RESEARCHES FOR THE UTILISATION OF SOLID WASTES IN CFBC BOILERS

Liviu G. Drago, Mihu I. F. Grjoab, Ctlin I. Flueraru ICPET-Cercetare S.A., Bucharest, Romania

Abstract The Romanian economic and energetic situation requires a large-scale implementation of clean and efficient technologies for the utilisation of different residues and wastes resulted in industry. The need of such technologies has become even more acute, since in January 1998 the new environmental regulations will become operational. This is why, in our institute, the research focused lately on the clean combustion of industrial wastes, sludge and biomass. The research works initiated aimed to establish a circulating fluidised bed (CFB) combustion and desulfurisation technology for these wastes. In the current paper, the results obtained for two types of residues are presented: a solid waste from the production of electrodes for metallurgic furnaces, and the high-sulphur petroleum coke, that is a residue from oil cracking.

Introduction The industrial development registered worldwide in the last decades has led, besides the growth of energy demand, to the accumulation of high quantities of wastes, with a negative environmental impact. As a consequence, the regulations for environment protection constrained the industrial producers to neutralise their residues or, if possible, to find an alternative use for them. Such examples are paper mill wastes, used tyres, biomass, sludge, petroleum residues. Not long ago, high sulphur petroleum coke was also considered to be among these, not being suitable for use in industry, where it should meet certain quality conditions. Once the desulfurisation technologies developed, its utilisation as a fuel in power plants became possible, in some countries - such as USA and Denmark - even being used on a large scale [1]. However, not all desulfurisation technologies provide a competitive price for the energy produced even from a low-price fuel such as delayed coke. A more attractive solution from this point of view is the circulating fluidised bed combustion (CFBC), which allows the in-bed limestone desulfurisation. The data achieved from several Romanian refineries allowed us to estimate the internal production of high sulphur petroleum coke at about 360.000 tons annually. Most of it is currently exported with 40 USD per ton, and the price could further decrease by the end of the year. In Romania, no power plants use petroleum coke as a fuel. Another potential fuel with no utilisation in Romania is the waste resulting from the production of electrodes for metallurgic ovens. In spite of its high heating value, this waste is not

being used because of the difficulties encountered at its combustion in classical furnaces. Since their utilisation seems to be both necessary and cost-efficient, we performed several combustion and desulfurisation tests with petroleum coke and electrode waste on a 1 MWt CFBC pilot plant to determine the optimum operating parameters.

Description of the test facility The experimental pilot plant is the first CFBC plant conceived in Romania. The fuel it was designed for is lignite with heating value higher than 5000 kJ/kg and the granulation 0 - 10 mm, or oil shale with 3800 kJ/kg and 7% fuel oil support. A simplified scheme of the boiler is presented in fig. 1; its main components are: furnace; cyclone; loop seal; external fluidised bed heat exchanger (FBHE); convective pass; air preheater.

The combustion takes place in the furnace, both in the dense bed from the bottom and above it, where the particle density is lower. The start-up of the plant is performed by means of a fuel oil burner that warms the fluidised bed up to the solid fuel ignition temperature, when the combustion process is initiated. Two water circuits ensure the cooling of the plant. The main cooling circuit consists of the heat transfer surfaces located in the fluidised bed ash cooler, the

furnace and the convective channel. The secondary circuit cools the ash screw conveyer and the fixed bed ash cooler. The solid fuel is introduced in the fluidised bed by means of screw conveyers. During the fluidising process, fuel particles are mixed with recirculated ash particles, burning as they are elutriated in the upper part of the furnace. For a complete combustion and in order to achieve the desired gas velocity, secondary and tertiary air is introduced at two different levels. The coarse particles that are not carried by the flue gas, are discharged through the drain pipe from the air distributor and evacuated by the water-cooled screw conveyer. The test rig is equipped with a tangential entry reverse flow single cell cyclone, where the coal and ash particles are centrifugally separated from the flue gas and fall through a duct into the siphon. The hot flue gas are evacuated at the upper part through a 90 turn and sent in the convective pass. An impact separator was fitted between the furnace and the cyclone for evaluation purpose, but only for a short period. The siphon consists of a mechanical valve and two compartments split by a dividing wall. The movement of the solids in the non-mechanical valve is facilitated by air, which is introduced through a bar grate similar to that from the furnace. Apart from its sealing function, this device makes possible the control of the fraction of ash directed into the external FBHE, by modifying the air flow in the two compartments. A strict control of the furnace temperature is obtained by diverting the solids either directly into the oxidising zone of the furnace, or into the external fluidised bed heat exchanger.

Legend:

1. 2. 3. 4. 5.

Furnace Cyclone Convective bank Secondary cyclones External FB heat exchanger 6. Coal hopper 7. Startup burner 8. Limestone hopper

9. Siphon
10. Air preheater 11. Mechanical valve 12. Heat transfer surfaces 13. Ash cooler 14. Impact separator

Fig. 1: The 1 MWt CFBC pilot plant

The external fluidised bed heat exchanger (FBHE) is used to supplement the furnace heat transfer surfaces, also making the boiler more fuel flexible. It is refractory lined, and a wall divides it in two compartments. The heat transfer surface consists of coils immersed in the bubbling bed. Bar grates are placed at the bottom as fluidising air distributors.

Fuel analysis Technical and elementary analyses have been performed for the selected solid fuels; the results are presented in table 1. The coke samples are from two different

refineries, while the electrode waste samples are from the same location, but the samples were taken at different times.

Table 1: Technical and elementary analysis of petroleum coke and electrode waste Feature Wti Ai Ci Hi Sc i Vi V (d.a.f.) Qii Unit % % % % % % % kJ/kg Petroleum coke Sample 1 Sample 2 1.6 7.80 0.45 0.30 88.02 81.80 4.78 3.95 4.02 3.65 7.85 9.35 8.01 10.17 34338 31219 Electrode waste Sample 1 Sample 2 9.30 4.30 37.00 31.10 52.56 59.77 0.84 1.57 0.30 0.57 2.19 3.80 4.08 5.88 17958 21524

Petroleum Coke Experiments The results of the petcoke combustion experiments are presented in table no. 2. Since the petroleum coke has practically no ash, during the preliminary combustion tests sand was introduced in the furnace as inert material. In the final tests, when limestone was added for desulfurisation, the sand was no longer needed. The particle size distribution of the petcoke was for all the experiments between 0 and 10 mm, with mean diameters between 1.9 and 3 mm, while the granulation of the sorbent was between 0 and 1 mm, with mean diameters between 92 and 500 m. A detailed presentation of the flue gas temperatures from test no. 2 can be seen in fig. 2. The bed temperature was above 1000C as long as the external fluidised bed heat exchanger (FBHE) was by-passed. The direct recirculating valve was then

gradually closed, so a larger fraction of the solids was recirculated through the FBHE. After about an hour, the combustion began here too and the full load was reached. The bed temperature stabilised at 980C.

1200 1100 1000 900

Temperature [deg. C]

800 700 600 500 400 300 200 100 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5

Time [h]
Bed Burner Furnace 2 Furnace 3 Cyclone Convective exit Siphon 1 Siphon 2 FBHE 1 FBHE 2

Fig. 2: Temperature variation during the start-up. Fuel: petroleum coke

The combustion efficiency was over 99%, in spite of the low air excess. In test no. 1, because the combustion was taking place only in the furnace, the bed temperature was between 1100 - 1200C. This fact, correlated with the high air excess, has led to a high NOx emission - about 300 ppm. SO2 concentration was also high, because of the low temperatures from the upper furnace and cyclone, that obstructed the sulfation of the CaO particles. A better desulfurisation efficiency was obtained in test no. 2, when the temperatures have been in the optimum range of 850-1050C [2] all along the recirculation loop: furnace-cyclone-FBHE-furnace. The analysis of solids proved

that the sulfation is higher at fine particles, due to their higher active surface, in spite of the shorter residence time in the reaction zone. The fraction of calcium carbonate decomposed into oxide was around 98%. However, considering the relatively high Ca/S molar ratio (2.4), the 80% efficiency that was obtained is still low. This is mainly because of the small size of the limestone particles - 50% less than 63 m, with an average diameter of 92 m - that were not recirculated, leading consequently to a very short contact time between CaO particles and the flue gas. Better results are expected in the case of industrial plants with greater furnace height.

Table 2: Experimental results obtained at petroleum coke combustion tests on the 1 MWt CFB pilot plant Unit MWt kg/h kg/h C C % ppm % % ppm ppm ppm % Test 1 0.79 91 20 1.44 850 150 88.7 2 2020 6.1 12.8 950 300 110 53 Test 2 1.12 128 35 1.14 950 800 90.7 2.4 2507 2.5 17.2 530 175 80 80

Thermal power Fuel feed rate Limestone feed rate Air excess coefficient Average furnace temperatures cooler Global thermal efficiency Ca/S molar ratio SO2 theoretical concentration O2 Measured CO2 concentrations SO2 NOx CO Desulfurisation efficiency

Electrode Waste Experiments Its particular characteristics, and the fact that this fuel has not been thoroughly studied until now, made the thermogravimetric analysing essential for electrode waste. One of the diagrams is presented in fig. 3, and it confirms once more the low reactivity of this fuel: up to 380C, the only phenomena that occur have a physical nature, and a complete combustion is only obtained at 950C. The limited heat amount released by the volatile matters at temperatures over 380C is not sufficient to start the burning of fixed carbon, meaning that in an industrial plant, the furnace temperature should be increased to at least 700C using liquid fuel, so the start-up procedure will last longer. These conclusions also came out during the many trials performed on the pilot plant for the autothermal combustion of the wastes. At the preliminary tests, it came out that the combustion conditions are very difficult to achieve. After more attempts to run the CFBC plant on electrode waste only, the maintaining of the adequate temperature conditions was thought to be impossible; in this situation, a mixture containing 60-70% electrode waste and 3040% lignite was fed. With this mixture as a fuel, the operation became steady and the first combustion tests have been performed. The combustion efficiency achieved was with up to 4% lower at the combustion of electrode waste and lignite mixture than in the case of petroleum coke. This can be explained by the non-uniform size distribution (the electrode waste having 22.5% under 200 m, compared to 10% for the lignite from the mixture). Thus, a large fraction was not recirculated and had an insufficient time for combustion, the average carbon content from ash being higher (11.6%). Limestone was not

introduced, but the measured SO2 emissions were with 10-15% lower than the theoretical concentration, due to the self-desulfurisation process.

Fig. 3: Thermogravimetric analysis of the electrode waste [3]

Table 3: Experimental results obtained at electrode waste combustion tests on the 1 MWt CFB pilot plant [3]:

10

Fuel Thermal power Fuel feed rate Limestone feed rate Air excess coefficient Average furnace temperatures cooler Global thermal efficiency Ca/S molar ratio SO2 theoretical concentration O2 Measured CO2 concentrations SO2 NOx CO Desulfurisation efficiency

Unit MWt kg/h kg/h C C % ppm % % ppm ppm ppm %

Test 1 Test 2 El. waste + Electrode 40% lignite waste 1.15 1.21 307 200 6.8 1.17 1.20 850 980 350 200 84.2 85.9 2 682 563 2.8 3.5 16.1 15.6 587 116 94 126 180 0 79

Another combustion test has been performed using a different size distribution, with the fraction under 200 m of only 15%, but with 35% between 10 and 15 mm. Average results of this test are presented in table 3 - test 2. The electrode waste was introduced in the lignite-fired furnace at a temperature of over 800C (see fig. 4). The feeding with electrode waste continued for more than three hours; with the external fluidised bed heat exchanger not being used, the furnace temperature reached 1000C. On the temperature diagram, the moments when the direct recirculating valve was opened and the air flow was reduced in the siphon are clearly marked by the temperature variation - a decrease in the external FBHE and an increase in the siphon. The high furnace temperature made the combustion take place in good conditions. It is to be noted the complete chemical combustion, no CO being detected in the flue gas for a long period. However, the short residence time of the fines has again led to a high carbon content in the ash, but

11

due the fact that no lignite was added, it can be considered that the electrode waste combustion efficiency is much better than in the previous experiments. The small height of the pilot plant should also be considered, since in the case of a higher capacity industrial plant, with a much greater furnace height, the unburned carbon content in the fines will significantly decrease due to the longer residence time.

1200

1000

Temperature [deg. C]

800

600

400

200

0 0 0.5 1 1.5 2 2.5 3

Time [h]
Bed Furnace 1 Furnace 2 Furnace 3 Siphon FBHE 1 FBHE 2 Convective exit

Fig. 4: Temperature levels at the combustion of electrode waste (Dragos, 1997)

Limestone was introduced at a Ca/S molar ratio of 2, obtaining a 79% desulfurisation. The relatively low value of the desulfurisation efficiency is due to the lower sulphur content of the fuel, giving thus a lower SO2 partial pressure, but also due to the temperatures above 1000C, that are higher than the optimum temperature range of the process. Increasing the Ca/S ratio to 3 did not prove to be practical, not only because the desulfurisation efficiency improved by just 5%, but also since SO2 emissions are low enough even without limestone addition.

12

Conclusions the experiments have proved that the CFBC technology is suitable for the clean combustion of both the high sulphur petroleum coke and the electrode waste; combustion efficiencies over 99% can be achieved in the case of petroleum coke with very low air excess; by a proper adjustment of the air staging, low NOx emissions can be obtained for both fuel types tested; for an efficient desulfurisation, the temperatures should be above 800C all along the recirculation loop, and the average size of limestone particles under 100 m; the very low volatile percentage gives the electrode waste a very low reactivity, in spite of its high heating value; electrode wastes can burn autothermally, but at higher temperatures, that the plant will have to ensure; an industrial plant using electrode waste as fuel should have bigger start-up burners and a greater furnace height.

13

References [1] Modern Power Systems, vol. 17, no. 3 March 1997, p. 10 [2] Ionel, I., Ungureanu, C. (1996), Termoenergetica i mediul, Bucureti, Editura Tehnic [3] Drago, L., Grjoab, M., Flueraru, C., Scarlat, N., Compared combustion researches in CFBC pilot reactor of solid fuels range with different volatile matters, PECO Joule 2 - Final Report, 1996

ICPET Cercetare S.A. Calea Vitan 236, Sector 3, 74369 Bucharest, Romania

14

Liviu Drago: chemical engineer, 1st rank senior researcher, head of department, Fluidised Bed Combustion Department ICPET Cercetare S.A. Calea Vitan 236, Sector 3, Bucharest, 74369 Tel.: 0040 1 323 42 77; 0040 1 323 42 70 ext. 108 Fax: 0040 1 321 41 70

E-mail: icpetc@com.pcnet.ro

Mihu Grjoab: mechanical engineer, scientific researcher, Fluidised Bed Combustion Department ICPET Cercetare S.A. Calea Vitan 236, Sector 3, Bucharest, 74369 Tel.: 0040 1 323 42 70 ext. 109

Ctlin Flueraru: mechanical engineer, 2nd rank senior researcher, Fluidised Bed Combustion Department ICPET Cercetare S.A. Calea Vitan 236, Sector 3, Bucharest, 74369 Tel.: 0040 1 323 42 70 ext. 114