sources. The following conclusions are drawn:
(1) Partial replacement with iron powder has a significant effect on the inhibition of carbonation and in preventing the corrosion of rebars. This is mainly because the iron powder is found to significantly reduce the porosity of a particular size, ranging from 0.075 to 7.500 μm, in which harmful ionic species from external sources can penetrate. Hence, an iron type that can reduce this porosity range, such as a direct reduced iron (DR type) that is susceptible to corrosion, will be the most helpful.
(2) Images obtained from an electron probe micro-analyser (EMPA) confirm that when carbonation of a cementitious material commences from the surface, the iron powders undergo corrosion, ionisation, and diffusion, and this process leads to filling, not only the surrounding pores themselves, but also the neighbouring pores that are far finer than their initial size before corrosion.
(3) The corroded, ionised, and diffused iron powder reduces the total porosity of cementitious materials.
This work is part of a study supported by JFE Steel Corporation. It is also supported by a grant (391) from Infrastructure and transportation technology promotion research Program funded by Ministry of Land, Infrastructure and Transport of Korean government.
