Modeling of a Highly Efficient Gasifier Engine System for Small-Scale CHP

Transcription

1 Modeling of a ighly Efficient Gasifier Engine System for Small-Scale CP Erik Karlsson Department of Chemical Engineering, Lund University, P. O. Box 1, SE-1 Lund, Sweden In this study an efficient and promising technique for small-scale combined heat and power production has been identified and investigated by the building of a simulation model in Matlab/Simulink. A Danish -stage gasifier called Viking has been used as the starting point in the study, due to its low tar content and effective way of using heat in the engine s exhaust gases to dry and pyrolyse the wood chips. This preheating technique increases the efficiency of the gasifier from about 75 % to over 9 %. In this study, both an Otto engine and a more advanced CCI (homogeneous charge compression ignition engine, which is a mixture of the Otto and Diesel engine, was simulated. The simulations showed that it would be possible to reach an efficiency of the engine of over %, which gives an electric efficiency of the system near %, based on the lower heating value. If a steam cycle is introduced as well the electric efficiency is approaching 5 %. It should be possible to reach a al efficiency of the system of over 9 %. 1 Introduction In Sweden today, different types of wood fuels are used to a high extent for district heating. In large scale thermal power stations (> MWth, it is common to have combined heat and power production, but in medium and small scale it is rare. This electricity is normally produced by direct combustion and heating of a boiler to generate steam to a steam turbines which powers the electric generator. In small scale this technique has an electric efficiency of just over %, but when the produced heat is included the al efficiency can be over 9 % [1]. Thermal gasification is predicted to be a technique for the future, both for production of electricity and synthetic fuels. There are many types of gasification techniques with different efficiencies, complexities and for different scales and fuels. At the same time the internal combustion (IC engines gets more and more effective and new techniques like homogeneous charge compression ignition (CCI is introduced. If the gasifier and the IC engine can be combined in a clever way, where the excess heat is used to power the gasifier, the whole system can be very effective. IC engines have been powered by wood gas (or producer gas all since World War, but usually with low efficiency. A conventional IC engine optimized for stationary operation has a high efficiency, over % for a heavy IC engine powered with gaseous fuels. If the engine can be optimized for wood gas and the heat losses used to power the gasifier, it should be possible to get a high efficiency of the whole system. The efficiency of a modern alternator is very high, about 95 % for a kva version, so this component would not limit the system. If the overall electric efficiency reaches % the system is competitive with large scale condense power plants which makes it really attractive. The purpose of this study is to build a model of a gasifier engine system powered by wood chips and simulate how the heat from the engine can be used to increase the performance of the system, for example by powering the gasification. Thermal gasification When gasifying biomass, the most common is to use wood in different forms, for example wood chips, but also other more or less dry and fiber rich materials []. When biomass is combusted it goes through four steps; drying, pyrolysis, gasification and combustion but the borders between these steps are not absolute. In the drying process all the water that is not chemically bonded vaporizes []. In the next step the temperature increases, the chemical bonds in the material start to break and gas, tar (or pyrolysis oil and charcoal are produced. The gasification step uses limited oxygen to convert tars and charcoal to gases (CO, CO,, O and C. In the last step enough oxygen is introduced to completely combust the gases leaving only carbon dioxide, water and ash. There are different technical solutions for how a gasifier can be designed []. The gasifier can work directly, by oxidizing part of the fuel or indirectly, driven by an external heat source. The operating parameters, like temperature and pressure, can also vary for different solutions. Circulating fluidized beds are most common in larger scale (>5MW and fixed-beds (also called moving beds in smaller scale

2 applications. The two most important fixed-bed gasifiers are the updraft and downdraft gasifiers. In the downdraft gasifier the airflow enters the reactor in the middle, creating an oxidation zone below the pyrolysis zone []. The gas flow is then moving downwards. In the reduction zone the charcoal is converted to carbon monoxide, manly by reducing carbon dioxide. This type of gasifier gives a clean gas with low content of particles and tars, but has problems converting all of the fuel to producer gas and leaving charcoal in the ash. The temperature of the producer gas out from the reactor is high. This is because the gas stream is not lead through the drying and pyrolysis zones and the heat can not be reused internally, which lowers the efficiency. The downdraft gasifier is also sensitive for bad fuel quality, i.e. high moisture content..1 Global equilibrium model The chemistry of gasification is complicated and the different zones in the gasifier have different reactions going on. To simplify, a so called thermochemical equilibrium model, also called global equilibrium model, can be used [3]. ere the process is reduced to only two reactions (see below, but in the reality it is thousands of different reactions going on, especially under the pyrolysis. In this model all of the reactions are assumed to reach equilibrium and the material in the reactor is completely homogenous and mixed. The model gives a fairly good prediction of the producer gas composition and the oxygen demand, calculated from the fuel properties, reaction temperature and steam supply. The global reaction in the gasifier can be written as: C O + w O + m( O N (1 x 3.76mN a b 1CO + xco + x3 O + xc + x5 + The ultimate analysis gives the mass percent of carbon, hydrogen and oxygen in the biomass and these can easily be converted to mole fractions to get variables a and b in equation 1 [3]. The w indicates how many moles of water that are added per mole of fuel. The water comes from the moisture in the biomass, and steam that is sometimes used in the process. The variable m is the molar flow of oxygen per mole of fuel. The production of the different species in the producer gas are not known so far but are labeled x i in equation 1. The elemental balance for carbon, hydrogen and oxygen (nitrogen already known because it is related to the oxygen are addressed as follows: 1. = x 1 + x + x ( a + w = x 3 + x + x 5 (3 b + w + m = x 1 + x + x 3 ( To be able to calculate the distribution of the species in the gas, two equilibrium reactions are needed. In the reaction zone of the gasifier methane is produced by the following reaction, were K 1 is the equilibrium constant: C + C K( T, P 1 P = ( P C x n x n 5 P P xn = x P 5 (5 If the producer gas can be assumed to be ideal the partial pressure is the same as the molar fraction (x i /n, where n is the al number of moles multiplied with the al pressure of the gas. The second reaction is the so called water-gas shift reaction: CO + O CO + PCO P x x5 K( T = (6 P P x x CO O The equilibrium constant of the methane production is dependent of temperature and pressure, but the water-gas shift reaction is only dependent of the temperature because the al pressure disappears when equation 6 is simplified. To calculate the equilibrium constants, thermodynamic data can be used as done by Zainal et al. []. Unfortunately the calculation of the equilibrium constant for the water-gas shift reaction seems to be incorrect in Zainal et al. This is because the value is more or less zero. Data from Gøbel gives a value of.97 at 8 C, which is reasonable and can be used instead [5]. To solve the equation system, the only missing equation is the energy balance, and the enthalpy of the reactant entering the system must be the same as for the products leaving it. react react prod = = h prod f, biomass = 3.76m + j= w, m 98 Tout j( h T, N 1 3 Tstart 98 T, j + h i 1 i 5 x ( + h f, j 98 Tout h T, i + h f, i (7 The sensible heat (h T and chemically bond energy (heat of formation, h f, gives the al enthalpy for the reactants and the products respectively. The reference temperature in the system is set to 98 K because the heat of formation (h f is given at this temperature. The heat of formation for wood can be calculated from the heating value by subtracting the h f of the products created under complete combustion. The temperature of the producer gas (T out is the same as the reactor temperature. In this model all tars and all of the charcoal and oxygen are assumed to react and there are no heat losses to the surrounding. owever heat losses and external heating can easily be included by adding and subtracting this heat in the energy balance. If for example the reaction temperature and steam input are known, the six variables m and x 1 to x 5 can be calculated from equations, 3,, 5, 6 and 7. The model seems to underestimate the production

3 of methane, and this is because many of reactions that occur in a real gasifier are excluded and that the reactions not always reach equilibrium because they are kinetically limited [5]. To compensate for this, the model can be calibrated by multiplying the equilibrium constants with a factor. 3 Internal combustion engines The two most important reciprocating internal combustion engines are Otto (or gasoline and Diesel engines [6]. The Otto engine uses a pre-mixture of fuel and air which is ignited by a spark plug [6]. If the charge is compressed too much it will get autoignited, which is called knocking and can damage the engine. ow easy a fuel auto-ignites is measured as the octane number. Otto engines fueled with natural gas is a well known technique and is among others used for combined heat and power production in the range of some hundred kw to a couple of MW. A Diesel engine is normally not using premixtures of air and fuel [6]. Instead, only air is filled in the cylinder which is then compressed to a high temperature and pressure. At the power stroke the liquid fuel is sprayed into the cylinder at very high pressure and is auto-ignited. This so called compression ignition is in some way the opposite of spark ignition. A new promising engine technique called homogenous charge compression ignition (CCI has been introduced [6]. This is a combination of the Otto and Diesel engine; hence it uses a pre-mixture of fuel and air and compression ignition. This gives a high efficiency and very low emissions of NO x, but is at the same time hard to control. To control the combustion duration and ignition timing, variable compression can be used and the fuel can be to be diluted with inert gases, for example by using exhaust gas recirculation (EGR. When using producer gas the gas is already diluted, making it suitable for a CCI engine. 3.1 Engine performance with producer gas When using gaseous fuels with low calorific value (LCV, i.e. producer gas, in a combustion engine there is always a problem with power derating []. When the fuel has a larger volume per energy unit, the energy input to the engine will decrease and since the heat and mechanical losses do not decrease to the same extent, the efficiency will decrease [7]. Due to the low energy density, the stoichiometric air/fuel ratio also decreases; from about to. This makes the different in power derating between concentrated and diluted fuels less significant than it first appears, because the energy density of the charges differs less. To overcome power de-rating supercharging (e.g. turbo has to be used. An increase in the compression ratio will increase the efficiency and therefore lower the effect of power de-rating. A study investigates the knocking sensitivity of a sparkignition engine fuelled with producer gas and found that a compression ratio of 17:1 shows no tendency of auto-ignition [8]. This means that it is possible to have a higher compression ratio with producer gas than for example with gasoline, which normally has a CR of about 1:1. The producer gas has also lower flame speed, which gives a longer combustion duration than for gasoline, and therefore a low engine speed is preferred [9]. The gasifier engine system The BTG Biomass Technology Group in the Netherlands used a downdraft gasifier and had a relatively high engine efficiency of 37 % (gas to electricity at their tests on a 15 kw e plant [7]. At the same time the gasifier efficiency was only 71 %, which gave an overall electric efficiency of 7 % gross. In their study the temperature of the producer gas was kept over the dew point to avoid condensation. The Biomass Gasification Group in Denmark has designed a two-stage gasifier called Viking [1]. In this design, heat from the engine s exhaust gases was transferred to the incoming wood chips in a pyrolysis reactor, where the temperature reached about 6 C. The feed of pyrolysis products, that is gas, vaporized tars and charcoal, was then entering a downdraft gasifier. In the upper part of the gasifier the tars were partially oxidized with air, which made the gas temperature reached about 1 C. In the lower part of the gasifier the charcoal was gasified, and here the temperature was about 7-8 C. If the scale up is achievable, the Viking gasifier seems to be a successful solution with both good gas quality and high energy efficiency. The fuel to gas efficiency of the 7kW in pilot plant was 93. %, the gas to electricity efficiency was 9.1 % and the overall fuel to net electricity was 5.1 % [11]. The efficiency of their engine was not as impressive as the gasifier and this was partly because no supercharging was used and that the engine only operated at part load. Unfortunately no other operating parameters are presented in the paper so it is not possible to make a deeper analysis. 5 System description From the literature review of gasifier engine systems above, it was found that the Danish -stage gasifier Viking was a good solution, and therefore this was the base for this study. The suggested system design in this study was for that reason more or less the same, except that the exhaust gases from the engine were not heated by the producer gas (figure 1. This configuration was chosen because it was interesting to investigate if the engine could give a higher temperature in the exhaust gases than the temperature of the producer gas. The supercharging

4 Cool water Cooler/condenser Wood chips Exhaust District heating Gasifier Preheater/pyrolysis reactor Air heater Gas cleaning Cooler/ condenser Engine and altenator Compressor Electricity Air Mixing valve Cool water Figure 1 The proposed system design, the exhaust gases is heating the wood chips without physical contact compressor is powered by the crank shaft because a turbo decreases the exhaust gas temperature. A Matlab/Simulink model of the system in figure 1 was built to evaluate its electrical and al efficiency. Using this model, it was investigated how variations in exhaust gas temperature/external heating influenced the producer gas. Both Otto and CCI engines were simulated and compared. The model was validated with results from the Viking gasifier pilot plant [11]. 5.1 Assumptions in the model The temperature in the gasifier was 8 ºC, and production of tars and ash was not considered, as well as charcoal losses. Parasitic losses were not calculated except for the charging compressor, and heat losses to the surrounding were only calculated for the gasifier. The alternator efficiency was assumed to be 95 %. Emissions from the engine s exhaust gases ware not considered at all. 5. Model description The main model was built in Simulink and has six sub-models, the gasifier, the internal combustion engine, the biomass preheater, the air preheater, the gas cooler and the exhaust gas cooler model. These sub-models were coded in Matlab. The cooling water from the two coolers and the engine are all used for district heating. The thermal efficiency of the plant is calculated by cooling this hot water to the return temperature, in this case from 9 C to 35 C. The model is quasi stationary, which means that reactions are at equilibrium and stationarity is reached for every single time step. If input or operating parameters are changing between the time steps, the system will react but there is no time lag or accumulation in the system. When starting the simulation the model needs a short time to adjust and this is because of the many feedback loops that use initial values set by the user. The more these guessed values differ from the right values, the longer time the adjustment will take. So in one sense there are time lags in the feedback loops of the model, but these are more a result of the calculation method, than attempts to model physical phenomena found in the real system. The gasifier Matlab function calculates the gas production from the input parameters using the global equilibrium model described in chapter.1. The molar flow of the different gas species, as well as the air required to reach the temperature are calculated. This is done by solving the equation system in chapter.1 using fsolve which is a numerical equation solver in Matlab Optimization Toolbox. 5.3 The pyrolysis reactor The biomass preheater, or drying and pyrolysis reactor, was modeled as a heat exchanger. Data for the heating of biomass was taken from Benny Gøbel, where 5 kw of heat was needed to pyrolyse kg dry wood chips/h, corresponding to 1 kw fuel based on lower heating value [1]. The temperature reached 6 ºC in their test, and therefore it was assumed that 86. W of external heat was required per MW fuel and K (5 / (1*(6ºC-ºC = 8.6*1-5. When heating dry biomass, the specific heat (CP can not be considered constant when the temperature reaches some hundred degrees Celsius, and this is because the properties of the biomass change under the pyrolysis. This fact is not considered in the model and therefore the temperature of the outgoing biomass stream should not differ too much from 6

5 ºC. The moisture in the wood chips is also vaporized and heated in the preheater, but here a variable C P was used for the heating of steam. 5. The engine model In the engine model there is a sub-function calculating the mechanical and thermodynamic cycles of one cylinder in the engine, giving how the energy is distributed between work, heat losses, exhaust gas losses and mechanical losses. This function was developed by Rolf Egnell with some minor modifications by the author [13]. A very important parameter in the cylinder function is the ratio of specific heat (C P /C V, labeled γ or κ [13]. For an isentropic change PV γ is constant, but in a real combustion engine γ is changing during the combustion because the temperature and gas composition is changing. In the model it is assumed to be constant, and therefore it is important to find an accurate mean value. This value has been estimated to 1.3 by Egnell, and is therefore a weak point in the model, but at the same time the results from the simulations seem reasonable. 6 Results and discussion 6.1 Validation To find out if the model of the gasifier and the biomass preheater was correct, it was validated with data from the Viking gasifier pilot plant [11]. Wood chips with the same elemental composition and moisture content as the Viking gasifier was used in the simulations. Two cases was simulated; a best case, which refers to a system with maximum efficiency under the given assumptions, and a standard case, referring to a system as near the Viking gasifier as possible. As mentioned in chapter.1, the global equilibrium model tends to underestimate the production of methane; and therefore a correction factor of.5 was used for the methane reaction in all simulations to calibrate the model. The water-gas shift reaction can also have a modified equilibrium constant, and in the standard case it is multiplied with to give a better match with the Viking gasifier. This gave a very close gas composition, though the hydrogen gas content is slightly higher and the nitrogen gas slightly lower than the Viking gasifier. The original engine model was validated when it was developed by Egnell [13]. In this study the engine efficiency was calculated to over %, and this was mainly because both high compression ratio and supercharging were used. Unfortunately, no other studies have been found with this engine configuration. owever, engine tests with producer gas at high compression ratios without supercharging were carried out by Sridhar et al., where an increase of CR from 11.5:1 to 17:1 increased the gas to work efficiency from 5.7 % to 3.5 % [8]. Stassen et al reached a gas to electricity efficiency of 37 % when using supercharging at their gasifier engine system [7]. The engine was operating on low speed (75 rpm, which was chosen because of the low flame propagation speed of the producer gas. The efficiency of their alternator was not mentioned. From these tests, it was considered reasonable that efficiency over % (gas to work can be reached when supercharging and high compression ratio are combined. 6. External heating of the gasifier It can be seen in figures a and b that there is a large potential for increase of the gasifier efficiency if excess heat can be used to heat the gasifier. In this study the biomass is externally heated to near 6 ºC, and for this an exhaust gas temperature of 79 ºC is needed. This gives an extra addition of heat to the gasifier, corresponding to about 15 % of the energy in the wood chips based on the lower heating value, and the lower heating value of the gas is increased from 5 to 6.5 MJ/Nm3 dry gas in the standard case. This increase can be explained by the increase of hydrogen gas and carbon monoxide content in the producer gas. Gas composition (vol % wet gas eating value (J/Nm 3 dry gas 5 (a External heating (Q extern /Q biomass 1 x 16 (b N CO CO O C Preheating to 6 ºC Preheating to 6 ºC External heating (Q extern /Q biomass Figure Wet gas composition and lower heating value of producer gas as a function of external heating of the gasifier If external heat corresponding to % of the energy in the fuel can be add, no oxygen is needed and all of the energy in the wood is converted to chemically bond energy in the producer gas. This can be seen in figure a, as the nitrogen gas concentration decrease to zero. Gasification without oxygen supply is called indirect gasification or pyrolysis. Unfortunately, it is very hard to get a good gas quality in this

6 type of processes and parts of the fuel will only be converted to tars and charcoal. 6.3 System performence During the simulations, it was found that the efficiency of the engine and the whole system increased if supercharging with intercoolers was used. With the Otto engine the optimal charging pressure was.3 bars absolute, while about 1.7 was optimal for the CCI engine (table 1. The higher combustion speed in a CCI engine gives higher efficiency but also higher maximum pressure, causing a lower charging pressure to be used in order not to damage the engine. The charging pressure was consequently limited by the maximum allowed cylinder pressure ( bar, but also by the fact that a larger compression work lowers the performance of the system. Table 1 Simulation of system performance Otto Otto CCI CCI Gasifier simulation case Standard Best Standard Best Energy loss in gasifier % % % % Charging pressure (bar abs Compression ratio Exhaust gas temperature Wood to gas efficiency Gas to electricity efficiency Wood to electricity efficiency 71 C 7 C 663 C 691 C 89.8% 9.% 89.3% 93.6%.3%.% 3.8% 3.8% 38.% 39.9% 39.% 1.% The results from the simulations showed that the electric efficiency of the system was near % when an Otto engine was used and slightly higher when a CCI engine was used (table 1. Without preheating of the wood chips, the electric efficiency of the system decreased to 31 % in the standard Otto case, mainly due to a decrease of the gasification efficiency from about 9 % to 75 %. To have a high amount of preheating, the engine s exhaust gas temperature needs to be high. In chapter. it was mentioned that a temperature of 79 C was required to preheat the wood chips to 6 C, but the simulation gives a slightly lower value (table 1. owever, these values are strongly affected by the assumed heat transfer efficiency of the preheater, and how well the exhaust gas temperature is preserved. 7 Conclusions There exist today a promising gasification technology that would be very effective if it can be scale up External heating of the wood chips by the engine s exhaust gases increases the gasification efficiency from 75 % to over 9 % It should be possible to develop and optimize internal combustion engines for operating efficiently on producer gas, and especially the CCI engine technique seems to fit well in combination with a gasifier because it prefers fuels with low energy density If the exhaust pipe and valves in the engine could be well insulated it would be possible to get an exhaust gas temperature high enough to effectively power the pyrolysis reactor 8 References 1. Aabakken, J. Power Technologies Energy Data Book. 6 [cited; Available from: /db_chapter_bio.pdf.. Reed, T.B. and A. Das, andbook of Biomass Downdraft Gasifier Engine Systems. 1998: The Biomass Energy Foundation Press. 3. Melgar, A.s., et al., Thermochemical equilibrium modeling of a gasifying process. Energy Conversion and Management, 7. 8: p Zainal, Z.A., et al., Prediction of performance of a downdraft gasifier using equilibrium modeling for different biomass materials. Energy Conversion and Management, 1. : p Jarungthammachote, S. and A. Dutta, Thermodynamic equilibrium model and second law analysis of a downdraft waste gasifier. Energy, 7. 3: p Johansson, B., Förbränningsmotorer. 6: Division of Combustion Engines, Department of Energy Sciences, Lund University. 7. Stassen,.E. and. Knoef. Theoretical and practical aspects on the use of LCV-gas from biomass gasifiers in internal combustion engines. [cited; Available from: 8. Sridhar, G., P.J. Paul, and.s. Mukunda, Biomass derived producer gas as a reciprocating engine fuel an experimental analysis. Biomass and Bioenergy 1. 1: p Porpatham, E., A. Ramesh, and B. Nagalingam, Investigation on the effect of concentration of methane in biogas when used as a fuel for a spark ignition engine. Fuel, 8. 87: p enriksen, U., et al., The design, construction and operation of a 75 kw two-stage gasifier. Energy, 6. 31: p Ahrenfeldt, J., et al., Validation of a Continues Combined eat and Power (CP Operation of a Two- Stage Biomass Gasifier. Energy & Fuels, 6. : p Gøbel, B., Dynamisk modellering af forgasning i fixed koksbed, in Institut for Energiteknik. 1999, Danmarks Tekniske Universitet. 13. Egnell, R., On Zero-dimensional Modeling of Combustion and NOx Formation in Diesel Engines, in Department of eat and Power Engineering. 1, Lund Institute of Technology.