Coking Pressure Phenomena and its Influencing Factors


Coking Pressure Phenomena and its Influencing Factors

Coking pressure is a phenomenon which has become important because of the use of the double-heated wall, vertical, slot-type coke ovens. In the round beehive ovens as well in the heat recovery coke ovens, which are also being used for coke production, the coal can freely expand upwards and thus the swelling of the charge is accommodated by this free expansion. On the other hand, in the slot-type coke ovens, the expansion of the coal horizontally to the heated wall is restricted. There are several cases of premature failure of oven walls during the coal carbonization process.

The erection of the new, larger and taller coke ovens has been accompanied by undesirable occurrences of distorted walls due to the coking pressure resulting in several studies regarding the expansion behaviour of coal during carbonization. The efforts have been focused on developing a reliable test so that coal blends can be tested for safety prior to their use in the coke ovens.

Development of coking pressure

During carbonization process, coal passes through the plastic stage and volatile matter (VM) evolves during and, to a lesser extent, after that stage. It is normally accepted that coking pressure arises in the plastic stage. In a coke oven chamber, two vertical plastic layers parallel to the heating walls are formed from the beginning of carbonization. As the carbonization proceeds these layers move towards the centre of the oven. At the same time, similar horizontal layers are formed at the top and bottom of the charge. These are joined with the two vertical layers and the whole forms a continuous region that surrounds the uncarbonized coal and it is usually referred to as the ‘plastic envelope’.

The permeability of the plastic layers is low in the case of coking coals. Hence, with a strongly plastic coal, there is a possibility that a pressure can set up within the cool zone of unconverted coal. The sum of the internal pressure within the plastic layers and between them is transmitted through the coke layers to exert a pressure on the walls.

When the plastic layers meet at the oven centre, the central mass of coal receives heat from both sides simultaneously whilst the temperature of the coal is being raised through-out the plastic temperature range. This results into acceleration in the coking rate thus leading to a faster rate of gas evolution with greater resistance to its escape. This then results in peak values in both internal pressure and wall pressure.

Since the plastic layers are also formed parallel to the oven doors and hence a complete continuous plastic envelope exists around the unconverted coal as soon as the coal is charged. Therefore, the VM evolved from the cold side of the plastic layer is entrapped within the envelope. At the time when the layers are about to meet, the rate of heating increases and so the amount of gas evolved increases which results in an increase in pressure. After the re-solidification into coke, a rapid decrease in pressure occurs due to the disappearance of the envelope.

As per another explanation, there is not sufficient heat for the formation of plastic layers parallel to the oven doors. Hence, a plastic sleeve (tube) is formed from the two principal plastic layers and the two secondary ones, which is roughly rectangular but widens near the doors because of the heat losses (Fig 1). The coking pressure results from the pressure of the gases within the plastic layers and it is transmitted to the walls via the coke and semi-coke already formed.

Fig 1 Schematic representation of the plastic zones

The gases which are evolved on the side nearest to the charge centre contain tar which condense and then revapourize as they are captured by the plastic layer. The impregnation of the coal by condensed tar modifies the viscosity of the plastic layer. The internal gas pressure depends very much on the nature of coal and the carbonization conditions. It appears that it is connected to the established equilibrium between the swelling of the plastic layer, the contraction of the semi-coke, and to a certain extent to compression of the uncarbonized coal.

The main phenomenon is that of the internal gas pressure within the plastic layer while the wall pressure results from the transmitting of the gas pressure through the semi-coke and coke to the walls. The magnitude of the internal gas pressure is likely to depend on the rate of evolution of the gaseous matter within the plastic layer and the resistance of the plastic layer to its flow i.e., the permeability of the layer.

However, the rates of devolatilization in the plastic temperature range are higher for the coals with high VM content and yet these coals do not give high pressures. Hence, the permeability can be the most dominant factor in the development of internal gas pressure. This is since the maximum rate of VM evolution for high rank coals is near the re-solidification temperature. In fact, there is a relationship between the maximum wall pressure and the rate of evolution of the remaining VM near re-solidification temperature.

During an investigation using a single-heated wall oven it has been observed that the permeability is minimum in the centre of the plastic layer and maximum in the product of re-solidification. The permeability is considered to be due to two contributions. The first is that of (i) the open porosity existing in the plastic zone, (ii) ultra-micro-porosity permitting the molecular flow, and (iii) macro-porosity permitting the viscous flow. The second contribution comes from the flow which follows the bursting of the devolatilization bubbles. This phenomenon is because in the region of the plastic layer adjacent to the semi-coke the bursting of the bubbles prevails over their formation while in the centre of the layer the opposite occurs. Hence, there is a transfer of mass from the centre of the plastic layer to the zone adjacent the semi-coke established by the maximum in porosity in the centre of the plastic layer.

In the case of coals which give high pressure, it has been observed that this maximum is replaced by a minimum in the coal zone adjacent to the plastic layer. This is because in these coals the second contribution in permeability i.e., the bursting of bubbles is low and this assumption is supported from the absence of a maximum in porosity in the centre of the plastic layer. Since the escape of gases is limited, the pressure in the area of the plastic layer increases and brings about a compaction of coal which results in the observed minimum of porosity.

Early studies for coking pressure

Laboratory tests were first conducted in order to ascertain the safety of coal blends. During these tests, a small quantity of coal, placed in a small crucible, was heated from one side either at constant pressure or at constant volume. A test developed in 1920 is considered as the first attempt to determine the swelling pressure generated by coals. In this test, 100 grams coal was heated in a vertical, cylindrical iron crucible under standard conditions, in a gas-fired furnace. A perforated piston was placed on top of the charge. The movement of piston was recorded appropriately, and the significant measurement was the change in the volume expressed as a percentage. This test was further developed in two directions and resulted on the one hand in the constant pressure test and on the other hand in the constant volume test.

In the modification of the constant pressure test, 80 grams of air-dried coal was placed in a steel crucible. A load of 1 kg/sq cm is applied to the coal charge and both expansion and contraction of the charge were recorded as changes in the volume. In the modification to the constant volume test which was carried out at constant volume, 120 grams of air-dried coal was carbonized in a steel crucible and the pressure required to keep the coal at constant volume was recorded.

Further, a number of larger scale tests which used several kilograms of coal were developed. Their concepts were similar to the smaller scale tests. A representative example was the large constant pressure test which differed from the smaller test mainly in size and had a charge capacity of around 5 kg. Only pressures generated in the coal charge greater than 0.08 kg/sq cm were measured. This pressure was regarded as the maximum permissible limit on a coke oven wall. Another large laboratory test was the ‘sole-heated’ oven test. In this test, a charge of around 35 kg was heated unidirectionaly from the sole and was subjected to a pressure of around 0.14 kg/sq cm. The movement of the upper slab was followed by means of a cathetometer.

Since heating in the above described tests was single-sided, the phenomena remained quite different from those occurring in a large oven. Due to the difficulty of recording the wall pressure in a coke oven, early investigators felt that they should try to simulate the carbonization conditions occurring in a commercial oven as closely as possible by adopting double-wall heating. Hence, a large-scale oven chamber in which both heating walls were movable was designed. By fixing one wall the pressure exerted on the other wall during carbonization could be measured with the aid of a hydraulic system. This oven was considered to be the prototype of the movable-wall oven allowing a direct measurement of the coking pressure.

Several movable-wall ovens exist throughout the world. All designs are based on heating from two sides. One of their walls is mounted on a trolley so it is free to move and the coking pressure developed during carbonization is expressed as a force exerted on the wall and is measured by means of a suitable device and is called the wall pressure. They are either gas or electrically heated and take coal charges ranging from 250 kg to 500 kg. As in the full-scale ovens, two vertical, plastic layers are formed, during carbonization, and progress gradually towards the oven centre where they get merged.

In one of the oldest designs of movable-wall oven, around 250 kg of coal was carbonized in a coking chamber which measured 300 mm x 700 mm x 1100 mm. One wall was mounted on rollers and was equipped with a lever weight system for measuring the pressure developed against the wall. This oven was used to study the coking pressures developed by American coals. 300 tests were conducted in which the coal was crushed 80 % below 3 mm and charged with low moisture content with the bulk density (BD) ranging from 785 kg/cum to 850 kg/cum.

The curves obtained by plotting pressure against coking time were categorized into six types according to several characteristics of the tested coals. Fig 2 shows examples of each of the types of pressure curves. Pressure curves of types 1 to 3, are all for those coals or blends which are normally safe to use in commercial ovens. Coals which give curves 2 or3 may give higher pressures when charged at higher BD. Types 4 to 6 are pressure curves of coals which are dangerous to use in the coke ovens.

Fig 2 Pressure curves of tests of different coals in movable-wall test oven

A larger 400 kg movable-wall oven was used for an extensive investigation of the phenomena of coking pressure at the Centre de Pyrolyse de Marienau (CPM). In this oven, the two walls were of corundum construction which allowed operation at higher temperatures and offered more strength and resistance to thermal shock. The force exerted on the movable wall was measured by a strain gauge balance which was mounted at the exterior of the wall at the geometric centre of the useful surface of the heating wall. During the various tests in this oven, four main types of coking pressure – time curves which were the most frequently encountered, were made and are shown in Fig 3.

Fig 3 Main types of coking pressure curves

By analyzing the form of these curves it was concluded that these resulted from the super-imposition of two basic curves, each one of them corresponding to one of the two stages. The first stage reflects the formation of the plastic layers parallel to the oven walls and their movement towards each other, and the second stage reflects the coalescence of the plastic layers at the ovens centre. Curves of types 1 and 2 were given by coals developing high pressures. Type 3 was characteristic of the stamped charged coals whatever the magnitude of the pressure, while type 4 was recorded for coals giving low pressures. Coals which gave high pressures showed similar type of curves.

In the case of coals with low VM, the pressure had risen regularly from the start to the maximum. The other type of curve typical of high coking pressure had shown a rapid rise in the pressure during the first hour. Then the pressure remained relatively constant during the course of the carbonization. When the plastic layers met a further increase occurred until the plastic coal disappeared.

Another approach to the issue of coking pressure was the measurement of the pressure of the gas at the centre of the plastic layer. These measurements were made both in the movable-wall and industrial ovens thus allowing comparisons to be made between the two sets of results. The gas pressure was measured by means of tubes introduced through the  holes either in the oven doors or in the lids of charge hole.

There were several studies of simultaneous measurement of gas pressure and wall pressure in test ovens. In these studies it was found that the maximum gas pressure in the centre of the charge is related to the peak wall pressure. The gas pressure increased with the distance from the oven wall and attained its maximum value in the centre of the charge at the time of meeting of the plastic layers. This maximum gas pressure usually coincided with the peak of the wall pressure and was always greater. Different values were given for the ratio of gas pressure to the wall pressure in different studies.

In one of the studies it was noticed that as the carbonization proceeded and as the plastic layer moved, there was a sudden rise in the gas pressure. When the plastic layer moved beyond the point the gas pressure dropped quickly and that probe did not record again any gas pressure. When the plastic layers met at the oven centre, the recorded gas pressure was higher than the pressures at other points in the oven. The ratio of gas pressure to wall pressure in this experiment varied from less than 1 for low pressures to upto 3 for high pressures.

In an another investigation the influence of both charging and carbonization conditions on the internal gas pressure had been studied and it was found that internal gas pressures was influenced by the same variables as the wall pressures.

In another investigation at the CPM, two observations were made They were (i) two plastic layers were formed at the beginning of the carbonization parallel to the sole and the roof and progress towards the centre of the oven, and (ii) no plastic layers parallel to the doors were formed. For the ratio of gas pressure/wall pressure a value around 0.5 was found.

The differences in the ratio of maximum gas pressure and peak pressure reported in various studies were attributed to differences in the test equipment and procedure. In the movable-wall oven although coking mainly proceeded from the heating walls simultaneously considerable coking proceeded inwards from sole, roof and questionably from the doors and this reduced the area of the plastic layers which met at the oven centre.

The CPM proposed a formula for the time of meeting of the plastic layers at the oven centre. The formula was Pw/Pi = Si/S, where Pw was the wall pressure, Pi was the internal pressure, S was the lateral surface of the charge, Si was the area of the projection of the plastic layer on the central plane at the time of the wall pressure peak. Pi was greater than Ps and these pressures were the internal and wall pressure at the same time. Si was evaluated by discharging test ovens at the moment when the two principal plastic layers joined together. The ratio k = Si/S by definition was 1 at the moment of charging, during the course of carbonization it decreased due to the end effects and it became zero after the coal was re-solidified. In a full-scale oven the end effects were small compared to the height of the wall so that the ratio Si/S was around 1 and coking wall pressure was to be equal to the centre maximum gas pressure. It was suggested that if a blend carbonized under certain conditions in the test oven produces a wall pressure P then the pressure to be expected in a large oven working under the same conditions is to be of the order of 2P.

Setting of safety limits.

Several studies carried out in movable-wall oven and full-scale ovens were aimed essentially to control the phenomenon of coking pressure and protect the ovens from damage caused by excessive pressures. For an assessment of a coal blend, the movable-wall oven test is generally employed and the resulting maximum wall pressure is used to classify the coal blend as safe or dangerous with comparison to previously established limits. The BD of the blend in the test oven is considered extremely important. It is to be at least equal to the average of the full-scale oven and preferably somewhat higher.

Whether a pressure is excessive or not, depends not only on what pressure is exerted but also on what pressure the oven wall can withstand. Hence, many efforts have been made to assess the strength of the coke oven walls.

The wall strength requirement is governed largely by the peak unbalanced coking pressure which is exerted on the walls during the carbonizing process. These unbalanced pressures cause wall bending in the horizontal direction which is required to be stabilized by the vertical gravity loading, including the weight of the roof and the wall, because the joints in the wall have no consistent tensile strength.

A study has been carried out in a specially built coke oven wall which has been subjected to lateral pressure from a hydraulic press. It has been shown that how low is the resistance of the coke oven wall against lateral pressure. Initial cracks appeared in the wall when the pressure has been around 0.09 kg/sq cm and the rate of bulging is increased rapidly as the pressure is raised to 0.13 kg/sq cm. On the basis of these results from a cold wall, it has been suggested a very Iow safe limit of 0.07 kg/sq cm. A very Iow limit restricts flexibility in choosing coal sources, coal blends and carbonization conditions. Thus, after comparing the results of several hundred coals carbonized in the movable-wall oven and, taking into account the behaviour of these coals in commercial ovens, the following safety limits for coals carbonized in coke ovens has been established.

  • Coals developing a pressure greater than 0.14 kg/sq cm are dangerous.
  • Coals developing a pressure greater than 0.11 kg/sq cm can be dangerous when carbonized regularly in ovens taller than 3 metres (m).
  • Coals which give pressures less than 0.11 kg/sq cm are safe.

On the basis on their work in 1948 and 1952, British Coke Research Association (BCRA) concluded that a blend of coals was safe if the pressure which was developed in their test oven was less than 0.14 kg/sq cm. This limit was confirmed by further work done at BCRA in 1956. In this work they charged several blends to the commercial ovens which were at the end of their working life and to a movable-wall oven. They measured the movement of the walls in the full-scale ovens by special apparatus and the walls were examined during the course of the test and after the battery had been cooled down to determine the movement and ascertain the damage. They suggested that an elastic deflection of commercial oven walls up to 0.13 mm could take place without the appearance of cracks. Beyond that point further deflection did not occur readily and cracking took  place.

In the 1960s and 1970s, the construction of tall (6 m or more) coke ovens became prevalent. These coke ovens were operated under the assumption that coking pressures under 0.14 kg/sq cm were safe. The result was that in some cases these ovens sustained serious, early refractory damage. It was reported that a 6 m tall battery had suffered progressive damage and had to be shut down after less than 5 years of operation. The investigations to determine the causes of the premature failure included a structural analysis of a 6 m wall.

A mathematical study was carried out of a 6 m oven wall subjected to pressure from one side. From this analysis the unbalanced lateral pressure which could cause collapse was calculated to be just above 0.12 kg/sq cm. By taking into account the recommended live load factor of 1.7 as well as serviceability relative to cracking, it was recommended that the allowable unbalanced lateral pressure was not to exceed 0.07 kg/sq cm.

Factors influencing coking pressure

From the early 1950s, the movable-wall oven and other similar-sized pilot ovens were being used to evaluate coals for coking. Pilot scale testing is cheaper than full-scale oven testing and generally provides good guidance regarding the behaviour of coals on the larger scale. Some reassurance comes from comparing internal gas pressures generated in the movable-wall and full-scale ovens. Most of the studies done concerning the factors influencing coking pressure have been done using pilot-scale ovens.

Many factors have been found to affect the magnitude of coking pressure. They can be categorized into three broad categories specifically (i) inherent characteristics of the coal, (ii) coal preparation and physical properties, and (iii) oven operating conditions.

Coal – Early studies have shown that dangerous pressures have been encountered when charging coals with VM content on the dry ash-free basis between 16 % and 30 % and never with coals with lower or higher VM content. It has been found that the strongly expanding coals mainly consist of bright coal while those with a higher proportion of dull coal show less expansive force. Hence, from petrographic investigations some conclusions can be drawn as to the degree of expansion of coals.

A series of experiments has been conducted by the Bethlehem Steel Corporation on the influence of coal composition on coking pressure using an 18 inch (457 mm) test oven and a large variety of single coals and coal blends. It has been found that a general relationship exists between coal rank (as shown by vitrinite reflectance) and coking pressure (Fig 4). From the Fig 4, it can be seen that high coking pressure tends to be generated by some, but not all, low VM coals with reflectance greater than 1.35 %. Further it has been found that the coking pressure not only increases with an increase in coal rank but also increases with decrease in inert content.

Fig 4 Relationship between coal rank and coking pressure

Coal rank and inert content have a synergistic interaction. At any given level of rank, coking pressure tends to be lower if the coal has a high inert content. When only the effect of the low VM coal is taken into account, the pressure generated by the coal blend increases with an increase in reflectance of the low VM coal and this increase accelerates when the reflectance of the low VM coal goes above 1.65 %. However, knowledge of the rank does not permit the assessment of the danger of coking pressure from a coal with any degree of certainty. It can be stated that between 18 % and 25 % VM the danger is great and between 25 % and 28 % VM it is still there though to a lesser degree. Low VM coals with reflectance greater than 1.65 % and low inert content produce high pressures whether coked alone or in blends.

Oxidation of coal – Oxidation of coal reduces the maximum plasticity and the plastic zone narrows. It has been found that oxidation of some coals initially increases the coking pressure but thereafter there is a fall in dilatation and an abrupt reduction in coking pressure. At the same time the index M10 increases.

Fourier transform infrared (FIR) studies indicate that in the early stages of oxidation, the main functional groups formed are carbonyl and carboxyl. At higher degrees of oxidation, evidence indicates a significant increase in ether, ester and phenolic groups. The loss of plastic properties on oxidation is attributed to the formation of ether and ester cross-linkages.

The effect of air oxidation of three Spanish coals was investigated. From the results, it has been concluded that, for coals which were characterized as ‘dangerous’ by the test, the oxidation increased their dangerous character to a maximum at a certain level of oxidation after which it decreased sharply. For a coal characterized as safe, there was no appreciable effect from air oxidation.

In a study at CPM of the influence of oxidation on the low permeability between 500 deg C and 600 deg C exhibited by high rank coals has shown that the permeability has increased with oxidation. This has been attributed to the increased open pore volume in the oxidized coals. There is a general agreement that it is difficult to reduce the coking pressure by oxidation without simultaneously impairing coke quality while moderate oxidation may increase the pressure.

Studies of coal blends – A series of experimental studies have been made on the behaviour of coal blends and their coking pressure. The blends have been charged by gravity at a moisture content of 3.5 % and simple crushing to 80 % less than 2 mm. From the results the following has been concluded.

  • Coal giving a medium coking pressure when charged alone affects the coking pressure of a dangerous coal almost linearly as a function of the blend composition.
  • High VM coals which give no coking pressure decrease the pressure more rapidly than in proportion to the amount added. So they are more effective in reducing the pressure. The effectiveness of such coals is greater the lower is their rank.
  • Semi-anthracite (low VM) behaves differently. If it is crushed with the blend the effect on lowering the pressure is marked but there is also a marked deterioration in coke quality. On the other hand if it is crushed separately (95 % to less than 1 mm) an addition of 15 % is sufficient to decrease the pressure and the coke quality is only slightly reduced. Semi-anthracite which is an inert increased the viscosity of the coal during fusion and this impairs the cohesion of the coal. It also reduces the contraction of the semi-coke while with the addition of high VM coals the contraction is increased.

It has also been observed that the coal blends give pressure higher than that generated by individual coals when charged alone. This occurs when a slightly fusible, low VM coal is associated with a very fusible coal. Then the fluid conditions allow the dangerous character of the low VM coal to be expressed.

At BCRA, it has been found that, as regards the development of swelling pressures, the behaviour of a binary blend depends largely on the character of the coal which is in excess in the blend. From their studies on the effects of blending the following has been concluded.

  • Coking pressure occurs only with coal blends with VM less than 24 %.
  • 30 % addition of weakly coking coal considerably diminishes the dangerous properties of a low VM coal.
  • In blends of high/low VM coals, a decrease in pressure occurs on increasing the portion of high VM coal.
  • In ternary blends with similar VM content, the addition of 15 % of an almost non-coking coal prevents the development of any coking pressure.

Effect of pitch addition – The decrease in availability of coals of high rank has led to an increased interest in the use of pitch additives in coal blends for the production of metallurgical coke. Pitch can behave as bridging agent and improve the strength of the resultant coke. The pitch (i) modifies the plastic state and this modification is associated with hydrogen (H2) transfer reactions involving movement of H2 from the pitch to free radicals thus stabilizing the plastic state, (ii) widens the plastic layer, and (iii) increases the evolution of VM. The addition of pitch to a coal tends to increase the coking pressure but the magnitude of the effect depends on the nature of the coal.

Effect of inerts – It has been found that relatively small proportions of inerts suffice to reduce considerably the pressure of a coal, while the particle size of the inerts has a very large effect. The inerts can have several effects namely (i) a diluent effect on the coal, since the inerts do not change in volume, the space available for the coal to swell increases, (ii) particle size effect meaning that if the size of the inert is finer than the coal it increases the average fineness, thus decreasing the pressure, (iii) specific action since the inerts absorb a certain amount of tar and bitumen, reducing the fluidity and swelling of coal, the finer and more porous are the inerts the greater the effect, (iv) modify the contraction of the charge after re-solidification, (v) inerts may increase the permeability of the plastic layer.

At BCRA, a study was made regarding the addition of coke breeze on coking pressure. Normally the average wall pressure decreases with increasing proportion of coke breeze. In another series of tests, with upto 30 % coarse breeze, it has been found that the maximum internal gas pressure decreases with the addition of more than 5 % coke breeze while the maximum wall pressure is not appreciably reduced until more than 10 % is added.

A patent for reducing the coking pressure suggests the addition of 2 % to 8 % of flakes formed from the sawdust or other inert materials. The action of flakes is considered to be the disruption of the plastic layers in the oven charge, thus providing the necessary passage for the gases.

Effect of oil addition – The addition of oil in coal blends, changes the BD at constant moisture content and decreases the pressure. The added oil acts as a diluent and a reduction in pressure can be achieved without damaging the mechanical properties of the resultant coke. It has been observed that the addition of anthracene oil decreases the wall pressure of coals considerably while the maximum in the pressure is observed at a very low temperature, less than 300 deg C. This has been attributed to the ‘balloon effect’, i.e., the volatiles from the oil formed at low temperatures become trapped between the advancing plastic layers and their accumulation leads to the development of pressure.

Effect of moisture – An increase in the moisture in the charge decreases the pressure by decreasing the BD. The use of wet charges makes the carbonizing process very uneven and highly disturbed. The steam acts on the coal particles affecting their ability to absorb the liquid pyrolysis products. The moisture breaks through the plastic zone at different points and so by travelling along the chamber wall reaches the gas free space. The result is a highly deformed plastic layer.

The joining of the plastic layers is spread chronologically and locally over individual sections. So the pressure on the wall manifests itself only in a reduced form. On the other hand during carbonization of preheated charges the plastic layers formed converge with parallel faces to the heating walls and at the same time practically join together over the entire length and height of the chamber thus the pressure on the wall attains its maximum value.

During carbonization in slot type ovens, the moisture content across the charge rises steadily, in comparison with the moisture of the original charge, as moisture is distilled towards the oven centre. .

Several investigators measuring internal pressure in the full scale ovens recorded a peak pressure coinciding with the temperature reaching 100 deg C at the oven centre. This peak is called the water or steam peak and is attributed to the fact that the steam can no longer condense in this region of the oven and must be expelled. Eventually it becomes trapped within the plastic envelope.

In a study, it has been found that the average water peak pressure depends only on the coal moisture and it decreases as moisture is increased suggesting that the BD is the important factor.

Effect of bulk density – BD is regarded as the most important of the variables affecting coking pressure. By increasing the BD, a safe blend can develop very high pressures. In several studies, it has been proved that the danger of damage to the oven increases with greater charge densities. With the erection of large ovens, the influence of BD of coal on the degree of expansion become more important because the coal charge become denser on account of the increased height of the fall during charging.

In several studies on the influence of BD on coking pressure, it is seen that a lack of agreement exists regarding the nature of the functional relationship between test oven wall pressure and BD. In some of the studies it is claimed that the logarithm of wall pressure with BD gives a straight line with slope, while in the other studies it is claimed that this relationship is better represented by a curve. In another it has been shown that at densities above 800 kg/cum (dry), the rate of increase in pressure with increasing BD is enhanced.

In a study at BCRA, it has been shown that gas pressure also displays the same variation with BD as wall pressure. It has been concluded that a given change of BD produces a given change in pressure irrespective of the method employed for BD control.

In a more recent work on the influence of BD on coking pressure, where a single blend has been used with different bulk densities, it has been seen that the variations in BD have been achieved in different ways such as oil addition, preheating treatment, variations in moisture content.  In this study it has been found that BD has a large influence on the coking pressure and it is probably the most important factor affecting the pressure developed by any coal or blend charge.

Effect of particle size – When coal particles differing in size are heated the conditions for the transport of the gaseous pyrolysis products from the middle of the particle to the surface differs. The larger is the particle size, the greater has the pressure developing inside it as a result of the pyrolysis of the organic mass. This promotes condensation which exerts a major influence on the subsequent thermal changes of the coal substance. It has been found that the initial softening temperature increases and the plastic temperature range becomes smaller as the particle size decreases but the viscosity of the plastic mass increases.

It is difficult in practice to analyze separately the influence of BD and particle size on coking pressure. It is known in fact that an increase in the fineness of the blend involves a reduction in charge density and a reduction in coking pressure. The independent influence of crushing on coking pressure has been studied. In the study examination has been carried out regarding the influence of the degree of overall fineness of the blend, the degree of fineness of each blend constituent (differential crushing) and finally the effect of the mode of crushing, i.e., the shape of the size distribution curve. It was found that at practically constant density, the particle size has a very marked effect on coking pressure, fine crushing reducing the pressure.

Crushing the constituents of a blend separately does not seem to present any particular advantage. A study has shown that systematic crushing can in certain cases have a specific effect, i.e., at the same degree of fineness the pressure is lower than with simple crushing. The conclusion of this study is that the pressure depends much on the proportion of coarse particles (higher than above 2 mm to 3 mm).

By examining the simultaneous action of bulk density and crushing it has been found that the effect of one of the factors depends on the level of the other. Thus the effect of density is greater, the coarser the crushing and similarly, the effect of crushing is more pronounced, the higher the density.

Effect of oven width – The effects of changes in width of the oven chamber, studied using pilot ovens on the wall pressure is not consistent. In the area of 150 mm to 330 mm, the oven width seems to have a substantial effect with decrease of wall pressure with increasing width. However, for greater widths the effect is small or negligible.

At BCRA by using two different ovens with widths 300 mm and 426 mm, it has been found no significant difference between the pressures developed by a given coal when carbonized under similar conditions of BD and flue temperature.

In one of the studies a good correlation has been reported between wall pressure measured in both 150 mm and 300 mm in wide ovens. The plastic area/wall area ratio has been about the same for both these widths.

The coking pressure value measured in a movable wall oven is associated with the ratio K given by surface area of plastic zone/surface area of heating wall. When the width of the chamber is increased so is the increase in the carbonization time. So more time is available for carbonization to progress simultaneously from the sole upwards and the roof downwards, the final plastic layer therefore has a smaller area in a wide oven than in a narrow one. This has been confirmed by direct measurement of the plastic zone area through premature pushing. Thus widening of the chamber involves a reduction in coking pressure but the magnitude of the variation depends on the blend considered while generally remaining small.

Effect of flue temperature – An increase in flue temperature increases the heating rate which in turn (i) shifts the softening and re-solidification temperatures to higher values widens the plastic temperature range,  (ii) increases the flow of gas liberated in the plastic layers, (iii) increases the thickness of the plastic layer, and (iv) reduces the viscosity. The increase in flue temperature also reduces the thickness of the plastic layer due to the increase of the temperature gradient. These effects act on opposite directions so the final effect can be very small or insignificant. By studying this effect the BCRA found that with two blends and one coal, higher heating rates resulted in some increase in wall pressure. However with one coal the opposite was observed and with two other coals they found no obvious effect.

In a study where testing of two blends has been carried out at three levels of temperature 1020 deg C, 1120 deg C, and 1200 deg C, it has been seen that, one blend has shown an increase in both wall and gas pressure with increasing flue temperature while the effect produced using the other blend was not significant.

In another investigation, where the influence of coking rate on gas pressure in commercial ovens has been studied, it has been concluded that increased coking rates has not significantly increased the pressure of the charged blends for the range of coking rates used (the centre of the oven has reached 900 deg C in 12.3 hours to 18.3 hours).

Other investigators have found that coking at faster rates increase the pressure in a consistent way (coking rate is based on time required to reach a coke mass centre temperature of 980 deg C) (Fig 5).

Fig 5 Effect of coking rate on coking pressure