B. S. Pathak et al.

/ International Energy Journal 8 (2007) 15-20

15

Design and Development of Sand Bed Filter for Upgrading Producer Gas to IC Engine Quality Fuel
www.serd.ait.ac.th/reric

B. S. Pathak*, D. V. Kapatel*, P. R. Bhoi*1, A.M. Sharma* and D. K. Vyas*

Abstract - Engine quality producer gas must be almost free of solid particulate matters (SPM) and organic contaminants (Tar) to minimize engine wear and maintenance. Except for the catalytic tar crackers, none of the gas cleaning systems commercially available can securely meet a tar removal exceeding 90 % and hence new concepts for tar removal are required. This paper presents a design and development work of sand filter for upgrading producer gas to IC engine quality fuel. The developed sand filter was tested for its performance with SPRERIS 20 kWe down draft circular throat type gasifier with engine set up. The experimental investigations show that the percentage reduction in tar and particulate matters is above 90 %. The total amount of tar and particulate matters was 319 mg N-1m-3 and 53mg N-1m-3 before and after filter respectively.
Keywords - IC engine application, Producer gas cleaning, Regenerative sand bed filter, Tar and particulate matters.

1.

INTRODUCTION

Table 1. Gas quality of raw producer gas from atmospheric air blown biomass gasifier [1]

The present study was carried out under quality improvement programme of thermo-chemical conversion division of SPRERI. The organics produced under thermal or partial-oxidation regimes (gasification) of any organic material are called tar and are generally assumed to be largely aromatic [1]. The producer gas must be free from tar and particulate matters and it should be cooled up to the ambient temperature for IC engine application. The gas quality requirement for the power generators is very strict. However, the postulated gas quality given in the literature should be interpreted carefully as it depends on the type of engine used and the methods of contaminant sampling and analysis procedures used during experiment. Gas cleaning systems are expected to reduce particles and tar content from the raw producer gas to the postulated levels. Typical values of particulates and tar content are given in Table 1 [1]. Table shows that the amount of tar is much higher in countercurrent than cocurrent gasifiers. Therefore in the present investigation cocurrent throat type gasifier with IC engine for power generation was used to evaluate newly developed sand filter. Gas quality requirements for the power generation are very strict (see Table 2)[1]. Numerous methods of contaminant sampling and analysis procedures are in use, which may lead to results, which are not strictly comparable. Large differences can be expected, especially, for the sampling and analysis of tar. In the present study, the field type tar and particulate sampling unit designed by IIT Bombay was used for filter evaluation.

Fixed Fixed bed CFB gasifier bed countercurrent cocurrent gasifier gasifier Particles, 100 - 8000 100 - 3000 8000 100,000 mg N -1 m -3 Tar, 10 - 6000 10,000 150,000 2000 30,000 mg N -1 m -3
Note: Some tar values are indicative since definitions are not specified

Table 2. Gas quality requirements for power generators [1]

Particles, mg N-1m-3 Particle size, m Tar, mg N-1m-3 Alkali metals, mg N-1m-3

IC engine < 50 < 10 < 100

Gas turbine < 30 <5 0.24

* Thermo-chemical Conversion Division, Sa rdar Patel Renewable Energy Research Institute, Vallabh Vidyanagar - 388 120, Gujarat.

Corresponding author: Tel.: +9 1-2 692 -23 133 2 / 23 5011 E-mail: thermo@spreri.org

Fax: + 91-269 2-2 379 82

Mukunda et al. developed sand bed filter to remove the particulates collected by the cooling water in spray tower. The filter is separated in to coarse and fine sections. The coarse filter is filled with sand of 0.5 2 mm size particles and the fine sand bed filled with 0.2 0.6 mm size sand. The size of the filter area is so chosen that the gas velocities through the filter bed do not exceed 0.1 m s-1. This low velocity coupled with the tortuous path causes the removal of a large part of the dust from the gas. The results indicated that the hot-end tar is about 100 mg m-3 and comes down to 20 +/- 10 mg m-3 at the end of the fine filter (cold end). The particulates level also comes down to 50 mg m-3 at the cold end from about 700 mg m-3 at the hot end [2]. Hasler et al. showed that a 90% particle removal is easier than a 90% tar removal. Based on experimental data for the removal of particles and tar, none of the investigated gas cleaning systems can safely meet the gas quality requirements for

16
satisfactory IC engine applications. The sand bed filter and wash tower have already been successfully tested in fixed bed biomass gasifier coupled to IC engines. The results show that particle collection is less critical than tar separation. The highest particle separation efficiency was observed in sand bed filters, rotational atomizers and in wet electrostatics precipitators (ESP). The highest tar reduction was also found in high temperature catalytic tar crackers, in venturi scrubbers and in sand bed filters. [3]. The gas quality for successful IC engine operation has been postulated as below 50 mg N -1 m -3 for th e particle an d less than 100 mg N-1m-3 for the tar [4]. Wet packed bed scrubber, spray towers or venturi scrubbers are often used for tar and particulates removal from producer gas. These commercialized cleaning devices make use of water for gas cleaning and cooling. For large-scale gasifier systems, water requirement would be in huge amount. In addition, there are several environmental regulations restricting disposal of polluted wastewater in the environment. The tar is present in vapor form at the gasifier exit where the temperature of the producer gas ranges from 350 C to 450 C. But as the gas cools to less than 150 C, the tar condenses and solidifies. For IC Engine application, the gas needs to be cooled to around 40 C i.e. ambient temperature. As stated earlier cooling of producer gas to this temperature will condense tar which affect the engine operation [5]. The aim of the present work is to design, develop and evaluate performance of sand bed filter.

B. S. Pathak et al. / International Energy Journal 8 (2007) 15-20

The filter consists of following five compartments: First compartment : Raw producer gas collection Second compartment : Sun dried wood shaves Third compartment : 0.60 mm coarse sand Fourth compartment : 0.15 - 0.35 mm fine sand Fifth compartment : Clean producer gas collection The second, third and fourth compartment were filled with wood shaves, coarse sand and fine sand respectively and each bed having bed height of 85 mm. The sun dried wood shave was used to remove the moisture from the producer gas. The coarse sand was used to remove coarse particulates and tar while the fine sand was used to remove fine particulates and tar. Wire meshes having size equivalent to that of respective sand size was used to separate the beds. The three beds of wood shaves, coarse sand and fine sand can be removed by opening flange provided at the bottom of the filter. The filter has facility to measure pressure drop at each bed and also across the filter. Two sampling port were also provided on the top of the filter in the raw gas collector and clean gas collector for measurement of tar and particulate in raw and clean gas. The schematic diagram of the regenerative sand bed filter is shown in Fig. 1.
Raw gas collector Wood shaves Coarse sand Fine sand Clean gas collector

Raw producer gas inlet

Clean producer gas outlet

2. EXPERIMENTAL METHODS Design of Sand Bed Filter

The design philosophy behind the sand filter is as follows: 1. The sand is neutral and non-reactive material. 2. It is inexpensive and easily available. 3. It is easily available in different grain size grades. 4. It can withstand high gas temperature 5. It is easy to clean and recycle i.e. regenerative material. The sand bed filter was designed, developed and tested with the existing 20 kWe down draft throat type gasifier with IC engine set up for power generation. The area of sand bed filter was designed for superficial velocity of 0.1 m s-1 [2] and gas flow rate of 60 Nm3h-1 i.e. the designed flow rate of the gasifier. The sand bed filter rectangular was fabricated from 3 mm thick mild steel sheet and wire meshes fabricated from stainless steel 304 were used to separate filter bed. SPRERI had conducted an experimental study on effect of bed height on pressure drop across the bed on dry packed bed filter. Experimental investigations suggest a liner relationship between bed height and the pressure drop across the bed. This in turn, increases power consumption of the blower. Literature and our experience shows that most of the gas contaminants (SPM and tar) were deposited in the initial 20 to 30 mm layers of bed height. Therefore, to be on safe side the bed height of each bed in the present filter was taken as 85 mm.

Sampling tap for tar and particulate matters Pressure tap Wire mesh

Dr a i n t a p

Fig. 1. Schematic diagram of regenerative sand bed filter to clean producer gas.

Fig. 2. Comparative photographic view of wood shaves, coarse sand and fine sand before and after test.

B. S. Pathak et al. / International Energy Journal 8 (2007) 15-20 8 (2007) 15-20 B. S. Pathak et al./
Water inlet

17 17

20 kWe Gasifier Sand bed filter Spray tower

IC Engine

Grate

Water outlet

Water circulated pump

Water pond

Fig. 3. Schematic diagram of experimental setup of the sand bed filter.

The properties of sun dried wood shaves, coarse and fine sand were measured at thermo-chemical conversion laboratory of SPRERI. The density of wood shaves was 68 kg m-3 at 9.5 moisture content (w.b.). The density and void fraction of fine sand ( 0.15 0.35 m) is 1399 kg m-3 and 43 % respectively. The density and void fraction of coarse sand ( 0.6 m) is 1450 kg m-3 and 46 % respectively. The comparative photographic view of wood shaves, coarse sand and fine sand before and after test is shown in Fig. 2. Experimental Setup The instrumented experimental set up was developed to evaluate the performance of the filter in terms of tar and particulates removal efficiency. The schematic diagram of the experimental set up is shown in Fig. 3. The experimental set up consists of 20 kWe down draft gasifier, spray tower, sand bed filter and IC engine. The details of gasifier and spray tower were the same as described in Patel et al [6]. System Operation and Measurements The gasifier was charged with charcoal up to the nozzles and then with woody biomass, is ignited by starting blower and using lighting torch at the ignition port. After 15 minutes a good quality of producer gas is generated which is flared at the burner. Then the producer gas was introduced to IC engine through the sand filter. The gas flow rate of producer gas of 60 Nm3h-1 was maintained constant throughout the experimental investigations. The performance of the sand filter was evaluated by measuring the following parameters. 1. Tar and particulate matters 2. Producer gas temperature at inlet and outlet of the filter.

3. Pressure drop across the filter. 4. Water temperature at inlet and outlet of spray tower. Chromel-Alumel (K type) thermocouples and a digital multichannel temperature indicator were used to measure various temperatures. Water-filled U tube manometers were used to measure the pressure drop across the filter and at the orifice meter. The field type tar and particulates measurement set up was used to measure tar and particulates of raw and clean producer gas. This unit measures total tar and particulate (TTP) contents of producer gas. It has the piston cylinder assembly, Non-return suction valve, and filter paper holder assembly, stroke counter and pedal and base. A Whatman glass micro fiber filter, grade: GF/C cat. No 1822 047 having 47 mm diameter was used in the field type tar sampler. 3. RESULTS AND DISCUSSIONS The system was extensively tested to evaluate filter performance. The performance parameters like pressure drop across the filter, temperature of the producer gas at the inlet and outlet of the filter and inlet and outlet temperature of the spray tower, tar and particulate matters in the producer gas before and after the filter were measured. The readings were taken at an interval of 15 minutes. The experimental results of tar and particulate matters measured by using field type tar sampler are shown in Table 3. Table 3 shows that the cleaning efficiency of the sand filter obtained in the range of 83 97 %. Table 3-shows significant reduction in collection efficiency with time because the filter area was reduced due to deposition of tar and particulate in the filter media. To avoid this problem water shower can be provided on the top of each bed and collect the contaminated water from the bottom of the filter, which continuously removes the tar and particulates from the filter media.

18

B. S. Pathak et al. / International Energy Journal 8 (2007) 15-20 Table 3. Tar and particulate matters measurement using field type tar sampler

Sr. No.

Time

Sampling

No. of Stroke 40 40 40 40

1 2 3 4

3.00 PM 3.00 PM 4.15 PM 4.15 PM

Filter Inlet Filter Outlet Filter Inlet Filter Outlet

Final Before Drying A 0.185 0.096 0.109 0.096

Weight of Filter Paper, mg Initial Diff. Final C A-C After Drying B 0.123 0.094 0.091 0.095 0.094 0.002 0.100 0.094 0.015 0.095 0.094 0.002

Diff. B-C 0.029 0.001 0.006 0.001

Tar and Particulate Before After Drying, Drying, mg/m3 mg/m3 4840 106 798 106 1542 53 319 53

Cleaning Efficiency, % 97 ---83 ----

Pressure drop, mm H2O

However, this techniques leads to another two problems first, requirement of large quantity of water and second, disposal of contaminated water to the environment.
1200 1000 Temperature, deg C 800 Experiment 1 600 400 200 0 12:15 PM 10:48 AM 3:15 PM 4:15 PM 4:45 PM 5:45 PM 6:15 PM 7:00 PM 1:15 PM 2:15 PM 3:15 PM 4:00 PM 4:45 PM Experiment 2

80 70 60 50 40 30 20 10 0 3:15 PM

Experiment 1

Experiment 2

Raw gas temperature Combustion zone

Pressure drop across the filter 4:15 PM 4:45 PM 5:45 PM 6:15 PM 7:00 PM 1:15 PM 2:15 PM 3:15 PM 4:00 PM 12:15 PM 4:45 PM
4:45 PM

Time, h

Fig. 6. Variation of pressure drop across the filter with time.

Time, h

70 Temperature, deg C 60 50 40 30 20 10

Experiment 1

Experiment 2

Fig. 4. Variation of raw gas temperature, combustion zone temperature and drying zone temperature with time.

Figure 4 shows the variation of raw gas temperature, combustion zone temperature and drying zone temperature of 20 kWe down draft gasifier with time, which shows that the variation of temperature is almost constant, which indicates the steady and smooth operation of gasifier.
70 Temperature, deg C 60 50 40 30 20 10 0 3:15 PM 4:15 PM 4:45 PM 5:45 PM 6:15 PM 7:00 PM 1:15 PM 2:15 PM 3:15 PM 4:00 PM 12:15 PM 4:45 PM 10:48 AM Experiment 1 Experiment 2

Outlet water temperature of sprayer Inlet water temperature of sprayer

0 3:15 PM 4:15 PM 4:45 PM 5:45 PM 6:15 PM 7:00 PM 1:15 PM 2:15 PM 3:15 PM 12:15 PM 4:00 PM 10:48 AM

Time, h

Fig. 7. Variation of water temperature at inlet and outlet of spray tower with time.

Filter inlet temperature Filter outlet temperature

Superficial velocity, m s-1

0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 3:15 PM 4:15 PM 4:45 PM 5:45 PM 6:15 PM 7:00 PM 1:15 PM 2:15 PM 3:15 PM 4:00 PM 12:15 PM 4:45 PM 10:48 AM Superficial velocity Experiment 1 Experiment 2

Time, h

Fig. 5. Variation of gas temperature at inlet and outlet of filter with time.

Figure 5 shows the variation of gas temperature at filter inlet and outlet with time. It is observed that the slight variation in the gas temperature at filter inlet and outlet which is due to variation in raw gas temperature. It is also observed that there is small temperature drop of producer gas in the filter.

Time, h

Fig. 8. Variation of superficial velocity with time.

10:48 AM

B. S. Pathak et al. / International Energy Journal 8 (2007) 15-20

19 [3] Hasler, P and Nussabaumer, Th. 1999. Gas cleaning for IC engine applications from fixed bed biomass gasification. Biomass and Bioenergy. 16 : 385 - 395. [4] Knoef, H. A. M. and Koele, H.J. 2000. Survey of tar measurement protocols. Biomass and Bioenergy. 18: 55 - 59. [5] Patel, J.B.; Jena, U.; Kapatel, D.V. and Patel, V.B. 2005. Recycling of wastewater generated during producer gas conditioning. Proceeding of 39th ISAE Convention & Symposium, ANGRAU Hydrabad, India. 9-11 March. [6] Patel, S.R.; Swamy, S.M. and Rao, C.S. 1993. Development and performance of a 20 kW gasifier system using woody agricultural residues as feed stokes for power generation. Fourth national meet on biomass gasification and combustion. Interline Publishing, Banglore, India, 187 - 197.

Fig. 6 shows the variation of pressure drop across the filter. It was observed that the pressure drop across the filter is increased with time, which shows blockage of cleaning area of the filter, which shows deposition of tar and particulates in to the filter medium. It was also observed that the pressure drop was low in the bed of wood shaves and coarse sand bed while pressure drop was high in the bed of fine sand. Fig. 7 shows the variation of water temperature at inlet and outlet of water spray tower. It is observed that the inlet temperature of water to spray tower remains constant while the outlet temperature decreases as raw gas outlet temperature decreases and increases as raw gas outlet temperature increases. Fig. 8 shows variation of superficial velocity with time. It is shown that during the first experiment superficial velocity is maintained constant as per designed value and in the second experiment it varies due to variation in flow rate of producer gas as superficial velocity is function of producer gas flow rate. Fig. 2 shows that the coloure of coarse and fine sand present in the third and final compartment has changed to black colour from yellow, which proves that the tar and particulate matter have been trapped. It is also observed that the fine sand of size of 0.15 0.35 (fine sand) are more darker in colour than the sand of 0.60 mm (coarse sand) which shows that the fine sand are more effective for cleaning than coarse sand. The cooling water washes part of the tar and the rest is deposited in the sand bed filter. Experiments have shown that the moisture carries over from the cooler causes slight wetting of the sand bed. It was observed that the percentage reduction of moisture content by the filter was around 10 18 %. 4. 1. CONCLUSIONS The developed sand filter worked satisfactorily and the tar and particulates removal efficiency was in the range of 83 97 %. The pressure drop across the filter was increased with the duration of test, which indicates that the moisture, tar and particulates being filtered out from the producer gas. ACKNOWLEDGEMENT The fund and infrastructure facilities for the research work provided by the management of Sardar Patel Renewable Energy Research Institute is gratefully acknowledged. REFERENCES [1] Milne, T.A.; Abatzoglou, N.and Evans, R J. 1998. Biomass Gasifier "Tar" Their Nature, Formation and Conversion. The Biomass Energy Foundation Press. 1810 Smith Rd. Golden, CO 80401. [2] Mukunda, H. S.; Dasappa, S.; Paul, P. J.; Rajan, N.K.S. and Shrinivasa, U. 1994. Gasifiers and combustors for biomass - technology and field studies. Energy for Sustainable Development 1(3).

2.

20

B. S. Pathak et al. / International Energy Journal 8 (2007) 15-20

Performance Test Techniques for EJECTOR VENTURI SCRUBBER*

L. S. HARRIS, Manager, Research and Development Department, R. HARTENBAUM, Research Engineer, Schutte and Koerting Company, Cornwells Heights, Pennsylvania

I he techniques presented were developed for the purpose of testing and evaluating the performance of the Ejector Venturi Scrubber System. A typical six-inch size system is shown in Fig. 1. The system is comprised of an ejector venturi scrubber and a scrubber separator. The separator is attached to the discharge end of the venturi unit. The ejector venturi scrubber8 is a wet type of contactor, which utilizes the jet action of the contacting fluid to pump the gas through the system. The performance characteristics of interest for such a system are as follows: 1 Pumping capacity and corresponding self induced draft. 2 Liquid carry-over in discharge stack. 8 Solid particulate collection efficiency. The tests and evaluation were performed on a four-inch scrubber system. The test scrubber and separator were transparent models made to the same dimensions as our stock units, and can be seen at the right in Fig. 2.
Pumping Capacity

A typical set of pumping performance curves of the ejector venturi scrubber at various nozzle pressures is shown in Fig. 3. Velocities in the suction duct vary from 139 ft/min at the low capacities up to a high of approximately 3300 ft/min. Air flow was measured in a four-inch diameter duct by two means; a specially calibrated Biram's type anemometer, for duct velocities from 0 to 2000 ft/min, and a 5/ie inch diameter pitot tube for duct velocities higher than 2000 ft/min. The pitot tube was located 10 diameters from the upstream opening of the straight inlet duct and three diameters from the downstream entrance to the scrubber, as illustrated in Fig. 4. These dimensions are greater than those specified by the National Association of Fan Manufacturers, Inc.,1 for * Presented at the 54th Annual Meeting of APCA, Commodore Hotel, June 11-15, 1961, New York, New York.
May 1 962 / Volume 1 2, No. 5

determining performance of fans and blowers by means of pitot tube traverse. The aforementioned specifies seven and a half diameters upstream and two and a half diameters downstream of straight duct. Air flow into the unit was controlled by two methods: (1) by a number of various size orifice plates located at the inlet of the duct when the air flow was determined by pitot tube, and (2) by a segmented circular type shutter located 13 diameters from the inlet of the duct when using the anemometer, as illustrated in Fig. 4. Air flows determined by horizontal and vertical traverse pitot tube readings, at velocities in the range of 0 to 1000 ft/min, were found to be subject to large errors, anywhere from 10% to greater than 100%. The lower the velocity the greater the errors and uncertainties. The errors and uncertainties incurred were due to (1) manometer reading accuracy in determining velocity head pressure in range of 0.00074 to 0.063 inch of water (139 ft/min to 1000 ft/min), (2) minor air flow fluctuations of the order of 0.01 inch water, (3) manometer zero adjustment and zero reading inaccuracies, and (4) the existence of negative (minus) velocity head readings near the duct wall at mean air flows less than 1000 ft/min, indicating internal eddying andflowreversals near duct wall. To obtain air flow readings within five percent accuracy in the range of 0 to 2000 ft/min, the specially calibrated four-inch anemometer was mounted into the entrance of the four inlet duct. The anemometer was calibrated for total air flow versus dial reading, by sucking air through the anemometer and HEI (Heat Exchange Institute) standard flow nozzles.2'7 The anemometer was sealed into the entrance of a duct the same size as used in our scrubber test and the HEI flow nozzle was mounted seven and a half diameters downstream. The calibration was achieved by critical pressure drop across the nozzle, which was created by using an S&K two stage steam jet vacuum pump. Five HEI nozzles XU to z/i inch were used to cover the range in which we were interested. The anemometer was not used for

determining flow above 2000 ft/min, due to mechanical speed limitations recommended by the manufacturer. During the first series of tests performed in the range above 2000 ft/min, the flow was determined by four horizonal traverse and four vertical traverse pitot tube readings. Since both vertical and horizontal traverse readings were found to be almost identical, within gage reading accuracjr, remainder of air flow readings were determined by horizontal traverse pitot readings only. The ejector venturi scrubber's corresponding draft was measured immedi-

Fig. 1. Typical six-inch figure 4014 Schutfe and Koerting Company ejector venturi scrubber system.

227

ately after the scrubber inlet flange, by means of a well-type manometer reading inches of water.
Liquid Carry-Over

2398

Fig. 2.

Solid particulate scrubber test stand.

20
Fig. 3.

40

60

80

100

120^

140

160

180

CAPACITY- ACFM
Air handling capacity for S&K four-inch 4010 ejector venturi scrubber.

The term "liquid carry-over" refers to droplets of contacting fluid in the discharged gas leaving the scrubber system. Obviously, in scrubber applications, it is desired to have minimum liquid carry-over. A typical liquid carry-over performance curve for the ejector venturi scrubber system is shown in Fig. 5. Since we were working with a small unit, total sampling by impaction was found to be the most practical method. The results obtained were reproducible within 10%. The method used consisted of ordinary six-inch diameter blotter paper rigidly mounted in a support one inch from the end of the scrubber discharge, shown in Fig. 6. The circumferential opening was equal to the flow area and was wide enough so that its effect on pumping capacity was nil. The liquid carry-over droplets were caught by inertial impaction and held fast by being absorbed by the blotter. Due to the relatively small amount of liquid carry-over, the sampling time even under the worst conditions was as long as five minutes before the blotter became saturated. The blotter support and sampling blotter were weighed in the vicinity of the test, immediately before and after each test. The blotter and support weighed approximately 600 grams and the amount of liquid sample collected varied between one to six grams. At the high efficiency points, it took as long as one hour to collect the one gram of liquid carry-over, and the discharge gas was dry to the "touch" of the hand. A triple beam balance having an accuracy of 0.1 gram was used. This resulted in a 20% uncertainty in the lowest readings, which was felt to be reasonable when considering all the factors involved. The above technique for sampling and determining the liquid carry-over was not the first method tried. Attempts were made to obtain representative samples by an isokinetic probe in the discharge stack with the droplets being caught in a series of drying tubes filled with glass wool. Results obtained were not consistently reproducible within 10% and the errors were as high as 50 and 100%. Several different size glass sampling probes were tried, x/4- to 1/2-inch diameter. Close examination revealed that the small liquid droplets were collecting on the walls of the probe, coalescing, and then falling back down into the stack. After witnessing the aforementioned, further attempts to use any type of probe sampling for liquid carryover were abandoned. Prior to using blotter paper as the means of collecting
Journal of the Air Pollution Control Association

228

4"DIA. DUCT

PITOT TUBE PORT.

PRESSURE TAP FOR DRAFT READING

ANEMOMETER FITTED INTO ENTRANCE OF DUCT (REMOVED WHEN USING PITOT TUBE)

CIRCULAR ORIFICES INSTALLED OVER OPENNING OF DUCT WHEN USING PITOT TUBE

SEGMENTED CIRCULAR-TYPE SHUTTER (REMOVED FROM LINE WHEN USING PITOT TUBE)

AIR OUT

WATER OUT Fig. 4. Air flow fest set up for S&K four-inch ejector venturi scrubber.

the water drops, fiber glass filter pads were tried. This was unsatisfactory due to the fiber glass pad becoming weak and falling apart. To avoid this, the fiber glass pad was backed up by blotter pad. This produced satisfactory results which were reproducible within 10%. As a matter of interest, we ran identical tests with the blotter paper alone and it produced the same results as the fiber glass pad backed with blotter paper. Since the fiber glass pad was not needed, it was eliminated and blotter paper was used alone.
Solid Particle Collection Efficiency t

fractional efficiency curve, particle size versus collection efficiency, as illustrated in Fig. 7. The obtaining of this performance characteristic involved the following work: (1) developing a method of generating the desired solid particle aerosol, (2) determining a method of obtaining a representative sample of solid particulate in both intake and discharge ducts, and (3) developing a method of conveniently determining solid particle size and distribution.
Aerosol Generator

The typical performance characteristic being sought is the grade or

100

The method used to generate the desired solid particulate aerosol in the range of one to 100 microns consisted of using an S&K Which a ir jet exhauster,3 a constant speed turntable type feeder

90 80 70

similar to the Harvard No. 4 unit developed by M. W. First, et al.,4 and an inexpensive pneumatic vibrator for the hopper, as illustrated in Fig. 8. The air jet exhauster sucks the particulate off of the turntable and mixes it with the air jet and discharges the mixture at high velocity into the entrance of the four-inch scrubber suction duct. Aerosol concentrations as high as three grams/cu ft were generated in this manner. The entire aerosol generator was enclosed in a box-like container with openings provided for suction air to the scrubber and for an exhaust venting system, as illustrated in Fig. 2. This arrangement permitted personnel to work in the immediate area without the use of gas masks for protection against breathing the aerosol. Prior to using the above method, unsuccessful attempts were made to use a similar type of aerosol generator not requiring the complexity of a constant speed turntable feeder. The system consisted of a V2-inch S&K air jet exhauster, a duPont Dry Dye Eduction Probe5 and a barrel or drum vibrated by a pneumatic vibrator. One end of the duPont dry dye probe is attached to the suction port of the air jet exhauster while the other end rests in the barrel of solid particulate. Although this arrangement works well for duPont in introducing dry dye into gasoline when using a liquid jet exhauster, it did not generate a steady enough aerosol for our purposes. The bulk densities of the solid particulate we used, fly ash and glass beads, were in the range of 90 to 300 lb/cu ft, and were very much higher than the dry dye used by duPont. It is suspected that this is the reason why the duPont probe did not work satisfactorily in this application. As a matter of general interest, we have used the duPont Probe successfully in applications where it was desired to uniformly suck or feed other solids directly from a barrel, such as Celite No. 379 (diatomaceous earth), having a bulk density of eight lb/cu ft.
Duct Sampling Technique

DC

I LJ

eO[-f

4 " E.V. SCRUBBER WITH S S K SCRUBBER SEPARATOR 4 " E.V. SCRUBBER WITH GRAVITY SEPARATOR

CO CO 5 0 DC Q_ 40 LLJ

20 10

SCHUTTE a KOERTING CO. CORNWELLS - HEIGHTS BUCKS CO., PENNA.

O.I

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

11 .

1.2

1.3

L4

1.5

CARRY OVER- GALOFWATER/1000 FT3 AIRx 10*

Fig. 5. Carry-over from a four-inch S&K scrubbing system.

The arrangement of the sampling ports and equipment is shown in Fig. 2. A y4-inch diameter null balance sampling probe6 was used for both the inlet and discharge ducts. The inlet duct sampling port was located 10 diameters from the entrance of the duct and three diameters from the entrance to the scrubber. The discharge duct sampling port was located in a straight duct seven and a half diameters after the separator. A Millipore plasticfieldmonitor, containing an aerosol type membranefilter,was attached to the end of the probe as shown at right-hand center portion of the picture. A 3/4-inch S&K air jet exhauster was used to draw the sample through the probe, Millipore filter, and
229

May 1 962 /

Volume 1 2, No. 5

WATER IN

PRESSURE GAGE BLOTTER IMPACT SURFACE

AIR IN

BLOTTER SUPPORT SWINGS SIDEWISE IN OR OUT OF CONTACT WITH DISCHARGE GAS

collected. Prior to using the transparent Millipore field monitors, metal filter holders were used. With metal filter holders, the membrane filter had to be removed to obtain the optical density readings. This method presented the possibility of losing some of the sample.

Evaluation of Particle Size

The samples collected were evaluated in three ways: (1) weight, (2) size and distribution, and (8) optical density. The weight of the sample collected was determined by standard laboratory procedures using an analytical balance having a sensitivity of 0.1 mg. The inlet samples collected were in the range of 0.2 to 0.4 gram and the corresponding sampling time varied between two to 10 minutes. The discharge sample collected was in the range of 0 to 40 milligrams and the sampling time was as long as 24 minutes. The particle size and number was determined by use of a vertical projection, light field microscope, shown in Fig. 9. This microscope was made especially for dust counting work. Its major advantage over other types of dust counting microscope equipment is that it can be used in a manner similar

PRESSURE TAP FOR DRAFT/ READING

I" VERTICAL OPENING

Fig. 6.

WATER OUT Liquid carry-over test set up S&K ejector venturi scrubber system.

diy gas meter. The null balance condition was indicated on an inclined manometer and was controlled by a valve on the suction line to the air jet exhauster. The samples were taken with the probe mounted at the point of average velocity. This was done in order to keep the amount of work involved to evaluate a test run to a

minimum. The Millipore plastic field monitors were chosen in favor of other methods because of the handling convenience, versatility, and low cost in our application. The transparent monitors permitted us to watch the buildup of particulate on the filter, and stop the sampling when a sufficient quantity was

IOO

f2 90

or
>58O z
UJ

U_ U. UJ

_70

O O

60

SCHUTTE a KOERTIN6 CO. CORNWELUS- HEIGHTS BUCKS CO..PENNA.

5
Fig. 7.

10

15 20 25 FART1CLE SIZE-MICRONS

30

35

40

Fractional efficiency curve for a four-inch figure 4014 S&K ejector venturi scrubbing system.

Fig. 8.

Aerosol generator.

230

Journal of the Air Pollution Control Association

Fig. 9. Leitz vertical projection dust counting microscope.

to nn ordinary photographic onlargor. This feature permits us to make full size eight-inch diameter prints of the projected view, as shown in Fig. 10. Larger or smaller prints could be made if desired by adjusting the projection distance. This photographic procedure permitted us to easily make permanent records and do the dust counting at any convenient time and under much less fatiguing conditions. The highest degree of magnification we used was 1000 X, and the total magnification due to projection was 1200X, when working with an eight inch view. At this degree of magnification, the 100 watt light source was

sufficient to permit viewing in a normally lighted room. Two methods were used for preparing the specimens for microscopic examination. For light loading, the membrane niters were used directly by mounting on a glass slide and rendering them transparent with immersion oil. For heavy loading, the membrane filter was dissolved in acetone and small portion placed in a counting cell. The distribution was determined by photographing five areas of each microscope specimen and then counting and sizing up to a total of approximately 500 particles. The number of particles that should be counted in any particular sample is dependent on the width of its distribution and the statistical level of confidence desired. The optical density was determined for each sample by light comparison reading of the membrane filter before and after sampling. The apparatus used consisted of a photoelectric cell, a meter calibrated in candlepower, and a light source whose intensity was varied by a variable transformer. The transparent Millipore field monitors permit us to take optical density readings of the membrane filter while mounted in the monitor. A correlation between dust loading and optical density was obtained but due to the limited amount of data collected to date we do not feel warranted publishing these results.7
REFERENCES 1. "NAFM Standards, Definitions, Terms, and Test Codes for Centrifugal, Axial, and Propeller Fans." Bulletin #110

Fig. 10.

Typical microscopic record at 5 2 5 X .

2.

3. 4.

5. 6.

7. 8.

National Association of Fan Manufacturers, Inc., 19. H. S. Bean, R. M. Johnson, and T. H. Blakeslee. "Small Nozzles at Low Values of the Diameter Ratio." ASME Transactions, 76: 833-871 (August 1954). Schutte and Koerting Company, "Air Jet Syphons," Bulletin 2A, October 1954, 3 and 15. M. W. First, et al., Air Cleaning Studies, U. S. Atomic Energy Commission, Progress Report (NYO-1581) (June 30, 1951). E. I. duPont de Nemours & Company, "duPont Petroleum Dyes," Bulletin A-2723, 11-56-10, p 12. Western Precipitation Corporation, "Methods for Determination of Velocity, Volume, Dust, and Mist Content of Gases," Bulletin WP-50, p 24. Heat Exchange Institute, "Standards for Steam Jet Ejectors," Third Edition, 18-28(1956). R. L. Webb, The Draft Producing Venturi Scrubber, Proceedings of APCA Semi-Annual Technical Meeting, 35-41 (December 1956).

Index to Advertisers
Academic Press, Inc Buffalo Forge Company Melvin F. Hall, Inc., Buffalo W. L Faith Hemeon Associates Lauren B. Hitchcock Associates William T. Ingram Jackson & Moreland, Inc 256 255 259 257 216 Inside Front Cover Outside Back Cover Resources Research, Inc. Staplex Company R. E. McGuire Associates, New York Travelers Research Center, Inc E. K. von Brand 216 216 216 215

Western Precipitation 222 Hixson & Jorgensen, Inc., Los Angeles John Wood Company Inside Back Cover Ray Ellis Advertising, Inc., Chatham, New Jersey

May 1962 / Volume 1 2, No. 5

231

58 Indian Journal of Science and Technology Vol. 3 No. 1 (Jan 2010) ISSN: 0974- 6846

Performance analysis of downdraft gasifier for agriwaste biomass materials

K. Sivakumar and N. Krishna Mohan

Dept. of Mechanical Engineering, Annamalai University, Annamalai Nagar, Chidambaram-608 002, Tamil Nadu, India
ksr_aufeat@yahoo.co.in Abstract In this study, combustible gas production from shell biomass materials such as coconut shell, groundnut shell and rice husk was experimentally investigated at 8000C using gasification technique by a downdraft gasifier. The combustible gases H2, CO, CH4, CO2, and N2 were formed. The calorific value of the producer gas for various shell materials was found. The calorific value of coconut shell is 23.01% higher than ground nut shell and 45.45% higher than rice husk. Based on performance analysis, it is concluded that coconut shell is the best suitable fuel for gasifier compared to the other two biomass shell materials. Keywords: Biomass, gasification, coconut shell, groundnut shell, rice husk, energy from waste. Introduction Biomass is recognized as one of the major potential sources for energy production. Biomass resources viz. forestry residues, energy crops, manufacturing food waste, coconut shell, bagasse from sugarcane processes and food processing residue all have been used for energy generation. Biomass gasification, the complete conversion of biomass to a gaseous fuel by heating it with a gasification medium such as air, oxygen or steam is fast becoming the most promising process for electricity generation. Biomass gasification products have been demonstrated for the generation of electricity via boilers and steam turbines or internal combustion engines, Fired in gas turbines and even proven suitable for some types of fuel cell. Combustible gases viz. H2, CO, CH4, CO2, and N2 were formed in the reactor (Adnan Midilli et al., 2001) reported the cashew nut as biomass fuel). Shell biomass materials maintains net calorific values between 6000 K.cal/kg and 3000K.cal/kg. This is almost half that of natural gas and fuel oil. This coupled with the low mass density of biomass means it has a low energy density compared to fossil fuels and therefore shell bio mass materials are suitable for small scale plants (Savitri Garivait & Ulonwan Chaiyo, 2006). There are various biomass shell materials are available in tropical countries. Where in the following three bio mass materials which are commonly available and reliable were taken for performance analysis (Singh et al., 2006): 1) Coconut shell was used as the feedstock with size range 0.75-1.0 mm. The proximate and ultimate analysis of biomass as follows: moisture 10.53%,fixed carbon 13.10%,volatile matter 57.96% and ash 18.4%.The Research article
Indian Society for Education and Environment (iSee)

ultimate analysis was C 50.2 %,H 5.30%, N 0.0% and O 43.4%. 2) Ground nut shell has great potential for commercial use. It is used as fuel filler in cattle feed, hard particleboard, cork substitute, activated carbon etc. The Proximate analysis reveals that moisture 8.76%, fixed carbon 15.50%, volatile matter 54.96 and Ash 20.3%. The ultimate analysis shows that C 48.3 %, H5.70%, N 0.8 % and O 39.4%. 3) Rice husk is recognized as an important source of energy, particularly in developing countries where it economies largely based on agriculture and forestry. Agriculture waste is one form of biomass which is readily available but is largely not utilized in energy recovery schemes. Rice husks are an agriculture waste produce as a by-product of the rice milling industry. The world wide annual production of rice husks is estimated to be about 100 million tones, 90% of which generated in developing countries. The proximate analysis and ultimate analysis of biomass as follows: moisture 7.9%, volatile matter 59.5%, fixed carbon 19.9% and ash 17.1%. The ultimate analysis was C 38.9%, H 5.1%,N 0.6%,O 32.0% (Van der Drift et al., 2001). Experimental setup and procedure The prototype small-scale biomass gasifier system comprised an agri-waste feed hopper and feed auger, air blower, cyclonic gasifier, external cyclone, cooler/condenser and engine (Fig.1).

Biomass fuels

The biomass gasification system developed within this project was a swirling flow gasifier. The concept of this entrained flow gasifier used air to entrain shell biomass in a turbulent vortex within the reactor, which incorporated two stages of separation to remove the char and ash produced in the process. The intense continuous Biomass energy Sivakumar & Krishna Mohan
Indian J.Sci.Technol.

Equipment

http://www.indjst.org

59 Indian Journal of Science and Technology reaction enabled gasification of high volumes of biomass in the compact reactor. Vol. 3 No. 1 (Jan 2010) ISSN: 0974- 6846

tests. Limited and controlled amount of air for partial combustion enters through the air nozzles (Fig.2). An array of 'U' tube design condensers were used to cool the Shell gas and condense any water and tars from the gas. The cooler was designed to reduce shell gas from temperatures of 7000C - 8000C down to 300C - 400C, so that the maximum volumetric energy content of the shell gas is achieved. The design factored in minimum pressure drop across the system. Dogru et al. (2002) concluded that, the pressure drop is directly proportional to the air flow rate for all bio mass.

Cooler/condenser

Fig.1. Schematic diagram of biomass gasifier

The system used for this work was sized to gasify up to 20kg/hr of shell biomass under atmospheric pressure Filter conditions; This gives a total energy input to the gasifier The cool gas was filtered in various filter systems of around 20kW. The system was tested over a range fuel such as in fine filters and cloth bag filters to evaluate the input rates from 12kg/hr up to 15kg/hr. For the final performance of the upstream separation systems. testing with engine operation the gasifier was turned down to half its peak capacity of 20kg/hr, results shown in Experimental procedure this report reflect the performance of the gasifier in turn In the down draft gasifier, both the fuel and the gas down operation (Philippe Mathieu & Raphael Dubuisson, flow downwards through the reactor enabling the 2002). The final prototype gasifier was manufactured in pyrolysis gases to pass through a throated hot bed of 316-grade stainless steel and proved suitable when char which is supported by a grate. This results in tested under continuous high temperature 800C cracking of most of the tars in to non condensable gases operation conditions. Khater et al. (1992) discussed and water. Furthermore, air is behavior of downdraft gasifier Fig.2. Down draft biomass gasifier plant admitted to the fuel bed through air and maximum temperature intake nozzles causing pyrolysis to range, Pattaraporn Chaiprasert charcoal and volatiles that partially and Tharapong Vitidsant burn as they are produced. The (2009) showed that increasing gaseous products of this flaming carbon conversion to gas from pyrolytic combustion then 44.13-78.43% where as tar consume the charcoal produced was decreased from 19.55during the pyrolysis and are 1.4% at temperature of 800C. reduced to fuel gas. In this way, tar

Heat-up gas burner

A burner system was designed to supply and ignite a supplementary gas to fire through the gasifier to heat it up to temperatures to initiate gasification reaction. Specification temperatures were up to NOT for the reaction initiation. The design also ensured no interruption of the gasification reaction and not be fouled by biomass particles.

vapors are typically lowered to 0.1% of the total feed. Whereas, in draft or cross flow gasifier, tar levels are higher than 0.1%. Paulo R. Wander and Carlos (2004) reported the various level of tar gasification. Avdhesh and Sharma (2009) revealed the characteristic parameters of gasifier. This gas is used for operation of, internal combustion engine. Results and discussion The gasification experiments were performed for three agriwaste biomass materials. The volumetric flow rate of wet product gas was measured with a gas flow meter. Drying, pyrolysis and oxidation zone temperature of the gasifier and the water scrubber and box filter outlet temperature were also monitored with the aid of R and K Sivakumar & Krishna Mohan
Indian J.Sci.Technol.

Hopper/feeder system

The biomass hopper and feed system have a small hopper with agitator provision. The hopper was small, however provided shell biomass loading of 15kg of feedstock, which under most test conditions was suitable but needed refilling and agitation was required for longer Research article
Indian Society for Education and Environment (iSee)

Biomass energy
http://www.indjst.org

60 Indian Journal of Science and Technology type thermocouples. The amounts of tar-dust and condensate in the product gas were determined before and after cleaning the product gas. The combustible gases from the shell biomass were obtained as H2, CO, CO2, CH4, and N2 and their amounts in the dry product gas were estimated in units of kg h-1 and as volume by volume percentage. The calorific value and total energy of the produced gases were determined and evaluated based on the dry product gas volumetric flow at STP hot cold and raw gas efficiencies of the dry product gas were estimated in order to investigate its usage in the CHP engines. During the experiment, pressure drops were measured as 0.10- 0.80 mm Hg at the gasifier outlet, 0.20 -1.80 mm Hg at the water scrubber outlet and 2.15 -4.60 mm Hg at the filter box outlet. It was found that they were small at the wet product gas flow rate of around 8.145 N m 3 h-1. Experiment was carried out at 800C for coconut shell, ground nut shell and rice husk. The producer gas analysed and the results were plotted in the graph. The percentage of composition of H2, CO, CH4, CO2, and N2 are shown in Fig.3. The calorific values of the producer gas for these three materials were founded out (Table 1). Conclusion Combustible gases can be produced from coconut shell, ground nut shell and rice husk; they were utilized as a feedstock in a down draft gasifier. The calorific value of the coconut shell is more than groundnut shell and rice husk. The coconut shell has more carbon content and producer gas. Vol. 3 No. 1 (Jan 2010) ISSN: 0974- 6846

The hydrogen amount in the producer gas for ground nut shell is more than coconut shell and rice husk. But carbon monoxide is 17.55% and 21.22% higher than ground nut shell and Rice husk in Coconut shell.. Also 6.15% and 38.71% methane are highly present in coconut shell than the other two shells. The coconut shell have more carbon content and also in producer gas. Coconut shell is one of the waste biomass and easily available material. It is the best alternative energy source shell material. References 1. Adnan Midilli, Murat Dogru, Colin R, Howarth Mike, Ling J and Teoman Ayhan (2001) Combustible gas production from sewage sludge with a down draft gasifier. Energy Conservation & Management; 42,157-172. 2. Avdhesh K and Sharma R (2009) Experimental study on 75 KW the downdraft (biomass) gasifier system. Renewable Energy. 34, 1726-1733. 3. Dogru M, Howarth CR, Akay G, Keskinler B and Malik AA (2002) Gasification of hazelnut shells in a down draft gasifier. Energy. 27, 415-427. 4. Khater EMH, El-Ibiary NN, Khattab IA and Hamad MA (1992) Gasification of rice hulls. Biomass & Bioenergy. 3, 329-333. 5. Pattarporn Chairprasert and Tharapong Vitidsant (2009) Promotion of coconut shell gasification by steam reforming on nickle-dolomite. Am. J. Appl. Sci. 6(2), 332-336. 6. Paulo R. Wander and Carlos R (2004) Assessment of a small saw dust gasification unit. Bio mass & Bio Energy. 27, 467-476. 7. Philippe Mathew and Raphael Dubuisson (2002) Performance analysis of a biomass gasfier. Energy Conversion & Management. 43, 1291-1299. 8. Savitri Garivait and Ulonwan Chaiyo (2006) Physical and chemical properties of Thai biomass fuels from agricultural residues. 2nd Intl. Conf. on Sustainable Energy & Environment. 9. Singh RN, Jena U, Patel JB and Sharma AM (2006) Feasibility study of cashew nut shells as an open core gasifier feedstock. Renewable Energy. 31,481487. 10. Van der Drift A, Van Doorm J and Vermeulen JW (2001) Ten residual biomass fuels circulating fluidized-bed gasification. Biomass & Bio-energy. 20, 45-56.

Fig.3. Gas composition of various biomass materials Table 1. Calorific value of various biomass materials
Shell Materials Coconut Shell Groundnut Shell Rice Husk Calorific value Kcal/Kg 5500 4229 3000

Research article
Indian Society for Education and Environment (iSee)

Biomass energy
http://www.indjst.org

Sivakumar & Krishna Mohan

Indian J.Sci.Technol.

SUPERFICIAL VELOCITY - THE KEY TO DOWNDRAFT GASIFICATION1 T. B. Reeda, R. Waltb, S. Ellisc, A. Dasd, S. Deutche
The Biomass Energy Foundation, 1810 Smith Rd., Golden, CO 80401; b Community Power Corporation, Aurora, CO, c Colorado School of Mines, Golden, CO, d Original Sources, Boulder, CO, e The National Renewable Energy Laboratory, Golden, CO.
a

ABSTRACT The superficial velocity (hearth load) of a gasifier is the most important measure of its performance, controlling gas production rate, gas energy content, fuel consumption rate, power output, and char and tar production rate. The superficial velocity, SV, of a gasifier is defined as: SV = Gas Production Rate/Cross Sectional Area = (m3/s)/(m2/s) = m/s It is easily estimated or measured by measuring gas production rate or fuel throughput and gasifier dimensions. It controls the rate at which air, then gas, passes down through a gasifier. This in turn exercises a primary effect on heat transfer around each particle during flaming pyrolysis of the volatiles, combustion of the tars and gasification of the charcoal. A low SV causes relatively slow pyrolysis conditions at around 600C, and produces high yields of charcoal - 20-30%, large quantities of unburned tars, and a gas with high hydrocarbon content and high tar (volatile) content. A high SV causes very fast pyrolysis, producing less than 10% char-ash at 1050 C and hot gases at 1200-1400 C in the flaming pyrolysis zone. These gases then react with the remaining char-ash to yield tars typically less than 1000 ppm, 5-7% char-ash and a producer gas with less energy. These relationships have been investigated in a velocity controlled inverted downdraft gasifier with a 7.5 cm diameter. As the superficial velocity was varied from 0.05 m/s to 0.26m/s , the gas production rate increased from 102 to 679 cm3/s, charcoal production decreased from 13.0% to 4.7% and tar in the gas decreased from 8330 to 300 mg/kg (ppm). At low Superficial Velocities (and low Biot numbers), the particles are heated slowly to pyrolysis temperature and remain essentially isothermal. At high superficial velocities the outside of the particle can be incandescent (> 800C) while the center is still at room temperature. This permits the escaping gases to react with the charcoal, thus reducing the charcoal yield and increasing the gas yield. We call this phenomenon simultaneous pyrolysis and gasification, SPG and believe that it is the fundamental reason why the Superficial Velocity controls all other aspects of gasification. In producing heat the tars in producer gas are a useful fuel, providing no cold surfaces intervene. The inverted downdraft stoves provide potentially simple, clean cooking for developing countries. At low SVs they also produce charcoal. In producing electric power, tars are detrimental to engine operation, and so high SVs must be maintained to minimize tar and char production.
*Presented at 4th Biomass Conference of the Americas; Oakland, CA, 8/29/99

Introduction Clean cooking and distributed power are the major problems of half the world population. Gasification of biomass can help provide solutions to both these problems; unfortunately very little fundamental research has been done on gasification and most new designs are made by guess and by golly. Over a million vehicles used gasified biomass successfully during World War II. Since then there have been many attempts to gasify biomass, most unsuccessful. Biomass gasifiers can be divided into three main categories depending on the source of heat for pyrolysis: Charcoal burning gasifiers (updraft, counter-flow ); tar burning gasifiers (downdraft, inverted downdraft, crossdraft, co-flow, open top, topless, ); and fluidized bed (with some char, some tar burning and many varieties). Due to the very high volatile content of biomass, the tar burning gasifiers are preferred when it is necessary to produce a very clean gas for power generation. This paper deals primarily with tar burning gasifiers, but many of the principles can be applied to other forms. The superficial velocity, SV, of a gasifier is the most fundamental measure of the expected behavior of a gasifier and controls most of the other aspects of gasifier operation as shown here: Gas Production Rate Fuel Consumption Rate Gas Energy Content
SUPERFICIAL VELOCITY

Char Production

Tar Production

It is as important to gasifiers as engine rpm to an engine designer or metabolism rate to a physiologist. It is defined as: SV = Gas Production Rate/Cross Sectional Area = (Nm3/s)/(m2/s) = m/s (It is sometimes defined as the Hearth load, Bh measured in Nm3/cm hr. Bh = 0.36 SV.) It is independent of gasifier size, and so permits comparison of gasifiers of very different dimensions. It is easily estimated or measured from air/oxygen fuel throughput and gasifier dimensions. It controls the rate at which air, then gas, passes through a gasifier. This in turn exercises a primary effect on heat transfer around each particle during flaming pyrolysis of the volatiles, on combustion of the tars and the on the degree of gasification of the charcoal. While the SV is the most important measure of performance for design, it is not generally recognized or discussed.
2

1.

HISTORICAL
In the book, Gengas1 G. V. Nordenswan says:

The concept of hearth load plays a very important role in dimensioning a wood gas generator hearth. The hearth load is the quantity of prepared gas, divided by the smallest

passage area of the hearth. Thus the hearth load is dimensionally a velocity. He goes on to say that the practical range for SV is 0.8m/s (where tars are quite high) to 2.5 m/s (above which charcoal dusting is unacceptable). These numbers only apply to constricted hearth (Imbert) gasifiers. Table 5-2 in Ref. 2 gives design parameters for various Imbert gasifiers ranging from 4 Nm3/hr to 230 Nm3/hr. More recent gasifiers would have other values of SV as shown in Table I. Table I Reported Superficial Velocities in various gasifiers2 GASIFIER TYPE Imbert Biomass Corp. SERI Air SERI Oxygen Syn-Gas air Syn-Gas oxygen Buck (Chern) Rogers PYROL. ZONE SV m/s 0.63 0.24 0.28 0.24 1.71. 1.07 0.23 0.13 CHAR ZONE SV m/s 2.5 0.95 0.28 0.24 1.71. 1.07 0.23 0.13

Buck Rogers (Wallawender)

The values in Table I are taken from literature reports and generally represent the highest values reported there. The Stratified Downdraft gasifier has the same values of SV in the pyrolysis and char zones since it uses a constant diameter gasification tube. Note that the values differ widely from those recommended for Imbert gasifiers.

2.

EXPERIMENTAL

4.1 Apparatus
The apparatus shown in Figure 1 has been built to measure the superficial velocity over a wide range in batch inverted downdraft gasification. (It can be configured with minor modifications to operate in the conventional downdraft mode.) The apparatus is well insulated with pressed fiber refractory (riser sleeve) to minimize heat loss. The air supply is measured with a calibrated flowmeter and the pressure drop, temperature, and gas composition, are measured during the runs. In use the cylinder is filled with fuel to the gas outlet. The runs reported here were made on mixed +1/4, -1/2 inch hardwood chips with a moisture content of 6.1% wet basis. The top

chips are saturated with alcohol and ignited and the cap is put on. The temperature and pressure drop are monitored as a function of time. The gas from the gasifier passes through the collection train shown in Figure 2. The outlet pipe is heated with a torch to prevent tar condensation at startup. The gas then passes through a condenser in an ice bath to cool the gas and remove most of the moisture and tar. At the end of the run the moisture recovered was measured and the condenser was rinsed with acetone to collect the majority of the tar. Residual moisture and tar can be removed after the condenser and total gas flow can be measured with a gas meter at the exit and the gas can be flared. Tars were also measured using a filter to draw aliquot portions directly from the gasifier during the runs. These results and new methods will be reported later. After the run the amount of charcoal remaining was measured. While the experiments were performed on an inverted downdraft gasifier, it is believed that the results will be similar for conventional and new downdraft gasifiers. Further experiments are planned.
Grate
10 cm 5 cm

Cap

Gas Out

Gas Sample, Pressure Char from Pyrolysis

40 cm 30 cm

Pyrolysis Zone

Temperature
Unburned Fuel 4 " Pipe Insulation Pressure

Air In

Figure 1 Gasifier for measuring effects of superficial velocity change.

3.

RESULTS
G as O ut G a s S a m p le , P re s s u r e T e m p e ra tu re P r e s s u re Tap

The results of these measurements at four SVs are shown in Fig. 3 and reported in Table II

F la r e , ESP, F lo w m e te r

4.1 Flaming Pyrolysis

Tar burning gasifiers consume the tars by a process we call flaming pyrolysis.2 The combustion of pyrolysis products in air in sufficient supply (as in a match) is termed flaming combustion. We apply the term flaming pyrolysis to the combustion of the same volatiles in an inadequate supply of air so that the products are largely CO and H2 rather than CO2 and H2O.
G a s ifie r

C o m p re s s e d A ir , F lo w m e te r

0 C Ic e B a th

Fig. 2 - Gas Train

4.2 Charcoal and Tar Yield

It is seen in Table II and Figures 4 and 5 that charcoal and tar yields depend strongly on the superficial velocity. The reduction in charcoal and tar with increasing SV is due to the increased temperature during pyrolysis shown in Figure 3 and Table II. Tars were measured both by total condensate in the condenser and by use of an absolute filter method.3

4.3 Air/Fuel ratio

It is obvious in Table II that many other variables depend on SV. In Table II it can be seen that the air/fuel ratio increases from 1.44 to 5.21 as more and more of the charcoal is gasified and the process approaches complete combustion. Table II Results of operation of gasifier at 0.05, 0.19, 0.26 and 0.44 m/sec SV
Run # 5 4.1 89.1 11.6 16.7 2.044 0.025 33.5 770 1.0 0.052 0.206 102 9941 8330 1.20 1.44 13.0% 6 18.8 91.1 6.8 13.8 1.463 0.11 14.4 993 1.0 0.187 0.423 489 3460 1330 2.08 3.80 7.5% 8 26.0 82.2 3.8 11.8 0.44 0.15 9.5 1033 0.6 0.260 0.387 679 1138 300 3.16 3.47 4.7% 9 46.4 84.2 0 13.8 0.189 0.33 8 1045 0.6 0.437 0.548 1141 346 10 3.75 5.21 0.0%

MEASURED VALUES
Air Flow - l/m Initial Fuel (g) Final Charcoal (g) Condensate (g)

Tar + Particulate (g)

Average Pressure Drop iwc Time of Run min Pyrolysis Temp Pyrolysis time min

DERIVED VALUES
Superficial Velocity - m/s Gas Produced kg (m3) Gas production rate cm3/sec Tar, Particulate in Gas - mg/m3 Tar mg/m3 -Filter Method Fuel Velocity - cm/min Air/Fuel Ratio Charcoal Yield - %

4.4 Gas Quality

The gases have not been analyzed. However, it is expected that the energy content will be very high at low superficial velocities (pyrolytic gas) and progress to lower values as more and more of the charcoal is consumed.
1200 SV= 0.26 1000 SV=0.44 Pyrolysis Time vs Superficial Velocity

800 Temperature - C

600 SV=0.19 400 SV=0.05

4.5 Reaction velocity

The flaming pyrolysis gasification 200 reaction in solids is analogous to the 0 combustion of a gas and proceeds at a 0 5 10 15 20 25 Time - min finite rate. In steady state gasifiers the fuel is consumed at the same rate Figure 3 Temperature Vs time for 4 sample runs as the reaction progresses, so fuel flow velocity is identical to reaction velocity. In the batch gasifier studied here the gasification reaction proceeds through the solid bed of fuel and the time required to pyrolyse all the fuel gives the reaction velocities of 1.2 to 3.8 cm/min.

30

4.6 Gasification-Combustion
Gasification can be viewed as a partial combustion reaction, and the declining energy content of the gas and the increasing air fuel ratio suggests that there may be a continuum between pure pyrolytic gasification, producing gas and charcoal, and pure combustion at high enough air/fuel ratios. Run 9 produced no charcoal and was probably close to complete combustion. measurements are in progress and will be reported separately. Gas

4.7 Time for Pyrolysis

The time required for pyrolysis to charcoal plus volatiles depends on the heat transfer rate and the amount of charcoal produced. The time for pyrolysis shown in Table II is taken as the time necessary for the first rapid temperature rise shown in Figure 3. It is remarkably short, 0.6 to 1 minute for these chips. We were surprised that there was not a steeper rise time at higher superficial velocities. We believe this can be explained by the less complete pyrolysis at lower velocities.

14.0% 12.0%

Char Yield vs Superficial Velocity

12000 10000 8000 6000 4000 2000 0
0.00 0.05

Tar vs Superficial Velocity

Char yield - %

10.0% 8.0% 6.0% 4.0% 2.0% 0.0% 0.00 0.10 0.20 0.30 0.40 0.50
Tar, Particulate mg/m3

Tar+Particulate in Condensor Tar only, Filter method

0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50

Superficial Velocity - m/s

Superficial Velocity - m/s

Figure 4- Charcoal yield Vs SV

Figure 5 Charcoal yield Vs SV

4.

SIMULTANEOUS PYROLYSIS AND GASIFICATION, SPG

In an earlier paper4 we described the rapid pyrolysis of single particle birch dowels 3/8-3/4 in diameter in a high temperature gas flame. We noticed that the charcoal yield was 3.111.7% rather than the 20-25% yield characteristic of slow pyrolysis. We also measured a heat for pyrolysis of 2.9-3.4 kJ/g, much higher than the 1.8-3 kJ/g required for slow pyrolysis. We now believe that this is due to simultaneous pyrolysis and gasification, SPG. In slow pyrolysis the particle remains close to isothermal and the surface temperature does not exceed 400-700C, not high enough for char gasification reactions. However, with fast pyrolysis at higher heat transfer rates, the particle surface temperature can exceed 1000 C while pyrolysis is continuing at the cool core. In this case the exiting gases and volatiles must pass through the surface charcoal and can react to gasify the charcoal and crack the volatiles. Since these are endothermic reactions, it is not surprising that the heat for pyrolysis is so high. We believe SPG is the root cause of reduced tar and charcoal at high superficial velocities. It also explains the success of downdraft gasification in eliminating tars.

5.

IMPLICATIONS FOR GASIFIERS FOR POWER AND STOVES

We said at the beginning of this paper that clean cooking and power were the primary needs of half the world. A great deal of progress is being made today in gasification for both stoves and power. Much of this work is reported on the Internet. In particular there are continuing forums on both stoves and gasification at www.crest.org.

4.1 Gasifier Stoves

Recently there has been a great deal of interest in cooking for developing countries. Wood cooking is inherently inefficient and dirty, but wood-gas can be burned like propane or natural gas to give fast, clean cooking. The inverted downdraft gasifier is a batch process in which a vessel of biomass is ignited at the top and produces a gas as the pyrolysis reaction proceeds to the bottom. This gas can then be mixed with air and burned much more cleanly than in conventional wood cooking. This can be operated with natural draft 5 or forced draft 6 . The apparatus described here is useful for characterizing biomass gasification at these very low SVs.

4.2 Power Gasifiers

The operation, cost and complexity of power gasifiers is very dependent on the tar produced. It is obvious from Figures 4 & 5 that one should operate a gasifier at as high a superficial velocity as possible, consistent with producing a combustible gas, corresponding to 0.26m/s on Figure 3 but not as high as 0.44m/s for these wood chips. We plan to operate this gasifier with different fuels & moisture contents to characterize the optimal velocities for producing gas. Power gasifiers are usually operated in the downdraft mode, but can also be

operated tar-burning updraft.7 The apparatus described here is useful for investigating the various factors affecting power gasifiers at high SV.

6.

ACKNOWLEDGEMENTS

The authors would like to thank their many friends at the Colorado School of Mines and the National Renewable Energy Laboratory who cooperated to help make this paper possible. The work was funded by the Community Power Corporation and the Biomass Energy Foundation.

REFERENCES
1. Generator Gas: The Swedish Experience with Wood-gas 1939-1945 3rd Edition, Reed, T. B. and Jantzen, D., Eds., The Biomass Energy Foundation Press, 1998.

2. Reed, T. B. and Das, A., Handbook of Biomass Gasifier Engine Systems, The Biomass Energy Foundation Press, 1988. 3. Das, A., Contaminant Testing for Gasifier Engine Systems 2nd edition, Original Sources, The Biomass Energy Foundation Press Press, 1999. 4. Reed, T. B. and S. Gaur, The High Heat of Fast Pyrolysis for Large Particles, in Developments in Thermochemical Biomass Conversion, Ed. AA. V. Bridgwater, Blackie Academic Press, 1996. 5. Reed, T. B. and Larson, R., A wood-Gas Stove for Developing Countries, in Developments in Thermochemical Biomass Conversion, Ed. A. V. Bridgwater, Blackie Academic Press, 1996. 6. Reed, T. B. and Walt, R., The Turbo Wood-Gas Stove, paper, this conference. 7. Brand, Stephen, Multi-Fuel Gasification for Energy Production in Rural Areas, in Proceedings, 2nd Biomass Conference of the Americas, Aug. 21, Portland Ore., the National Renewable Energy Laboratory, NREL/CP-200-8098, 1995.