A clean, efficient system for producing Charcoal, Heat and Power ...

25 downloads 24173 Views 660KB Size Report
a Cardiff School of Engineering, Cardiff University, Queens Buildings, The Parade, Newport Road, Cardiff CF24 0YF, UK b BioEnergy Devices, Unit 28, ...
Fuel 85 (2006) 1566–1578 www.fuelfirst.com

A clean, efficient system for producing Charcoal, Heat and Power (CHaP) C. Syred a,*, A.J. Griffiths a, N. Syred a, D. Beedie b, D. James c a

Cardiff School of Engineering, Cardiff University, Queens Buildings, The Parade, Newport Road, Cardiff CF24 0YF, UK b BioEnergy Devices, Unit 28, St Theodores Way, Brynmenyn Industrial Estate, Bridgend CF32 9TZ, UK c James Engineering Turbines Ltd, 5 St Johns Road, Clevedon, Somerset BS21 7TG, UK Received 13 March 2005; received in revised form 12 October 2005; accepted 26 October 2005 Available online 5 December 2005

Abstract There is a strong domestic and industrial market for charcoal in the UK and is still used in many developing countries for cooking and heating as well as for many industrial applications. It is usually made in small-scale simple kilns that are very damaging to the environment, very inefficient and labour intensive. The Charcoal, Heat and Power (CHaP) process offers a method for producing clean efficient charcoal under pressurised conditions and uses the product gas from the carbonisation process to drive a small gas turbine to produce heat and power. The charcoal is produced using waste forestry matter and other waste wood, including that from sustainably managed forests. The CHaP system can also be used in developing countries where there is an excess of forestry waste and a shortage of fossil fuels. The CHaP process was initially designed, developed and a prototype system built. This paper discusses the CHaP design and the various components used, their separate development and integration into a system. Tests showed the process successfully produced a high quality charcoal and the product gas effectively used to drive a gas turbine. The CHaP technology was proven and a new novel system of producing charcoal under pressurised conditions was created coupled with a novel use of the product gas whose output was green heat and power. The initial CHaP prototype showed the process was capable of producing low emissions and is virtually carbon neutral. q 2005 Elsevier Ltd. All rights reserved. Keywords: Charcoal; LCV wood gas; Combustor; Small gas turbine

1. Introduction Long before its development as a fuel, charcoal was used as a drawing medium by artists. Cave paintings made with charcoal have been found, dated to 30,000 years BC. The ‘charcoal’ used here was more likely to be charred sticks from a fire, rather than charcoal produced intentionally. The bronze and iron ages, starting around 5500 years ago, are probably the first use of charcoal as a fuel. Wood could not produce the high temperatures needed to smelt, or reduce the ores, and then to melt the resulting metal in order to cast it. Copper was first reduced with charcoal around 3000 BC, starting the Bronze Age, and around 1200 BC, the Iron Age began. It is possible that the Egyptians also used charcoal in the early development of glass. A by-product of producing charcoal, tar or pitch, was used to waterproof wooden structures, in particular ships, as far back as Roman times. In addition, the pyroligneous acid * Corresponding author. Tel.: C44 29 2087 4318; fax: C44 29 2087 4317. E-mail address: [email protected] (C. Syred).

0016-2361/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2005.10.026

(another by-product of charcoal manufacture) was used by the Egyptians as an embalming material [1]. The production of charcoal involves burning the raw material in an atmosphere free of oxygen (or air) and the earliest method of charcoal production was probably with a pit kiln, positioned in the forest, close to the point of wood collection. This involved digging a shallow, level, pit and stacking the timber to be used longitudinally along the bottom of the pit. The complete pile was covered with vegetation, straw and earth to make an airtight seal around the wood. The wood was lit and the burning allowed to progress from one end of the pit to the other, a process taking around 10–15 days [2]. Further developments led to the classical ‘forest kiln’, a hemispherical woodpile built around a central shaft, which acted as a chimney. Again, the woodpile was covered with soil and turf to shut out the air, and lit by pouring several bucket loads of hot embers down the chimney, which was then sealed. Air supply to the heap was controlled by ensuring that any cracks in the earth covering were repaired and opening or closing purpose-made vents built at the base of the woodpile. The charcoal burner had to attend to the kiln throughout the burn to ensure that maximum charcoal was produced without

C. Syred et al. / Fuel 85 (2006) 1566–1578

the wood being burned to ashes, a process which would take around 10 days. During this carbonisation process, the pile would contract in size as the volatile matter was lost from the wood. Average yield of charcoal from this type of kiln was around 35–40 bushels of charcoal per chord of original wood (i.e. around 35–45% of the original volume) depending on operating conditions and wood-type. One major disadvantage with this method of charcoal production was that a percentage of the feedstock was burned to produce heat in order to power the carbonisation process [1]. 2. Charcoal production developments Improvements to the traditional forest or pit kilns involved building more permanent structures with bricks and more recently, metal. This, however, presented the problem of transporting large amounts of wood from the forest where it was felled to the site of the carbonising facility. Initially, the first development was to replace the forest kiln with a very similar structure built with brick bases, in order that the tar and pyroglineous acid could be collected in pits and put to further use. Later, domed brick kilns were built, which were themselves replaced with cast-iron retorts, where the wood to be carbonised was held in a cylinder separate to the fuel used to provide heat for the process. In this system, a brick-built chamber incorporating a firebox remained hot while the castiron cylinder holding the wood could be rapidly replaced, saving time and heat energy. Quite a number of different designs were produced using this basic design principle, with additions for collecting the tar, acid and wood-gas by-products of the process. During the late 19th and the 20th century, much larger industrial plants were built for larger quantities of charcoal to be produced. Here, the wood and the final charcoal products were held in railway-style wagons, which were pushed on tracks into cast-iron tunnel retorts, and pulled out at the far side when the process was complete. In some designs, the gases produced by the carbonising wood were burned directly in the furnace, reducing the fuel requirements of the system. A large system was developed for refining and treating the by-products of the carbonising process, similar in form to the plant used today for refining oil. A number of large charcoal producing plants were built, incorporating both retorts and refinery processes, enabling both charcoal and many other products to be produced, thus providing the raw materials for a wide range of other processes [2,3]. Modern charcoal production methods have changed little from the traditional forest kiln and remain inefficient, time consuming and environmentally unfriendly with over 60% of process energy loss. An extensive literature review found that although some advances in charcoal production had been made in the last century involving multiple batch loads, new kiln designs, etc., these processes still remain inefficient and time consuming. These modified processes are now no longer used and UK production of charcoal has reverted back to the more traditional kiln methods. Few references could be found to work on charcoal production under pressurised conditions.

1567

Charcoal in Europe is mainly used for the barbeque market, although there are many other uses and the UK imports over 90% of its requirements. Interest is also growing in charcoal as a ‘renewable’ fuel. Developing countries and those short of fossil fuels however, use charcoal as their main cooking fuel as well as for many industrial processes, such as smelting and steel refining. Charcoal can also be ‘activated’ by further refining and in this form is used in filters for water and air. Charcoal can be used for medical purposes, both internally and externally. It is used in sugar refining, agriculture, horticulture, and as an ingredient in animal foodstuffs. Specific charcoals (i.e. those resulting from particular wood species) are used for gunpowder and fuse powders, and also for artist charcoal. The Charcoal, Heat and Power (CHaP) process discussed in this paper offers a cheap, clean and efficient method of producing charcoal with the waste energy being utilised in the production of heat and power. This process can be used in many situations both nationally and internationally. In the UK the CHaP system could be used at forest management sites, also with traditional and urban forestry. The completed system uses wood sustainably derived either from ‘urban forestry’— highway, amenity and domestic tree management operations— or from revitalised deciduous woodlands. It could, if required, utilise wood-chips from ‘energy plantations’ or waste from conventional forestry. In developing countries, the CHaP system could with modifications, use a range of different biofuel and biomass materials. With increasing concerns over climate change and the UKs commitment to increasing green energy, reducing CO2 emissions, the process can make a useful contribution to sustainability. The process can also use a sawdust fed gasifier to provide heat to feed the carbonisation process of the lumpwood. The hot gas (volatiles, tar, etc.) driven off from the wood, combined with the gasifier gas, is then fed into a combustor. This combustor then fires a small gas turbine to produce green heat and power. The whole system is operated under pressure. The CHaP system is thus an attempt to improve the efficiency of the charcoal manufacturing process by utilising available energy to generate electricity and heat efficiently and economically whilst also reducing emissions. The original system had a number of features in order to achieve these aims: 1. As the carbonisation of wood is a cyclic process, a modulated source of heat is required that can serve to preheat the carbonisation vessel. This must be a gasifier to avoid direct combustion of the lump wood intended for charcoal production in the carboniser vessel. This can be provided either by an available design of cyclonic gasifier for sawdust or bio-oil. This means that none of the feedstock intended for charcoal production is burned for heat generation, whilst the evolved low calorific value fuel gas can be blended with those from the gasifier and fed to a special design of combustor. The combustor is used to fire small gas turbine, thus the polluting gases resulting from the carbonisation process will be burned cleanly and used to produce heat and power.

1568

C. Syred et al. / Fuel 85 (2006) 1566–1578

2. The vessel for carbonising the wood will be operated at moderate (gas turbine) pressures (3.2 bar absolute for the pilot unit), higher pressures are envisaged later. 3. As the small robust gas turbine is direct fired to avoid expensive gas cleanup systems, a special combustor had to be evolved to deal with the variable mix of medium to low calorific gases (LCV) from the carboniser and gasifier. This incorporated novel vortex collector pockets (VCPs) to remove and collect ash particles down to 5 mm without the need for additional cyclone collectors in the system. This paper describes the origins of charcoal use and production, and the techniques for producing it. As CHaP uses a pressurised system, results from the available literature on the effects of temperature and pressure on the production of charcoal are also discussed. The CHaP system, a clean, efficient system for the production of charcoal, heat and power is then described and a detailed discussion of the results from a prototype system made. 3. Charcoal quality standards Few standards exist which define the quality of charcoal, particularly for the domestic market. Large industrial users have a much tighter specification for the charcoal they can use, particularly in the metal industry. In the domestic (barbeque) market, no British standard exists whilst in Europe standards exist in Germany, Belgium and France, i.e. the German DIN 51749, the French EP 846E and the Belgian NBN M11-001, respectively. The German standard is quite specific on fixed carbon content, giving a minimum of 78%, and quoting maximum percentages of volatiles, ash and moisture. The

French and Belgium standards define the sizes of charcoal pieces that can be sold to the public: the French quote 85% to be in the 20–120 mm range and Belgium quotes a maximum of 10% below 20 mm and none over 160 mm. In Britain, generally, charcoal is sold in pieces between 20 and 80 mm in size at time of packing. Proximate analysis results of a range of charcoal products resulting from a number of woods, as manufactured by traditional processes are shown in Table 1 [4]. Most charcoals have a carbon content greater than 65% (with the exception of the soft-burned sample) and a volatile matter content less than 26% (again, with the exception of the soft-burned sample). Moisture content is generally below 8% and typically ash is below 3% (although some exceptions exist). Moisture present in the charcoal reduces the calorific or heating value of the charcoal, since energy is required to heat and evaporate the moisture. For comparison, Table 2 shows the characteristics demanded by a steel blast furnace plant in Brazil using charcoal as a fuel. The table shows the range and yearly averages of the charcoal used. The charcoal is a mixture of 40% eucalyptus charcoal produced in company kilns and 60% heterogenous natural wood charcoal manufactured by privately operated kilns. The charcoal considered to be ‘good to excellent’ is that produced from eucalyptus wood in company kilns. 4. A qualitative description of the carbonisation process The process of charcoal manufacture is known as the destructive distillation of wood, and essentially involves heating the wood to a temperature beyond 270 8C in an

Table 1 Typical charcoal [4] Wood species

Production method

Moisture content (%)

Ash (%)

Volatile matter (m.c./%)

Fixed carbon (%)

Bulk density (raw) (kg/m3)

Bulk density (pulverised) (kg/m3)

Gross calorific value (oven dry basis) (kJ/kg)

Remarks

Dakama

Earth pit

7.5

1.4

16.9

74.2

314

708

32,410

Wallaba

Earth pit

6.9

1.3

14.7

77.1

261

261

35,580

Kautaballi

Earth pit

6.6

3.0

24.8

65.6

290

290

29,990

Mixed tropical hardwood Mixed tropical hardwood Wallaba

Earth pit

5.4

8.9

17.1

68.6

Earth pit

5.4

1.2

23.6

69.8

Earth mound Earth mound Portable steel kiln Portable steel kiln Retort

5.9

1.3

8.5

84.2

5.8

0.7

46.0

47.6

3.5

2.1

13.3

81.1

32 500

Pulverised fuel for rotary kilns (1) Pulverised fuel for rotary kilns (1) Pulverised fuel for rotary kilns (1) Low grade charcoal fines (1) Domestic charcoal (1) Well-burned sample (1) Soft-burned sample (1) (2)

4.0

1.5

13.5

83.0

30 140

(4)

5.1

2.6

25.8

66.8

Wallaba Oak Coconut shells Eucalyptus saligna

Key: (1) Guyana, (2) UK, (3) Brazil, (4) Fiji.

(3)

C. Syred et al. / Fuel 85 (2006) 1566–1578 Table 2 Characteristics of charcoal for a Brazilian blast furnace [4] Chemical and physical composition of charcoal (dry basis) (by weight)

Max

Min

Yearly average

Charcoal considered good to excellent

Carbon (%) Ash (%) Volatile matter (%) Bulk density—as received (kg/m3) Bulk density (dry) (kg/m3) Average size—as received (mm) Fines content—as received (!6.35 mm) (%) Moisture content—as received (%)

80 10 26 330

60 3 15 200

70 5 25 260

75–80 3–4 20–25 250–300

270 60

180 10

235 35

230–270 20–50

22

10

15

Max 10

25

5

10

Max 10

1569

(vi) Temperature rises from 290 to 400 8C: Further breakdown of the wood allows a number of gases to be released such as carbon monoxide, carbon dioxide, hydrogen and methane, in addition to condensable vapours such as water, acetic acid, methanol, and acetone. Wood tars begin to predominate as the temperature rises further. (vii) Temperature levels around 400–600 8C: The main process of carbonisation is complete, and the charcoal is known as ‘soft-burned’. This type of charcoal can contain up to 30% weight of tar, trapped in the internal structure of the material. Further heating drives off more of the tar and increases the fixed carbon content of the final product.

6. The effects of temperature and pressure on the products of carbonisation oxygen-free environment. This breaks down the complex cellulose and hemicellulose molecules mainly into H2O, CO, CO2, and char (solid carbon). The process of carbonisation is generally described in terms of ‘Primary’ reactions and ‘Secondary’ reactions. Primary reactions are conversions of the basic wood constituents to products including gases, liquid tars and solid char, whereas secondary reactions reduce the products of the primary reactions (in particular, the tars) to lighter fractions and result mostly in gases. 5. Temperature–time characterisation The production can be typically described as a three-stage process: 1. Drying the wood to expel all remaining moisture. 2. Raising the temperature of the oven dry wood to 270 8C. At this point, the wood begins to decompose, and an endothermic reaction with spontaneous pyrolysis begins. 3. Final heating to 500–600 8C to drive off tar and increase the fixed carbon content to an acceptable level. These three stages can be further refined into the following five stages: (i) Temperature rises from 20 to 110 8C: Wood absorbs heat energy, and releases water vapour. (ii) Temperature will remain at or slightly above 100 8C until all moisture is driven off (bone dry). (iii) Temperature rises from 110 to 270 8C: Wood starts to decompose, releasing some gases such as carbon monoxide and carbon dioxide, and liquids, such as acetic acid and methanol. (iv) Temperature rises from 270 to 290 8C: Endothermic reaction commences in the wood. (v) Temperatures remain above 270 8C. This allows the further breakdown of the wood to occur spontaneously, provided that the temperature of the wood is not cooled below 270 8C.

Some research studies provided experimental results detailing the effects of temperature and pressure on the gaseous and liquid products of carbonisation, and are shown in Fig. 1. Sadakata et al. [5] used mulberry wood in a laboratory scale experiment, rapidly heating the wood at over 1000 8C minK1. Although the CHaP apparatus will not be capable of heating the wood feedstock at this rate, their results provide some useful trends. Fig. 1(a) shows the temperature effects of the decomposition products during wood carbonisation. In general, the graph indicates that the gas products increase while the solid char products decrease. The fraction of condensed liquids and tars appear to decrease slightly with increased temperature, although this seems only to have a significant effect when the temperature rises above 600 8C, above the operating temperature of the CHaP carboniser, and therefore beyond the scope of this project. Zaror and Pyle [6] collected data from a ‘slow’ pyrolysis process (in contrast to the results shown by Sadakata et al. [5]). Fig. 1(b) shows the effect of final pyrolysis temperature on charcoal yield. The graph suggests a decrease of solid charcoal production with increasing temperature, with a corresponding increase in the carbon content of the solid fraction. This supports the results shown in Fig. 1(a) where increasing temperature causes an increase in the gaseous products and therefore a corresponding decrease in the solid products. The gas emitted from the wood during carbonisation (termed ‘wood gas’) is a mixture of a number of products. Fig. 1(c) shows the relationship between these component gases and process temperature. Fig. 1(d) shows the variation of the calorific value of the wood gas with temperature, as given by two different sources [5,7] and illustrates the range of wood gas calorific values that may be expected at a specific temperature. Antal et al. [8] used small amounts of biomass (w1 kg) in experiments to determine the effects of pressure on the charcoal process. The design of the CHaP process requires that the carbonisation vessel operates at an elevated pressure of 3 bar absolute, and these results provide an indication of

1570

C. Syred et al. / Fuel 85 (2006) 1566–1578

Fig. 1. Temperature effects on charcoal process [5,7]; (a) temperature effect on wood, (b) final pyrolysis temperature on charcoal yield, (c) component gas relationship with temperature, and (d) calorific value of wood gas with varying temperatures

the likely effect of pressure on charcoal production. Fig. 2(a) shows a comparison of charcoal yield when operating a carbonisation system at pressures 1 and 10 bar. The results clearly show the charcoal yield is significantly increased with pressure for all wood types. Fig. 2(b)–(d) shows the effect of

different pressures on charcoal volatile matter, fixed carbon content and ash content respectively. Softwoods (Pine and Spruce) showed a decrease in volatile matter, an increase in fixed carbon content and an increase in ash content at the elevated pressure. Hardwoods (Alder and Oak) showed

Fig. 2. Effect of pressure on various wood species [8]; (a) charcoal yield, (b) volatile matter content, (c) fixed carbon content, and (d) ash content.

C. Syred et al. / Fuel 85 (2006) 1566–1578

an increase of volatile matter, a decrease of fixed carbon content and a decrease in ash content at the elevated pressure. Birch wood, classified a hard wood, however does not follow the trend of the other hardwoods and shows very small decrease on volatile matter with elevated pressure with a increase of fixed carbon content and a decrease of ash. Apart from Antal et al. [8–10] very few studies have been found which investigate the effects of elevated pressure on the results of the carbonisation process. These results suggest that the CHaP system will increase the yield of charcoal compared to atmospheric processes whilst maintaining acceptable charcoal quality. A detail review of the production and properties of charcoal is given by Antal and Gronli [10]. Antal et al. [11] also investigated flash carbonisation of a fixed bed of biomass to form charcoal and gas to utilise their green waste. This work is ongoing and no literature could be found on modern techniques for utilising the gas produced from the carbonisation process. The CHaP project thus offers a very novel and efficient process that can effectively utilise the process gas from carbonisation to produce green heat and power. It can be seen from the literature review that the Charcoal, Heat and Power (CHaP) system is a further improvement in the development of the charcoal manufacturing process. This system offers the possibility of manufacturing charcoal with a lower environmental impact, higher yield, as well as simultaneously producing heat and electrical power. Charcoal can be considered a renewable fuel, capable of producing the high temperatures required of many industrial processes. It is used in many parts of the world both for domestic cooking and heating, as well as an industrial fuel.

1571

7. CHaP design The CHaP system uses four major subsystems and is shown in Fig. 3 in schematic form. (1) The carbonisation vessel and its ejector/flow recirculator. (2) A secondary combustor capable of running on dual fuels, oil (for start-up/shut-down or during certain operating periods) and LCV gas. (3) The combined support fuel-gas supply and carboniser heat source. (4) The gas-turbine based ‘turbo-alternator’ unit. The turbine is initially spun by the alternator to a selfsustaining speed. The combustors oil burner is then started and compressed air is supplied to it from the turbine compressor, through an air manifold and control valve 1 (CV 1), Fig. 3. Additional secondary air is then supplied through CV 2. Compressed air is then supplied through control valves 3, 4a, 4b, which is fed into the combustor pressure vessel and cools the combustors surface. This air then mixes with the combustor exhaust gas and reduces it temperatures so it is suitable for firing into the turbine. Once full speed is reached, the combustor is stabilised, and the turbine inlet temperatures are reasonably constant then control valves 5 and 6 are open to initially warm the carbonisation vessel and lump wood. After a predetermined time, the gasifier is then turned on to provide heat to the lump wood and start the carbonisation process. The gasifier gas and carbonisation waste gas are then fed directly to the combustor. As this gas enters the combustor, the combustor oil burner flow rate is turned down automatically by a control system.

Fig. 3. Schematic of the CHaP system.

1572

C. Syred et al. / Fuel 85 (2006) 1566–1578

Fig. 4. Schematic of the ejector-carboniser system showing the operating principle.

† Generation of non-premixed or diffusion flame to exclude the danger of flashback. † Maintaining high efficiencies whilst giving low NOx and CO. † The necessity of using larger fuel nozzles and swirlers to handle the higher fuel gas volume. † Issues of fuel quality restrictions such as hydrogen content, particulates, alkalis, heavy metals, tars, fuel gas temperature, etc. † The issue addressed in this paper are of redesigning the combustor to avoid any drop in efficiency by essentially increasing available residence time, whilst simultaneously dealing with the contaminants in the LCV gas. There is a wide range of work in this area as discussed in the literature [12–17] where the issues raised above are more fully discussed. These combustor designs are conventionally derived from conventional gas turbine combustor systems fired on conventional liquid fuels or natural gas. They are all designed to be fired on cleaned bio-gas, this arises from the type of turbine equipment used with sophisticated turbine blades incorporating numerous fine cooling passages susceptible to blockage. Conversely, CHaP addresses a different problem involved with small-scale power systems. Here, gas turbine systems are generally of simpler construction with un-cooled turbine blades and can sustain modest levels of fine particulates less than 5 mm in size. Indeed, some small turbine systems are derived directly from turbochargers. Turbine inlet temperatures are up to 900 8C. Low pressure drop across the system, low

emissions and good flame stabilisation are also necessary requirements of the system. The next section of this paper describes the design and development of the CHaP process and its main components. Initial tests performed on the system are also described. 7.1. Carbonisation vessel and its ejector/flow recirculator The Carboniser vessel and ejector recirculator was a main component of the CHaP system. The carbonisation vessel holds lump wood under pressure in a flow of hot oxygen-deficient gas (generated separately by a gasifier). Hot gas is recirculated around the vessel by the use of an ejector and flow recirculator. As the hot gas passes over the wood, the pyrolysis process starts, volatiles are driven off, and charcoal forms. The volatiles given off in this carbonisation process enrich the hot gas and raise its calorific value. Fig. 4 shows the design of the carbonisation and recirculator vessel. A tenth scale prototype of the carboniser and ejector was initially built and isothermal and hot gas tests performed under atmospheric conditions with conditions representative of those under pressurised conditions. The performance of the ejector was maximised by testing isothermally and determining the optimum position of the nozzle in relation to the carbonisation vessel outlet port 900 800

Temperature (°C)

The combustor is an integral part of the CHaP system and must be capable of fully burning dual fuels, the LCV gas produced from the carbonisation and gasification processes and a range of supporting fuels (initially oil). The combustor must have good heat storage capacity, produce low emissions and fully burn out any tars remaining in the flow. Several studies have been undertaken to develop LCV and dual fuel gas turbine combustors. Problems encountered are numerous and include:

700

inlet

600 Port 1

500 400

Port 2

300

Outlet

200

Centre of carboniser

100 0 0

100

200

300

Time (min)

Fig. 5. Temperature with time for carbonisation vessel when full of Lump wood.

C. Syred et al. / Fuel 85 (2006) 1566–1578

Fig. 6. Temperature profiles across combustor.

(Fig. 4, port 1), to achieve the maximum recirculation ratio. The optimum position of the nozzle was in line with port 1 exit. Tests showed the successful recirculation of the hot gas giving a minimum recirculation ratio of 2 to 1. CFD modelling of the system had been initially performed and matched the experimental results closely. The LCV gasifier gas was simulated using diluted natural gas and tests undertaken to study the performance of the system for charcoal production. Fig. 5 shows temperatures during a test with the carbonisation vessel full of wood. The inlet temperatures are similar to those expected from the gasification process [20] and thus are appropriate to simulate conditions that would occur in the complete CHaP system. Temperatures inside the carbonisation vessel were seen to slowly rise through the process and peak around 600 8C. This process follows the carbonisation process described earlier in this paper, and tars and other volatiles should be driven of by this peak temperature and an acceptable level of fixed carbon achieved. Proximate analysis was performed on the charcoal produced and the fixed carbon content was 79.17%, moisture content 2.87%, volatiles 16.57% and ash 1.37%. These results show the charcoal produced corresponds to a good quality charcoal that is similar in characteristics to that produced from more traditional methods, Table 1. 7.2. Combustor A cyclonic type combustor was chosen as previous research showed robustness, stable flow and uniform outlet conditions could be achieved through this type of design. A novel tangential outlet would minimise pressure drop and create uniform exit conditions. The combustor had two tangential inlets besides the novel tangential outlet. An oil atomiser and combustor can, originally used to fire the Rover gas turbine, was attached to one of the inlet pipes and secondary air to the cyclone combustion chamber was supplied through the second inlet. The secondary inlet was also capable of supplying product gas from a gasifier. A programme of testing was carried out to characterise the combustor design and the oil burner operating conditions. The viability of using LCV fuel gas from the carboniser/gasifier (feeding the ejector) in the combustor had also to be established, and the turn down ratios

1573

with the various fuels determined. The final design was to operate at a maximum output of 515 kW at a pressure of 3.2 bar. Thus atmospheric tests on the prototype combustor were run up to an output of 200 kW with conditions representative of those under pressurised conditions. The combustor prototype was successfully matched to the Rover gas turbine oil atomiser using kerosene. The tangential inlets created a stable, strongly swirling flow that gives good mixing and burn out rates. The combustor could be run over a range of operating conditions from 50 to 200 kW, with varying air/fuel ratios. Output from a gasifier was introduced to the combustor and successfully operated with both fuels. The oil burner flow rate could be turned down whilst keeping the gasifier flow rate constant and maintaining a stable flame. CFD modelling using the package Fluent 6 was initially performed on the prototype combustor, inlet and exit temperatures and emissions closely matched those measured experimentally. The model was created with a vortex collector pocket (VCP) at the outlet to collect fine particulates and a central drop out pot for larger materials. The discrete phase model was used to inject particles into the combustor to simulate those occurring in the gas and investigate their capture. The model showed the combustor was capable of removing particles above 5 mm from the flow. This work also supported this design process in that it identified the optimum position for the tangential multi-fuel inlets as well as the position of the VCP relative to the inlets and outlet. The successful testing and modelling of the prototype combustor confirmed the suitability of its general design and the capability of burning dual fuels. The final design of the combustor was tested against the Fluent predictions. This showed the combustor ran well on LCV gas and oil as well as a combination of these, producing a stable swirling flow with good mixing and burnout, with early combustion initiated near the inlets (Fig. 6). This was achieved with low pressure drop across the combustor. A detailed discussion of the combustor modelling is available in then literature [18,19]. The vortex collector pocket (VCP) is positioned just before the outlet such that the flow is drawn past the VCP as it is forced into the tangential off-take that forms the exhaust. This mechanism causes most fine particles to be projected into it. Fig. 7 shows

Fig. 7. Particles collected in combustor for (a) conical bottom section and (b) VCP.

1574

C. Syred et al. / Fuel 85 (2006) 1566–1578

Tests on the full-scale combustor were performed at atmospheric conditions with inlet parameters representative of those under pressure. The tests successfully proved the combustor could run under a range of operating conditions whilst maintaining a stable flame and uniform exit conditions, as well as maintaining relatively low emissions. The oil burner had a high turn down ratio (10 to 1) and maintained stable conditions with varying equivalence ratios. LCV gasifier gas combined with oil was successfully burnt in the combustor producing a stable flame and uniform exit conditions. 7.3. The combined support fuel-gas supply and carboniser heat source

Fig. 8. Full-scale combustor.

particle trajectories inside the combustor. Larger particles are collected in the bottom conical section of the combustor, and smaller particles collected in the VCP. Most of the particles that escape the combustor are less than 5 mm. This size of particles in the exhaust gas is an acceptable value for direct feed into the gas turbine. A full-scale cyclonic combustor was therefore constructed to the required specification and inlet conditions, Fig. 8. The combustor was designed with three tangential inlets, an air inlet, a high CV fuel inlet for support fuel and a low CV gas inlet. The combustor was operated at a maximum thermal input of 500 kW. The combustor was mounted vertically and with the cone section at the base collecting larger particles from the flow. The combustor was designed with a long chamber to allow flame movement axially with varying thermal input and quality whilst giving sufficient residence time for complete fuel burnout and thus low emissions. The central section of the combustor was refractory lined allowing substantial heat storage capacity helping to create stable flames. The tangential off take gives low system pressure drop whilst forces the exhausting flow tangentially across a VCP aperture, hence increasing separation capability. The combustor fires a Rover Gas turbine operating at an inlet temperature of 800 8C, which later will be extended to 900 8C. The exhaust gas of the combustor was at higher temperatures than this, and was diluted by a co-flowing air stream that was passed through a jacket surrounding the combustor. This co-flow air lowers the temperature of the exhaust gas and acts as a diffuser to the flow, lowering the pressure drop across the combustor. The VCP removes fine particle above 5 mm, which if carried through the exhaust could damage the turbine. It also removes the need for a separate cyclone separator to remove the particles, which would increase the pressure drop across the system further.

The last main component of the CHaP system is the support fuel gas supply and carboniser heat source. A gasifier acts as a source of support fuel-gas; this gas also acts as a heat-source to drive carbonisation. Support fuel-gas is required to augment the cyclically varying thermal output of the carboniser and maintain a near-constant level of total gas thermal input to the combustor and gas turbine. During the middle of the carbonisation cycle, when carbonisation is occurring most rapidly, carboniser-gas provides the main fuelling for the combustor (the initial design point being 70% of total gas thermal input). At the start and finish of the carbonisation cycle, little energy is contained in the carboniser-gas and the support fuel-gas provides the entire gas thermal input to the combustor and gas turbine. The support fuel-gas also contains the extra energy required to raise the temperatures of the various thermal inertias within the gas generation system. A pressurised design of an inverted, sawdust fired, cyclone gasifier previously tested at Cardiff University [20] was to be developed. However, due to the feedstock delivery problems and time constraints other solutions had to be adopted, namely a bio-oil gasifier. This produces the required hot oxygen-free gas and is relatively much easier to engineer as fuel injection and ignition may be accomplished by fairly standard fuel-oil injection and ignition systems. Sub-automotive-grade biodiesel is a readily available, clean and carbon-neutral fuel-oil. It is also a direct substitute for fossil-derived fuel-oil support fuels for the main combustor. Thus a gasifier feedstock change to fuel-oil had the beneficial side-effect of enabling elimination of all requirements for fossil-derived fuels in the CHaP process. The sawdust fired gasifier is being currently developed and will eventually replace the oil fired version. The bio-diesel gasifier is of similar configuration to the combustor, that is, a tangential-inlet, tangential-outlet, single swirl chamber. It is single-skinned and fully refractory-lined. The fuel oil is injected into the gasifier using similar components to those employed on the support fuel inlet to the combustor. 7.4. The gas turbine unit A turbo-alternator system based on a kerosene-fired Rover derived unit manufactured by Lucas for military auxiliary powered purposes was supplied by James Engineering

C. Syred et al. / Fuel 85 (2006) 1566–1578

1575

Turbines (JET) Ltd. It had previously been demonstrated by JET to run with an alternative external combustion system fired by up to 50% sawdust augmented by kerosene. 8. System analysis Design ranges for thermal inputs to the combustor were established (figures given for turbine at full load): Effluent from carboniser: 0–70% of gas fuel (0–318 kW). Gasifier gas (output): 30–100% of gas fuel (136–454 kW). Combustor support fuel: !20% of total (100 kW). These targets arose from the following considerations: † Cyclonic sawdust gasifier thermal output is controllable in a wide turndown range, expected to be approximately 100– 500 kW in CHaP conditions. † The gasifier response is slow relative to the turboalternator’s acceleration response to input energy variations. † The carboniser energy output will contain both slow and fast components: slow corresponding to the batch timescale; fast due to wood settling and wood fracturing transients. † It is preferable to run the turbo-alternator at maximum power as this produces the highest electrical output. † The input energy to the turbo-alternator must be regulated to, and not exceed the maximum to prevent turbine overspeed. † Since the magnitude of carboniser output energy variation is unpredictable, for precautionary purposes in this first CHaP prototype, a fast response control on the turboalternator input energy is needed of sufficient magnitude to counter transient increases in carboniser output. The Rover gas turbine’s original kerosene injection system has a sufficiently fast response and an existing proven control system based on turbine speed and jet pipe temperature. † A nominal power level of 100 kW by the support fuel oil was the target. Modulation of this power level would compensate for variation in the gasifier-gas power level control. † The maximum kerosene consumption of the original Rover turbine’s combustor was 42.7 kg/h corresponding to 514 kW thermal (net CV basis). Additional estimated heat fluxes are heat losses of gasifier and combustor (25 and 40 kW, respectively) and maximum power absorbed by thermal inertia of wood (47 kW). With the adoption of a bio-diesel gasifier to solve the fuel feeding problems encountered with the sawdust gasifier, a much lower turndown needed to be factored into the system design. The effect was to transfer the main modulation requirement to the combustor’s support fuel burner (to which it is well suited). In a commercial CHaP system this would not be required as multiple carbonisation retorts would be phased so as to generate a near-constant production rate of effluent gas energy, minimising support fuel requirement.

Fig. 9. Complete CHaP system.

The full-scale CHaP rig was designed and manufactured as shown in Fig. 9. The combustor was operated at a maximum thermal input of 550 kW, with the gasifier rated at a maximum 200 kW. All components of the CHaP system were designed and pressure tested to appropriate standards. The full-scale combustor was placed in a pressure vessel that had cooling inlets direct at the outer walls of the combustor to cool hot spots identified during atmospheric tests. This cooling air then acts as dilution air at the combustor exit. Testing at elevated pressure occupied two phases. The first focussed on that part of the gas circuit comprising the gas turbine and the combustor; the second phase covered the complete gas circuit (including the gasifier and carboniser). The first stage of the testing involved the combustor coupled to the gas turbine. An isolation valve was positioned just before the inlet for the LCV carboniser gas which was closed during initial combustor/turbine tests. An initial proving test with the gas turbine, prior to connecting it to the new combustor, showed that the turbine was performing as expected. Having installed the full-scale combustor, further tests were carried out to check the control and instrumentation systems. The turbine was started using its auxiliary motor and the combustor then fired on gas oil and the stability of the system monitored. Pitot tubes and thermocouples were position throughout the rig and temperature, pressure drops and flow rates were measured across the rig. The flow rate of fuel was regulated by a control system. 8.1. Complete pressurised CHaP system testing The complete CHaP cycle was tested initially with the carboniser empty. The system had a compressor air manifold (Fig. 3) with seven air outlets, two for the combustor primary air and secondary air, three for the combustor cooling and dilution air and two for the gasifier primary and secondary air. An important consideration was the system instrumentation. Mass flow measurements were taken at each of the compressor manifolds air valves and the bell-mouth inlet of the compressor as well as oil flow rates for both oil burners. Pressure transducers were used across the ejector/carboniser, and the

1576

C. Syred et al. / Fuel 85 (2006) 1566–1578

pressure difference measured across the compressor exit and turbine inlet. Thermocouples were used to measure temperatures in the compressor exit and manifold, the combustor, at the system exhaust and turbine inlet as well as measurements in the gasifier and carboniser. 8.2. CHaP testing with lump wood The carboniser vessel was filled with wood and the CHaP system was prepared for a full run. The gas turbine was started by the auxiliary motor and then switched over to the oil burner. The combustor was initially fired using the oil burner to allow the combustors thermal mass to heat up. After w12 min a steady combustor exit temperature was reached of 850 8C and a turbine inlet temperature of 720 8C. The turbine jet pipe exit temperature stabilised at 400 8C. At this point the air valves from the compressor manifold to the gasifier burners air inlet was opened allowing the compressors hot air (at w200 8C) to circulate around the carboniser vessel and heat up the lump wood. The LCV gas inlet valve to the combustor was then opened. The combustor exit and turbine inlet temperatures then stabilised at 800 and 700 8C, respectively. The jet pipe exit temperature remained constant at 400 8C. The system remained stable and the carboniser inlet temperature increased to around 120 8C. After 35 min, the gasifier was turned on to provide heat to the carboniser vessel and the gasifier secondary air inlet opened. The gasifier was turned down and the air valves closed to the gasifier and combustor. The lump wood slowly pyrolised and a LCV gas given off. This gas was fed into the combustor and burnt. A stable turbine inlet temperature was maintained by a control system that controlled the combustor oil inlet flow. As more LCV gas entered the combustor the combustor oil burner flow rate was decreased automatically. A stable combustor exit temperature and turbine inlet temperature of 820 and 700 8C, respectively, was maintained. The combustor oil burner flow rate was turned down from a full load of 10.5–1.5 g/s when the LCV gas was at its maximum safe output. The output from the carboniser was maintained at a stable rate by controlling the gasifier. To control the system the gasifier needed to be switched on and off several times throughout the cycle to achieve steady carbonisation and control the amount of LCV gas produced to maintain stable combustor conditions. After typically 3 h and 40 min, the system was turned down when no further gas was produced from the carboniser. This corresponded to previous calculations as to the length of the carbonisation process. (The system is a batch process and for commercialisation a second carboniser would be used with a switch over valve, maintaining continuous turbine use.) The carboniser was opened and examined, charcoal had been produced. Proximate analysis of the charcoal was performed, and showed the process made a high quality product. The lump wood was weighed before and after carbonisation and gave a yield of 38%. Overall mass balance results from the various fuels used in the system showed the wood produced a gas giving a calorific value of approximately

Fig. 10. RPM and oil flow with time.

9.8 MJ/kg. Emissions from the turbine were monitored throughout the tests. NOx levels remained at approximately 80 ppm and could have been reduced by detailed attention to the secondary combustor. CO levels were generally low w10 ppm throughout the process except when the combustor oil burner and gasifier oil burner was being ignited or extinguished. This is due to large fluctuations in flame temperatures and air to fuel ratios resulting in momentary incomplete combustion. As to be expected when the gasifier was switched on and off these levels rose to several hundred parts per million, and gradually dropped back to a low level once the system had restabilised. Fig. 10 shows the turbine speed and combustor oil flow rate throughout the run. The turbine reaches it full speed after 2 min where it becomes stable. Throughout the run the turbine speed was reasonably constant with small increases when gasifier was modulated. The oil flow rate to the combustor was monitored by a control system attached to the turbine. This was controlled by the turbine inlet temperature. As LCV wood gas was introduced into the combustor the oil flow rate drops to maintain similar exit conditions, this is seen from the fluctuations in oil flow rate in Fig. 10. Fig. 11 shows temperatures in the CHaP cycle whilst running. The compressor exit temperatures remain reasonably constant (w200 8C) throughout the run. The temperatures of the combustor products from the oil fired burner at inlet to the main combustor was varied across the run (average 1350 8C)

Fig. 11. Temperatures in CHaP cycle.

C. Syred et al. / Fuel 85 (2006) 1566–1578

1577

11. Conclusions

Fig. 12. Pressure difference across CHaP system.

as this helps maintain reasonably uniform exit conditions. The combustor exit temperature was reasonably constant around 800 8C. This exit gas was diluted with the cooling air inside the combustor pressure vessel and reduces the turbine inlet temperature to approximately 700 8C (this will be increased later). Minimising pressure drop across the system was an important design consideration as this has considerable impact on the system efficiency and turbine performance. Fig. 12 shows the pressure difference between the compressor outlet and turbine inlet. An average pressure drop of 0.2 bar is seen, which is an acceptable value for the turbine rating. Due to time limitations the system was run only for a few hours and future testing and validation are ongoing. 9. Charcoal production The dried wood was weighed before and after the CHaP process. Successful carbonisation of the wood occurred and the charcoal left gave a yield of 38%. Calculations from mass flow of fuels into combustor and gasifier showed the combined hot wood gas had a calorific value of approximately 9.8 MJ/kg. Proximate analysis showed a high quality charcoal was produced. 10. Effects of volatilised alkali salts The effects of volatilised alkali salts from wood combustion are well known [17,20] as are the problems of their condensation on hot surfaces such as turbine blades. However, in the system described, rugged turbine systems with uncooled blades are used which are much more tolerant to deposition than conventional turbines. The low inlet temperature of the turbine also assists in this matter. Other work has shown that typically O50% of the alkali salts condense on fine particulates generated and are removed by the VCP and other particle collectors in the system, this reducing turbine problems. Compared to experiences with pressurised fluidised bed power generation with coal [14] where direct firing of gas turbines is also used with gas clean up via two stages of cyclone dust separators, the fuel gases produced by this work appear to be no less deleterious and thus suitable for direct firing of appropriate gas turbines with uncooled turbines blades.

The main objectives of the project were to research a novel energy technology and create a prototype machine for clean and environmentally benign small-scale conversion of wood to charcoal, heat and power. This new novel technology was successfully researched, developed and commissioned in the given time constraint and successfully demonstrating the CHaP process and its commercial feasibility. The CHaP project has developed a clean and efficient system to produce charcoal heat and power. No research could be found on pressurised charcoal production and harnessing the energy produced in the associated LCV gas. Similarly no existing, gas turbine combustors were capable of efficiently burning the range of fuel inputs and or simultaneously removing fine particulates from the gas stream to minimise damage to the turbine. This research will have a significant impact in producing efficiently low cost charcoal and electricity, for the right application as well as waste heat. There will be clear benefits both nationally and internationally in producing a more sustainable environment. The next phase of the CHaP project is to attract the interest of companies and commercialise the system. Several local companies have shown interest in installing such a system. Commercial systems will use much more modern designs of gas turbines with un-cooled turbine blades to permit direct firing. Because of the low pressure ratios of many designs of small gas turbines a heat exchanger can be inserted between the turbine outlet and compressor outlet to recover heat and improve cycle efficiency by a predicted 8–10%. Development will also be continued to produce a pressurised sawdust fed gasifier to replace the oil gasifier. As well as promoting the system nationally the CHaP system will have great commercial benefits in developing countries that have vast supplies of waste wood and significant markets for charcoal. The system will therefore be promoted both in Europe, North America and developing countries where there are significant supplies of appropriate wood. Acknowledgements The authors would like to thank EPSRC, contract GR/N16587/01, and the DTI for their support in this project. References [1] Kelley DW. Charcoal and charcoal burning. Bukinghamshire: Shire Publications Ltd; 1986. [2] Crumrin T. Fuel for the fires: charcoal making in the 19th century. Chronicle of the early American industries association, vol. 47, June 1994; 1996, !http://www.connerprairie.org/historyonline/fuel.htmlO [3] Churchouse AH. Wood distillation. In: Heilbron IM, editor. Thorpe’s dictionary of applied chemistry. London: Longman’s, Green and Co; 1954. [4] Anon. Industrial charcoal making. Food and Agriculture Organization of the United Nations; 1985.

1578

C. Syred et al. / Fuel 85 (2006) 1566–1578

[5] Sadakata M, Takahashi K, Saito M, Sakai T. Production of fuel gas and char from wood, liginin and holocellulose by carbonization. Fuel 1987; 66:1667–71. [6] Zaror CA, Pyle DL. The pyrolysis of biomass: a general review Wood heat for cooking, Indian Academy of Sciences. London: Macmillian Press; 1983. [7] Fagbemi L, Khezami L, Capart R. Pyrolysis products from different biomasses: application to the thermal cracking of tar. Appl Energy 2001; 69:293–306. [8] Antal Jr MJ, Allen SG, Dai X, Shimizu B, Tam MS, Gronli M. Attainment of the theoretical yield of carbon from biomass. Ind Eng Chem Res 2000; 39(11):4024–31. [9] Antal MJ. Biomass pyrolysis: a review of the literature part 1— carbohydrate pyrolysis Advances in solar energy. Colorado: American Solar Energy Society Inc; 1982. [10] Antal Jr MJ, Gronli M. The art, science and technology of charcoal production. Ind Eng Chem Res 2003;42:1619–40. [11] Antal Jr MJ, Mochidzuki K, Paredes LS. Flash carbonisation of biomass. Ind Eng Chem Res 2003;42:3690–9. [12] Adouane B, Hoppesteyn P, De Jong W, Van Der Wel M, Hein KRG, Spleithoff H. Gas turbine combustor for biomass derived LCV gas, a first approach towards fuel-NOx modelling and experimental validation. Appl Therm Eng 2002;22(8):959–70.

[13] Beer JM. Clean combustion in gas turbines: challenges and technical responses-a review. J Inst Energy 1995;67:2–10. [14] Beer JM, Garland. A coal fuelled combustion turbine cogeneration system with topping combustion. J Eng Gas Turb Power 1997;119(1): 84–92. [15] Hasegawa T. Study of low Nox in medium-btu fuelled 13008C—class gas turbine combustor in IGCC. ASME International gas turbine and Aeroengine congress and exibition, 98-Gt-331, Stockholm, Sweden; 1998. [16] Hoppesteyn PDJ, De Jong W, Andries J, Hein KRG. Coal gasification and combustion of LCV gas. Bioresour Technol 1998;65(1–2):105–15. [17] Nakata T, Sato M, Ninomiya T, Hasegawa T. A study of low NOX combustion in LBG-fuelled 15008C class gas turbine. ASME J Eng Gas Turb Power 1996;118(94-GT-218):534–40. [18] Syred C, Griffiths AJ, Syred N. Turbine combustor with integrated ash removal for fine particulates. ASME Turbo Expo, June 14–17, 2004, Vienna, Austria; 2004. [19] Syred C, Griffiths AJ, Syred N. Combustor development for dual fuels with gas oil and biomass. 42nd Aerospace sciences meeting and exhibit conference, AIAA, 5–9 January 2004, Nevada, USA; 2004. [20] Syred C, Griffiths AJ, Syred N. Cyclone gasifier and cyclone combustor for the use of biomass derived gas in the operation of a small gas turbine. Fuel 2004;83(17–18):2381–92.