The mechanism of hydrogen gas generation of high moisture dri C at temperatures below 100 degrees centigrade

Dr. Oscar Dam G.

PhD University of London. IIMA Technical Committee Chairman

Abstract:

This paper presents a thermodynamic analysis and matemathical model of the mechanism associated with the hydrogen gas release from humid Direct Reduced Iron Fines or DRI C. It should be noted that no such work have been previously published in authorized technical literature. The present analysis aims to clarify the concerns raised over the safe carriage of DRI (C), specifically on the hydrogen gas release mechanism from the wet material and compare the same fenomenon as show by dry DRI.

Key words: Direct reduction, direct reduced iron, HBI, DRI, HRD

1.- INTRODUCTION

The theories regarding the heat generated by the direct reduced iron DRI have focused on the corrosion and heat released by the iron oxidation reactions. DRI A (briquettes) and DRI B (pellets – lumps and or chips) typically were rather dry. IDRI by product fines is a material that normally water passivate due to its origin as well as the way it is stored (open yards and in under rain conditions in most cases). The mechanism at work in the case of DRI C with water is results in the continuous production of hydrogen that when in contact with oxygen across the surface of the stowed material. This paper presents recent findings on the effect cooling effect and gas releasing mechanism of high moisture, self-containing DRI C.

2.- THE MECHANISM

2.1.- Previous Observations

It has been demonstrated that wetted material has a strong resistance to penetration by any gas (1). The material when stowed is very closely packed with virtually no free air or oxygen present. Such free oxygen as may be present will be rapidly, depleted resulting in some temporary heating, but thereafter, there will be no available oxygen and none can penetrate the material. In addition, the noted loss of moisture during a voyage will be the result of evaporation, which requires heat and therefore has a cooling effect on the cargo. Therefore, self-heating does not and almost certainly cannot occur in this cargo.

It also has been noted by means of temperature monitoring of real shipments, that during long haul or sealing periods as for example from Venezuela to China and Korea, the temperature readings in the monitored ship holds, show in Figure 1, and have shown the general trend for temperatures to drop, as show in Figure 2. Althoug several thermocouples are used for monitoring the cargo temperatures, the actual thermocouple E show in Figure 1 is the one used to detect the temperature in the hold open spaces.

2.2.- The Proposed Mechanism

Under such conditions, the previous theories and mechanism must be reviewed. This review involves in a most comprehensive consideration of the reactions occurring at low temperatures. In equation 1 (3), the classic approach is the simple consideration of heat generation and its dissipation to the environment, expressed as:

dT/dt = [ Q Oxidation Reaction] - [Q(losses to the Environment)] (1)

Where (dT/dt) is the rate of increase of the temperature of the material measured by thermocouples A,B,C, and D and the heat released to the environment.

In practice the “excess heat” will be measured by thermocouple E. This equation only is valid at temperatures above 100 degrees Celsius. The “excess heat” according to equation 1 will vary directly with the:

• Directly with the exposed surface area of the material and the oxygen presence

• Indirectly with the heat capacity (Cp) of the material

• Indirectly with the heat transfer coefficient of the system.

To review the application of equation 1 for understanding of DRI C material, new terms related to the low temperature of the sequence for corrosion or galvanic reactions, such as water vapor formation must be included. Then equation 1 needs to be modified in the following way:

dT’/dt = Q Oxidation - ₍₂₎(Q losses + Q Water Evaporation)

The above equilibrium equation 2, is affected by the exposure of the metallic iron to the oxidation reaction producing an oxide layer. Thus the “heat of reaction” gives a resultant increase in the reaction rate related with an increase of temperature and a reduction in the reaction rate by delaying effect of the oxide film that is formed. This term of the equation is not noticeably at 50 ºC for DRI B. In fact only 6 ºC was measured at 120 ºC for the same material.

Based on operational experience, the mechanism for DRI C involves two oxidation reactions:

1. Oxidation of the covering rusty crust in the presence of water and oxygen (applies for stock yards storing and in stowage in cargo holds). For this condition, rusting of the cargo surface, the equation that applies is Eq. 1.

2. Oxidation of metallic iron in the absence of oxygen and the presence of water vapor inside the cargo holds. For this condition, equation.2 is applicable. The condition inside the cargo is water vapor present but very little oxygen, as show in Figure 3 (4).

When the second term of the left hand side of the equation is greater than the heat of oxidation, the resulting dT’/dt<0 and the cargo cools down. On the other hand if lower then dT’/dt>0 the cargo will gets hotter (3).

Thus, by applying heat transfer equation, assuming that the oxidation rate of the metallic iron is delayed by water mass evaporation, where K= Heat evolved by the oxidation of metallic iron, equation 3 is obtained

dT’/dt < K - [Q (losses to the Environment) +Q Water Evaporation)] (3)

From these conditions the chemical reactions considered are:

1. 3 Fe(OH)2 ® Fe3O4+ 2 H2O + H2 - Heat

2. 3 Fe + 4 H2O ® Fe3O4 + H2 + Hat

3. ( H2O) liq. + Heat ®(H2O) vap.

The energy required to heat 1 gr. of water from room temperature to 100 degrees Centigrade is 80 cal per gr. to evaporate it and requires an additional 539 cal/gr extra, which means 619 cal per gr. of H2O will be required for the evaporation of water.

For the case of the reaction of iron to produce magnetite, the heat release at 100 degrees Celsius is about 500 KJ/mol or 9 Kcal/mol, of iron reacted. According to reaction 2 the total heat released for 3 mole of reacted iron is 27 Kcal/gr. Fe. Therefore, for every 1% moisture to be evaporated in a cargo hold of 5000 MT, (about 50 MT of water), about 30950 Kcal, will be provided by 1.15 tons of oxidized iron.

Therefore, if the mass of water moisture to be evaporated consumes most of the heat released by a) the metallic iron oxidation and b) the heat release of the upper rusted surface produced by the oxidation of the cargo, the overall temperature effect to be measured by the thermocouple located in the hold open space will result in a temperature drop

3.- Analysis of DRI C cargoes (5).

To corroborate the above calculations, two cargoes were sampled at the loading and unloading ports, to check the chemistry. The 5,000 metric ton cargoes chemistry variations are show in Table 1.

From Table 1, the following variations can be observed:

• Metallic Fe max.: 2.09%

• Moisture max.: 0.65%

For these values, it is expected that the total heat released by the oxidized iron will be in the order of 2821500 calories and the water evaporation required 20117.5 cal. The remaining heat is retained within the DRI C totaling 2801.4 Kcal during the voyage period. When considering these values for the whole voyage, the amount of reacting iron counts for 62.7 Kcal/day (2.2 Mt/day). This value indicates a very slow rate of reaction.

Under surface ventilation, some moisture is allowed into the hold from marine sea with a average moisture content of 26 gr/m3. This effect adds 0.173 m3 of water that it is expected to react and be evaporated thus representing an additional heat requirement

5.- OTHER RELATED REACTIONS (6,7)

In view of the corrosion mechanism involved at low temperatures, the following reactions need to be considered, because there are competing cathodic reactions

4. H2O + O + 2e = 2OH

5. 2H + 2e = H2 (g)

This, although against expectations has been related with the formation of porous magnetite layer in the presence of high oxygen contents.

A cargo of DRI B under sea water environment, has a maximum evolution rate of hydrogen gas of 0,0l6 lt//kg.hr, whereas in air it is 0.0043 lt/ kg.hr in air, which is 3 times higher (6).

In comparison, the gas evolution rate for DRI C in ship holds is 0.00191 lt/kg.hr maximum, which the order of magnitude is about 10 times lower than DRI B under sea water conditions.

The gas composition of iron and its oxides at any given temperature is expressed by the following equation (7):

4 log (P H2/ P H20) = 1973/ T (ºK) – 2.20

Using this equation, for the range of temperature below 100 ºC, and more specific from 25 ºC to 60 ºC, for the (PH2/PH2O) ratio the following values are obtained 26350 and 5380 respectively (7). This means although a hydrogen atmosphere is equilibrated with iron can reduced the iron oxides in the temperature range 40-60 ºC, as if the oxides were in the range 100-120 degrees C, the rate of the reduction reaction is very slow or not noticeable, in spite of the large specific surface area. In any case, it has also been estimated that only about the 0.1 % of the hydrogen evolved is available for reduction at the assumed “ignition site”. Therefore any other non reacting amount of hydrogen will be released to the environment (7).

6.- CONCLUSIONS

From the above analysis and calculations, it is possible to conclude the following

1. DRI C material must be considered as a nondry water passivated material.

2. In the range of temperature from ambient to 100 degrees C, the evaporation of the DRI C water content, as well as the water corrosion reactions must be taken into account.

3. The active role of the evaporation and the corrosion reactions has a previously unknown cooling effect, by means of the heat consumed by these reactions.

4. These reactions takes place until the temperature of the RI C reaches the threshold reactivity above 100 degrees C when it is expected to react as a dry DRI material. At this point, the classical theories can be applied...

5. Besides reacting with the surface metallic iron, alongside the original moisture, sea water has an extra cooling effect on the DRI C.

7.- REFERENCES

1. Joint Report Trinidad and Venezuela on DRI C. IMO DSC 14 Meeting. London 2009.         [ Links ]

2. J. Castillo. Monitoring of the MV Sea Pearl Shipment. Venprecar Internal Report 2009.         [ Links ]

3. Ciccariello, R. et.al. Sidbec R&D Dept..- Report to IMO Sub-Committee on Containers & Cargoes. 22nd Session, DC XXII/5/9/Add.1.Agenda Item 5. Circulated under Cover of BC XXII/5/9 London.         [ Links ]

4. Dam. O., J. Castillo, I. Montiel and U. Farias.- IIMA Technical Report on The Effect of Sea Water on the Hydrogen Gas Generation. Sept. 20011.

5. Dam, O.-Analysis of DRI C cargoes. Internal Report for Trinidad&Tobago DRI C Shippers. IMO Report London.2009.         [ Links ]

6. Gallagher, W.P.- Hydrogen Evolution from DRI Products Reacting with Water. ISS Transactions. Vol. 1, pp. 69-73. 1982.

7. Birks, N. et.al.- Mechanism in Corrosion Induced Auto-ignition of Direct Reduced Iron. Materials Science and Engineering Department, University of Pittsburgh.         [ Links ]