Hydrogen – a future fuel gas for brick plants? (part 2)
The use of renewably generated hydrogen instead of natural gas is regarded as an important element contributing to the transformation of the clay brick and tile industry to CO2-free production. In comparison with the electric heating of tunnel kilns, which is also widely discussed, using hydrogen as fuel gas in brick plants offers the advantage that only minor changes to the burner equipment are necessary. However, with regard to the application of hydrogen, wide-ranging questions arise in respect of its cost efficiency, combustion properties and the influence of the changed exhaust gas composition on product properties. Focussing on these aspects, a research project was conducted at the IAB Weimar (Weimar Institute of Applied Construction Research) from 04.2021 – 02.2023 specifically for clay brick and tile plants in Thuringia; the project was financially supported by the Thüringer Aufbaubank (development bank of the German state of Thuringia). (The article will be published in three parts)
4 Tunnel kiln operation with natural gas and hydrogen
4.1 Energy balance
The energy balance of a tunnel kiln for roof tiles for a firing with 100 % natural gas and 100 % H2 was calculated with the help of SimKilnT software. The basic operation of this program is described in [28]. The data from »Table 8 were used for the calculation.
The results of the energy inputs and outputs for the combustion of natural gas and H2 are summarized in »Table 9.
For the example shown, the fuel consumption for natural gas operation is 788 kJ per kg fired ware. For operation fuelled with H2, it is reduced by 17 kJ per kg fired ware. The reduction corresponds to around 2 % of the fuel consumption and can be attributed to the reduction in the quantity of combustion air and exhaust gas. For the calculation, the set-point of the exhaust gas temperature was selected at 180 °C. An increase in the exhaust gas temperature on account of the possible dew point increase of the wetter exhaust gas of H2 (cf. Section 4.3) can, however, offset the calculated saving again.
4.2 Water balance of the exhaust gas
In »Table 6 (in part 1 in ZI 1/2024), it was shown that for the stoichiometric combustion of H2 with air, relative to the same amount of energy, 1.67 times the mass of water is formed in comparison with CH4. It can be expected that the increased water content in the kiln atmosphere has an effect on the sintering and fired colour of the products [29]. In the real tunnel kiln, the fuel gas, however, is usually burned with air excess. In addition, large quantities of partly wet ambient air is used for cooling. Other moisture is released during the dehydration of the raw material in the kiln (»Table 17).
For determination of the actual water content in the atmosphere of a tunnel kiln, for the example of a facing brick, first a water balance for the exhaust gas with heating with CH4 as well as with H2 was compiled. Added up in the exhaust gas are the air flow from the cooling zone and all gases, like, for example, CO2 and water vapour from the preheating zone (principle of a countercurrent heat exchanger). In a further step, the water content was determined along the kiln axis (cf. point 4.4). For the calculation, the following data were used (»Table 10).
The physically and chemically bound water contained in heavy clay bodies generally varies from 3 – 9 mass%. It is dependent on the specific mineral content and the residual moisture content after drying and ends up completely in the exhaust gas. The physically bound water escapes together with a residual water content after drying from 0.5 – 2.0 mass% at temperatures up to about 250 °C. The chemically bound water from the clay minerals was released mainly between 450 and 800 °C depending on the mineral composition. For the calculation of the water balance, a water content of 6.0 mass% was used. This results from a clay mineral content of 50 mass% (mix of two-, three- and four-layer clay minerals) and 50 % non-clay minerals (mix of quartz, feldspars, carbonates and iron minerals) as well as a residual moisture of 1.0 mass% after drying.
In the real firing process, the kiln is supplied with cooling air (supply air, rapid cooling) and combustion air from the kiln surroundings is supplied to the burners. Both introduce more moisture into the kiln. The humidity of the surrounding air varies both depending on the prevailing climate zone as well as on the time of year and day, and for Germany is generally between 4 and 20 g water per kg air (»Table 11). For calculation of the water balance, a mean water content of 12 g water per kg dry air is used.
The results of the exhaust gas – water balance are shown in the following tables »Table 12, »Table 13 and »Table 14. The fuel data used for the calculation are listed in »Table 6 (in part 1 in ZI 1/2024). The results show that relative to the total water mass, only 48 % of the water in the exhaust gas comes from CH4 while 61 % comes from H2. In addition, water from the raw material (31 % / 23 %) and the ambient air (21 % / 16 %) finds its way into the exhaust gas.
The results of »Table 12 confirm that the water mass flow in the exhaust gas from the H2 combustion is a factor of 1.67 greater than for CH4 (»Table 6 in part 1 in ZI 1/2024). The water mass from the raw material remains unchanged for the two fuels and was confirmed by results of thermal analysis and laboratory firings (cf. Section 6.2 and 6.3.1 in part 3 in ZI 3/2024). The water mass from the ambient air decreased slightly for H2 as the quantities of combustion air and the exhaust gas are reduced.
In total, therefore, the entire water mass in the exhaust gas of the tunnel kiln for H2 combustion was increased only by a factor of 1.31. This factor, calculated with the moisture contents of the ambient air from »Table 11, lies at 1.37 in winter and at 1.28 in summer. In winter, the ambient air is dryer, as a result of which the influence of the water content from the H2 combustion increases relative to the entire exhaust gas from the tunnel kiln. The water mass is consequently lower in real conditions in the tunnel kiln than stoichiometrically calculated. It becomes clear that with the tunnel kiln always large quantities of ambient air is present in the kiln, which thin the exhaust gas from the combustion. In addition, the constant water mass from the raw material has, on account of the mass throughput, a moderating share in the overall water balance of a tunnel kiln.
»Table 13 shows the origin of the water from the different ambient air quantities like the combustion and leakage air and air injections. On account of similar pressure conditions in the kiln for CH4 and H2, the masses of leakage air and air injection were assumed to be constant while the combustion air mass for H2 is reduced by the factor 0.83 owing to the lower air requirement (»Table 6 in part 1 in ZI 1/2024). The resulting water mass for the entire ambient air consequently decreases for H2 only by the factor 0.96.
The exhaust gas volume and the water content in the exhaust gas are shown in »Table 14 for both fuel gases Because of the lower combustion air volume, the exhaust gas volume is reduced by the factor 0.91 for H2 in comparison with CH4 (»Table 6 in part 1 in ZI 1/2024). As a result, on the other hand, the water content in the exhaust gas of the tunnel kiln increases for H2 to the factor 1.35 compared to CH4 (winter = 1.41, summer = 1.31). This ultimately leads to a higher exhaust gas moisture content of 16.9 vol% for H2 in comparison with 12.6 vol% for CH4.
Empirical values from measurements of the exhaust gas composition show that the calculated water content in the exhaust gas of 12.6 vol% for natural-gas-fired tunnel kilns is comparatively high. Real water contents in the exhaust gas can be considerably lower in a natural gas firing. As shown above, they are dependent on the tunnel kiln settings (mass ratio air / bricks, quantity of combustion air), the specific energy consumption, the amount of water in the raw material and the air humidity.
4.3 Dew point of the exhaust gas
The dew point of exhaust gases depends on their water content. It the exhaust contains sulphur besides water, the dew point increases additionally. For this reason, the temperature in the flue should not fall below the dew point of the exhaust gas as otherwise condensation may form in the exhaust gas system and ultimately lead to corrosion damage.
According to [30] and [31], the dew point of the exhaust gas is calculated according to the following equation:
1/TH2O = 4,924 ∙ 10-3 - 1,945 ∙ 10-4 ∙ ln (pH2O)⇥(5)
pH2O Partial pressure of water vapour [Pa]
TH2O Dew point temperature of water vapour [K]
dTH2SO4 = 251,3∙(pH2SO4 /pH2O )0,1173⇥(6)
dTH2SO4 Dew point increase caused by sulphuric acid, relative to the dew point of the pure water vapour [K]
pH2SO4 Partial pressure of sulphuric acid H2SO4 [Pa]
Equations (5) and (6) were used to calculate the moisture content in the exhaust gas, the dew point and the dew point increase for a sulphur dioxide (SO2) content of 500 mg/mN³ exhaust gas for the facing brick tunnel kiln and the water contents from »Table 11. For the calculation, the maximum permissible value for SO2 in the exhaust gas in compliance with the German Clean Air Act (TA Luft) was chosen [32]. The results are shown in »Table 15.
Because of the higher water contents in the exhaust gas after the H2 combustion, for the example kiln, an increase in the exhaust gas dew point by 5.6 – 7.0 K can be expected compared to with CH4. The dew points vary seasonally for the combustion of H2 by 4.3 K and for CH4 by 5.7 K, independent of the SO2 content in each case.
The calculated dew points are comparatively high with the presence of sulphur in the exhaust gas as low quantities of air and high water and sulphur contents in the exhaust gas was expected. In practice, depending on the quantities of air, the existing air humidity, the sulphur content and the specific raw material, they can vary.
4.4 Composition of the kiln atmosphere
In the tunnel kiln, the raw material is exposed to the specific prevailing atmosphere, which along with the firing temperature and time influences product properties like colour, strength and water absorption. Investigations conducted by the former Institute of Building and Heavy Clay Ceramics in Weimar in cooperation with the School of Architecture and Building and later Bauhaus-University of Weimar showed that a high water vapour content in the kiln atmosphere from the breakdown of the clay mineral at around 500 °C has a substantial influence on the ceramic properties. Also detected was that material reactions shift to other temperatures and the pozzolanicity (metaclay formation) increases. The earlier sintering led to a strength increase as well as to an increase in shrinkage and enlargement of pores. This mechanism was termed the “hydrothermal effect [29], [33], [34], [35], [36]. Knowledge of the kiln atmosphere is therefore a key precondition for the laboratory firings.
In the literature, details on the composition of the atmosphere in the tunnel kiln are seldom found. On the one hand, this is because of the increased measurement effort and, on the other hand, because for the operator only the composition of the dry exhaust gas at the entrance to the kiln is significant for compliance with the statutory required emission limits. Comparable measured values at 800 °C for a masonry brick plant with natural gas firing are given in [37]. In volumetric terms, the contents amounted to 3.4 % H2O(g), 1.8 % CO2, 17.2 % O2 and 77.7 % N2. The exhaust gas of a tunnel kiln heated with natural gas, on the other hand, consists of 2 – 15 % H2O(g), 2 – 5 % CO2 and 12 – 15 % O2. Empirical values from process analyses show that the composition of the kiln atmosphere varies widely from plant to plant as it depends on the operation of the kiln (quantities of supply, burner and leakage air, air/brick ratio), the environmental conditions (season), the kiln sealing (dry with sand, water bath) as well as the water content in the raw material. This dependence is described, for example, in [38].
The water content inside a tunnel kiln depends on the position in the kiln, and accordingly is sometimes significantly lower than in the exhaust gas. This is shown based on the example of a roof tile tunnel kiln with H setters in »4 for the combustion of natural gas and in »5 for the combustion of H2 and was calculated with the help of the simulation software SimKilnT [28]. The basis for calculation is shown in »Table 8. Equipment and kiln settings, raw material, tile size, setting structure and firing curve were the same for both combustion gases.
The upper diagram in each case shows the equipment along the kiln axis, shown as a function of the firing time, from left to right: kiln entrance with exhaust gas extraction, preheating zone with burners, cooling zone with rapid cooling, extractions and supplied air at the kiln exit. In addition, the set temperature curve of the tiles is shown as a function of the firing time. The dry shaped tiles are heated in the preheating zone to a the maximum temperature. This is followed by a holding time, which serves homogenization of the temperatures and sintering in the setting. In the cooling zone, the tiles are cooled to almost ambient temperature. The tiles pass through the specified temperatures in the diagram from left to right. The air is supplied in the opposite direction according to the principle of the countercurrent heat exchanger.
The second diagram in each case shows the air mass flow in the kiln as a function of the firing time. Here, “air” refers to both the ambient air and the combustion products as well as the water vapour and CO2 from the raw material. The air supply begins timewise on the right in the diagram (supplied air at the kiln exit) and initially increases substantially. At the two extraction points in the cooling zone, the quantity of air decreases. With the removal of the cooling air, the temperature gradient at the quartz inversion is reduced and hot air is extracted for the dryer. At the beginning of the cooling zone, the rapid cooling is installed and increases slightly the quantity of air in the kiln. From the end of the firing zone, the quantity of air increases in the direction of the kiln entrance almost continuously because of the combustion exhaust gases as well as the excess combustion air (air factor λ > 1.0). In this range, dehydration and decarbonization of the minerals as well as the oxidation of organic components take place so that in the preheating zone, additionally water vapour and CO2 from the raw material find their way into the kiln atmosphere. At the kiln entrance, the entire quantity of air (cf. Section 4.2) is extracted and sent to the flue.
The air mass flow shows in the cooling zone with the use of H2 the same curve as for natural gas. From the end of the firing zone, the air mass flow increases in the kiln in the direction of the exhaust gas extraction. On account of the lower combustion air requirement and accordingly falling exhaust gas quantity, this increase is – as described above – somewhat lower for H2 than for natural gas.
In the two bottom diagrams, the volumetric concentrations of O2, H2O(g), CO2 and N2 in the kiln are shown. The notes on this follow the air mass flow and begin at the kiln exit. Here, ambient air with 78 vol% N2, 21 vol% O2 and approx. 1 vol% H2O(g) from the air humidity is supplied to the kiln (supplied air). This composition in the cooling zone remains constant to the end of the firing zone as here no combustion takes place and there is no outgassing from the raw material. From the end of the firing zone, for both fuel gases, in the direction of the kiln entrance the content of O2 and N2 decrease continuously. The water vapour content increases as a result of combustion for H2 somewhat more strongly than for natural gas. In the case of the natural gas, the CO2 additionally increases. In exchange, part of the air (O2 and N2) in the kiln is displaced. For the H2 combustion, the CO2 concentration amounts to 0 % as expected. Between 250 – 400 °C, a slight increase can be detected as a result of the combustion of organic components in the raw material. For a high-carbonate raw material, irrespective of the fuel used – the CO2 content would increase between 700 and 850 °C in the direction of the kiln entrance. The results shown were used to set the atmosphere for the firing tests in the laboratory kiln.
5 Conducted laboratory tests
For the tests, 13 individual clays and four clay bodies from the Thuringian brick plants were available. The raw materials were first characterized in order to determine their material composition (mineral content, grain size, inorganic oxides, organic components).
The influence of different water vapour contents in the kiln atmosphere on the thermal behaviour of the raw materials was tested with the help of simultaneous thermal analysis (STA). The STA F449 F3 Jupiter from Netzsch, as a combination of thermogravimetry (TG) and differential thermal analysis (DTA), was connected with a water vapour generator. A test specimen and an inert material were heated up together and the mass loss as well as the temperature changes in the specimen are determined as a function of time and temperature. With calibrated Differential-Scanning-Calorimetry (DSC), the energies of the different raw materials reactions can be quantified. For the measurements, laboratory air with a water content of approx. 10 g water /kg air was used. The water vapour and air flow content remained constant during the entire measurement.
For the determination of the ceramic properties, firings in different kiln atmospheres were conducted in an electrically heated laboratory kiln from the company MUT advanced heating (»6).
The laboratory kiln with heating coils fitted in the inside has a maximum working temperature of 1 200 °C. The interior chamber is gas-tight both for vacuum as well as for positive pressure to 2 bar. The double-leaf outer wall is temperature-controlled to 130 °C with oil to avoid the formation of condensate. Air, N2 and CO2 can be supplied into the kiln. At the kiln door, there is a nozzle for the addition of liquid water and with which the gas mix is atomized. The gas quantities are adjusted by means of the mass flow controller, the quantity of water with a liquid flow controller with the software for the firing curve segment. The total quantity of the wet gas mix results from the sum of the individual gases and amounted to 1 200 standard litres per hour.
As the kiln atmosphere, the gas contents for natural gas and H2 shown in »4 and »5 were used in the tunnel kiln. As the two gas contents only differ slightly, in addition an atmosphere without and one with constantly high water vapour were also set.
The preheating and cooling gradients were the same for all firing curves, the holding time at maximum firing temperature lasted two hours in each case. Cooling was not controlled below 400 °C. The set firing curve for the water vapour atmosphere is shown in »7.
Further firing conditions are listed in »Table 16. As the maximum firing temperature, that of the respective brick plant was chosen. Depending on the raw material, prisms were fired for determination of the ceramic-related properties, while cylinders for determination of the compressive strength were fired for each raw material. From the raw material masonry brick plant 2, plates were made and fired for determination of the thermal conductivity. To reduce the sample mass per firing, the sample specimens from masonry brick plant 2 were distributed over two firings per kiln atmosphere in each case.
The samples had the following dimensions:
Prisms L x B x H [mm]: 70 x 25 x 10
Cylinders D x H [mm]: 23 x 70
Plates L x B x H [mm]: 100 x 100 x 25
For each raw material, prisms were fired in the gradient kiln from 800 – 1 050 °C. The preheating rate was 200 K/h, cooling was uncontrolled.
The fired prisms were used to determine the bending strength, shrinkage, mass loss as well as the water absorption, while the cylinders were used to test the compressive strength. The shrinkage was measured on the basis of shrinkage marks. After the determination of the flexural strength, the water absorption was determined based on one-hour boiling of one sample half. The results are mean values from five single samples in each case.
The thermal conductivity was determined with the two-plate process in compliance with ISO 8302, DIN 52612 and DIN EN 1946, part 2 by means of the TLP 900/100 – SG and Lambda V.2012 software, two-plate from Taurus Instruments. For the measurements, the fired plates were ground to a standardized measurement. The mean thermal conductivity was calculated from three individual measurements on two plates in each case.
The pore structure was determined on the samples tested for thermal conductivity. The pore distribution was determined with Hg porosimetry with the Poremaster 60 from Quantachrome. With the Phenom ProX scanning electron microscope from Phenom World, images of the microstructure were captured with 500x, 1 000x, 2 000x and 5 000x magnification. For the XRD measurement (reflexion), the STADI MP powder diffractometer from STOE & Cie was used.
(Read part 3 in ZI 3/2024)
