Co-firing of natural gas and Biomass gas in biomass integrated gasification/combined cycle systems

Energy 28 (2003) 1115–1131www.elsevier.com/locate/energy

Co-firing of natural gas and Biomass gas in biomassintegrated gasification/combined cycle systems

Monica Rodriguesa,b, Arnaldo Waltera, AndreFaaijb,∗

a Mechanical Engineering College, State University of Campinas, P.O. Box 6122, 13081-970, Campinas, Brazilb Department of Science, Technology and Society—Utrecht University, Padualaan 14, 3584 CH Utrecht,

The Netherlands

Received 22 August 2001

Abstract

This work aims to evaluate the co-firing of gas derived from biomass and natural gas in combined cycles.It is suggested that co-firing can solve some of the initial technological problems associated with the gasturbines of BIG-GT (Biomass Integrated Gasification/Gas Turbine) plants. De-rating is the simplest strategythat allows continuous gas turbine operation with low calorific value fuels, but it also reduces cycle powerand efficiency. The proposed biomass gas is derived from sugar-cane residues and has a lower heatingvalue of around to 6 MJ/Nm3. Modeling results show that if the natural gas content is higher than 35–50%(energy basis) no de-rating will be necessary. At these proportions, the efficiency of electricity generation isnot substantially reduced vis-a`-vis the reference case. Another important outcome of the modeling is thatthe peak in power that occurs for natural gas contents slightly higher than 50% in energy basis. Anotheradvantage of co-firing is that it will entail only small hardware modifications to the gas turbines. A compari-son is made between the co-firing strategy and the improvements likely to arise from retrofitting gas turbinesfor LCV fuels. 2003 Elsevier Science Ltd. All rights reserved.

1. Introduction

The BIG-GT (Biomass Integrated Gasification/Gas Turbine) technology has the potential toenhance the use of biomass use since it can substantially improve the efficiency of electricityproduction. Whereas conventional biomass power plants based on steam cycles have efficienciesin the 15–30% range[1], BIG-CC cycles (the same principle of biomass gasification but integrated

∗ Corresponding author. Tel.:+31-203537643; fax.+31-302537601.E-mail address: [email protected] (A. Faaij).

0360-5442/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved.doi:10.1016/S0360-5442(03)00087-2

1116 M. Rodrigues et al. / Energy 28 (2003) 1115–1131

into a combined cycle) are expected to achieve efficiencies of about 40% for wood and atmos-pheric gasification at modest scales [2,3]. Demonstration projects committed with engineeringdevelopment and aimed at cost reduction are in progress. The first unit using this technology—a CHP plant (6 MWe and 9 MWth) based on pressurized gasification of wood residues and chips—operated in Varmano (Sweden) from 1996 to 1999. A combined cycle unit of 8 MWe net capacityis under construction in Yorkshire, UK (ARBRE project), it is based on atmospheric gasificationof short rotation coppice and forestry waste. Other demonstration units are under consideration,the largest of which (for 32 Mwe) is likely to be built in Bahia, Brazil.

Cycle efficiency is a crucial parameter for the cost-effectiveness of BIG-CC systems. One majortechnological problem with the first generation of BIG-CC units is the reduced efficiency resultingfrom the use of low calorific value (LCV) fuels in gas turbines that are originally designed fornatural gas. Most of the losses stem from control strategies applied to the gas turbine to keep thecompressor operation safely beneath the surge line, which is its limit of stable operation. Com-pressor surging is associated with a sudden drop in delivery pressure and with violent aerodynamicpulsation, which is transmitted throughout the machine [4]. The rise in surge risk is associatedwith an increase in compressor pressure ratio caused by a larger mass flow passing through thegas turbine when a low heating content gas is used.

The strategy that will probably be used in the short-term to control surge limits is the simpletechnique of rating due the gas turbine [3,5]. De-rating consists of imposing a lower gas turbineburning temperature and thus, lowering the compressor pressure ratio. Co-firing the biomass-derived gas with natural gas will increase the heating content of the resulting fuel reducing theimpact of de-rating or eliminating it. Furthermore, mixing biomass gas with natural gas wouldprevent the need for significant burner modifications. Co-firing will also increase combustion stab-ility.

The co-firing proposal analyzed in this paper is backed by the increasing availability of naturalgas and sugar cane residues in the State of Sao Paulo in the Southeast of Brazil. Sugar-cane trash(leaves and tops of sugar-cane plant) is the biomass considered in this paper. Trash availabilityis likely to increase rapidly in the coming years due to the reduction of pre-burned harvesting inthat region. The end of the practice of burning sugar cane leaves and tops might lead to an annualbiomass availability of between 100–200 PJ in Sao Paulo state alone. Natural gas is supplied bymeans of a newly built gas pipeline with a capacity to supply 30 million Nm3 per day that crossesthe main sugar-cane areas. The natural gas is imported from Bolivia and the investment maderequires high consumption for pay-off.

2. Co-firing experience and previous work

The term co-firing is used to refer to the combined use of biomass and fossil fuel, in powerplants as well as in industrial steam boilers. The most acceptable idea is to burn a mix of biomassand coal in power plants that are adapted for this purpose whenever the useful life of the existingboilers expires. Owing to the substantial reduction in technical and economical risk, co-firing hasbeen considered in some countries to be the first step in the enhancement of biomass use in powergeneration [6]. In the USA, for instance, some coal burning electric utilities are becoming inter-ested in biomass co-firing as a low-cost option for reducing greenhouse gas emissions. Co-firing

1117M. Rodrigues et al. / Energy 28 (2003) 1115–1131

is a cost-effective way of utilizing biomass because it takes advantage of the relatively highefficiency of large coal boilers but does not require a large investment [7].

So far, no co-firing experience of biomass and natural gas on a gas turbine has been reported.Co-combustion of LCV fuels and natural gas has been commercially implemented in Italy inpower plants for the steel industry. Stambler [8] and Thoraval [9] report that the major steel-maker in Italy—Ilva—has replaced old steam turbo-generators by a new combined cycle instal-lation nominally rated at 530 MWe. The power plant has three similar units based on the GEMS9001/E gas turbine. The system has improved overall plant efficiency from 36 to 45%. Thecombined cycle runs on a mixture of three different sources of waste gases from the steel mill:blast furnace gases with a low heating value (LHV) of 3.35 MJ/Nm3, coke oven gas (LHV =18.84 MJ/Nm3) and LD gases (LHV = 8.7 Nm3). The blast furnace gases provide the largestvolume and the coke oven gases the lowest. The mixture with natural gas increases the LHV tovalues between 6.3 and 8.4 MJ/Nm3. The lower heating value in that range ensures that theminimal thermal requirement for the combustion chamber of the gas turbine and for flame stabilityis achieved. By March 1999 the plant had operated for more than 60,000 hours. Another similarinstallation was planned for the process of repowering in Livorno, Italy (189 MWe). It wasexpected to become operational by the third quarter of 2000.

A report on co-firing has been produced in The Netherlands [10]. The work focused on thepossibility and potential of co-firing LCV fuels and natural gas for different power configurations.The gas turbine suppliers contacted for this preview, i.e., General Electric, Asea Brown Boveri,Allison and Rolls Royce, have been informed about the gas turbine constraints and requiredadaptations. A similar study was developed some years ago at the National Renewable EnergyLaboratory—NREL, USA, but only a preliminary analysis of technical options was conducted atthat time [11].

The objective of this study is to present technical issues, modeling assumptions and performanceresults on the co-firing of biomass gas and natural gas in combined cycles. The next item (item3) discusses the hardware adaptation required for the use of biomass gas in gas turbines. Item 3also includes a discussion on constraints and modifications related to the use of low calorificgases (LCV) and how the modifications can be reduced with co-firing. Furthermore, this itemaddresses issues like flame temperature, NOx formation and limitations associated with the hydro-gen content of biomass gas. The BIG-CC system is described on item 4, whereas the modelingassumptions are described on item 5. The model, developed in excel spreadsheet, predicts theperformance of the gas turbine and steam cycle in design when 100% natural gas is used andoff-design for mixtures of biomass gas. Three different cycles are considered, namely aero-deriva-tive, medium and large size combined cycles. These cycles were chosen as representative ofcombined cycles of certain scale ranges. The results are presented in item 6 and consist of graphsthat show the effects on cycle efficiency and power when the biomass gas content in the mixtureis varied from 0 to 100%. A comparison with the hypothetical case of a turbine that has beenretrofitted for the use of biomass is presented in item 7.


3. The use of biomass fuel gas in gas turbine

3.1. Constraints and modifications imposed by the use of LCV fuels

Besides reducing the need for control strategies for gas turbines, a mixture with natural gascould minimize the hardware modifications required for the use of LCV fuels in machines thathave been designed for natural gas. The extent of the hardware modifications depends on theheating value of biomass gas and on the type of gas turbine.

Combustion stability could be a problem for the use of LCV gases in gas turbines, particularlyduring the start-up, when ignition can also be difficult due to lower temperatures [8]. The combus-tion stability for biomass-derived gas has been verified for the modified LM2500 combustor [12].

When a LCV fuel gas is used, pressure loss is increased due to the injection of a larger volumethrough a nozzle originally designed for a fuel with much higher energy content. It might provenecessary to replace the gas turbine nozzle and, eventually, the whole combustion chamber.According to Consonni and Larson [3] can-type combustors used in many industrial gas turbinesgenerally provide adequate cross section and volume for complete and stable combustion withacceptable pressure drops. In fact, many industrial gas turbines have operated successfully foryears using LCV fuels from, for instance, steel mills (e.g., blast-furnace gases). However, accord-ing to De Kant [10], it is essential to replace the combustion chamber, originally designed toburn natural gas, in order to adapt industrial GE gas turbines. De Kant estimates that this modifi-cation will increase the gas turbine cost by about 20%. Neilson [12] describes modificationsintroduced into the aero-derivative GE-LM2500 so that it can operate with biomass-derived gas.The modifications include a new fuel nozzle for the combustion chamber with a new fuel nozzle,a new swirler venturi and a larger cowl opening. The fuel pipeline and manifolds also have tobe adapted.

Another concern is the hydrogen content of the biomass-derived gases (between 10–20% vol.),which is much higher than for natural gas. Whereas this content aids flame stability [3], it maylead to back stream flame propagation in a dry low NOx combustion chamber [10].

3.2. Reducing gas turbine modifications with co-firing

Table 1 presents the estimated composition of the gas the derived from sugar-cane trash andof the natural gas considered in this paper. The biomass-derived gas composition was taken fromWalter et al. [11,13]. The fuel mix LHV presented in Fig. 1 was calculated on the basis ofthese compositions.

The fuel mixture containing natural gas increases the energy density and, hence, reduces theneed for hardware modifications. According to De Kant [10], if the fuel mixture contains morethan 75% (energy basis), or about 18 MJ/Nm3 (see Fig. 1), there will be no need to replace thecombustion chamber of industrial gas turbines manufactured by General Electric. The same reportstates that ABB has a special combustion system for gases with a heating value of between 8and 16 MJ/Nm3, such as gases derived from coal gasification. This LHV range is between 30and 70% of natural gas (energy basis) in the fuel mixture with biomass-derived gas, as can beseen in Fig. 1.

Further tests are necessary before final recommendations can be given regarding the suitability


Table 1Composition of the biomass-derived gas and natural gas (% volume)

Component Clean syngas Gas natural

H2 16.69 1.00CO 19.98 –CO2 10.49 0.52CH4 2.63 80.61C2H2 – 6.00C2H4 – 6.00C2H6 – 4.00C6H6 0.33 –H2O 3.24 0.07N2 46.64 1.80

Fig. 1. Estimated LHV as a function of the proportion of natural gas in the fuel mix—r(NG) (energy basis).

of biomass-derived gases for industrial fuel injection systems. However, primary indications arethat co-firing has the potential to reduce the need for modifications in the combustion system or,at least, to increase the heating content of the fuels mix so that it reaches heating values equivalentto those of gases derived from coal gasification, the use of medium LHV gases has proved com-mercially viable.

3.3. Flame temperature and NOx formation

The flame temperature of the mixture biomass-derived gas and natural gas will be higher thanthe fuel flame temperature for biomass derived-gas only. In gas turbines thermal NOx is a majorconcern for natural gas combustion and its formation increases exponentially with the flame tem-perature [4]. Whereas thermal NOx formation is likely to be very low for the biomass deriveddue to its lower flame temperature and NOx emissions should not be a concern [3,14].

Variations in flame temperature as a function of the fuel mix composition are here represented


by the adiabatic flame temperature (AFT) in stoichiometric conditions. Fig. 2 presents AFT vari-ation as function of the natural gas content in the fuel mixture. As it can be seen, the AFT forbiomass-derived gas is about 20% lower than for natural gas (1700 °C for Biomass gas and 2100°C for natural gas).

The actual flame temperature will also depend on the degree of gas dissociation and on combus-tion excess air. Dissociation was not considered here. Theoretically, flame temperature is at itsmaximum in stoichiometric conditions and will fall off with both rich and lean mixtures. Resultsshow that the excess air with simulated gas turbines operating at full load with natural gas variesbetween 270 and 300%. It is estimated that for de-rated gas turbines burning biomass-derivedgases, the combustion occurs with a 250–270% excess air. Due to the composition of biomassgas and the degree of de-rating that is imposed on the gas turbine, excess air does not vary linearlywith the composition of the fuel gas mix. The excess air for a fuel mix with 50% of natural gas(energy basis) is between 210–215%. This is considerably less than that for the case with 100%natural gas (270–300%). The conclusion drawn from those values is that for the mixture biomassgas and natural gas, NOx emissions may be a concern and the problem should be further addressed.

Numbers presented previously are preliminary and contain inaccuracies associated with a gen-eric compressor map and estimated cooling air so that they can only present trends. Moreover,the mixture will not be homogeneous so that detailed chemical and fluid analysis is necessaryand out of the scope of present work. Further detailed studies on combustion and tests are neededwith regard to NOx emissions for co-firing, particularly due to limitations on the use of gaseswith some hydrogen, such as biomass gas, in dry low NOx chambers. As an illustration, Fig. 3presents the estimated stoichiometric air/fuel ratio as a function of the fuel mixture.

3.4. Hydrogen content

The hydrogen content of biomass-derived gas can be an advantage in that it improves the flamestability of the fuel-mix combustion. On the other hand, it can be a drawback in the sense thatlimits the use of biomass-derived gas in dry low NOx combustion chambers. Results achieved byGeneral Electric suggest that the higher the hydrogen content, the lower the heating value required

Fig. 2. Adiabatic Flame Temperature as a function of the proportion of natural gas share in the fuel mixture—r(NG)

(energy basis).


Fig. 3. Stoichiometric air/fuel ratio as a function of the proportion of natural gas in the fuel mixture (energy basis).

for stable combustion [3]. This is explained by the fact that hydrogen has a higher flame propa-gation speed than CO or CH4, which are the other principal combustible gases in biomass-derived gases.

The hydrogen content is also associated with the possibility of back stream flame propagationin a dry–low NOx combustion system that premixes the fuel and air prior to the combustion.Whereas General Electric recommend no hydrogen in the fuel for some of its dry–low NOx

combustion systems, ABB sets a maximum of 5% for the hydrogen content in its dry–low NOx

systems [10]. Co-firing could help to limit the hydrogen content of the mixture, as natural gascontains low levels of hydrogen. However, as can be seen in Fig. 4 the 5% maximum volumecontent suggested by ABB will only be achieved for fuel mixtures with more than 90% of naturalgas (energy basis).

Further research and testing are necessary in order to establish more reliable limits—and eventu-ally, to reduce restrictions. In addition, manufacturers might consider redesigning their dry-lowNOx systems in order to suit the biomass-derived gas. However, that is unlikely to occur in

Fig. 4. Hydrogen content as a function of the proportion of natural gas in the fuel mixture (energy basis).


the short term and so other options, such as steam injection for NOx emissions control, shouldbe considered.

4. BIG-CC system description

The co-firing option analyzed in this paper is based on a combined cycle fueled by a mixtureof natural gas and gas derived from biomass (sugar-cane trash) gasification. The capacity of thebiomass gasifier and the size of the gas-cleaning module depend on the proportion of biomassgas in the gas mixture. To evaluate various capacities of power production, three combined cyclesbased on aero-derivative and on industrial gas turbines were considered. Biomass gasification isassumed to take place in an atmospheric air-blown gasifier similar to the gasifiers developed bythe Swedish company TPS (Termiska Processor AB). This type of gasifier is technically provenand is likely to achieve a commercial status in the short-term [2,15]. Gasification occurs with airinjection into a circulating fluidized bed (CFB) at 2 bar. After gasification, the raw gas passesthrough another CFB reactor in which the tars are cracked. Dolomite is used to catalyze thecracking of tars to gases and lower molecular weight vapors [15].

After the tar cracker, the raw gas is cooled and the heat is recovered sequentially to increasesteam generation, to preheat the blast air and the biomass fuel gas itself (before it enters the gasturbine) and to preheat the feed water for the heat recovery steam generator (HRSG). After coo-ling, a bag house filter removes the particulates and the remaining components are removed in awet scrubber. The biomass gas is then compressed and preheated and injected in a gas turbine at370 °C. The energy of the exhaust gases is recovered via an unfired HRSG. Wet, incoming thebiomass is dried by using flue gas from the HRSG (down to 15% moisture content). It is predictedthat drying design conditions requires 200 °C for the flue gas from the HRSG [3]. Fig. 5 representsa scheme drawing of the system.

Some characteristics of commercial gas turbines such as efficiency, exhaust temperature, press-ure ratio, heat rate and taken from information on literature [20] (GE LM2500, GE PG6101 (FA)and GE PG7001) have been used for cycle modeling. The LM2500 model considered in thispaper is the aero-derivative machine modified by General Electric for use with biomass-derivedgas [12]. The other two machines are heavy industrial gas turbines. Parameters that are not avail-able in the literature have been estimated on the basis of reference performance data for suchmachines. Combined cycles based on such machines are identified as aero-derivative (LM2500),medium-size (PG6101) and large (PG7001). These three gas turbine models were chosen in orderto cover a broad range of power capacities.

5. Modeling assumptions

In BIG-CC or co-fired cycles, a larger mass flow goes through the expander because of thelower heating value of the biomass-derived gas. The only way that this higher flow can be accom-modated without any design change or gas turbine de-rating is to increase on the compressorpressure ratio. The main problem associated with an increase in pressure ratio is that the com-pressor operation could approach its surge limit [4]. Off-design operation with increased com-


Fig. 5. Scheme of the co-fired combined cycle considered.

pressor pressure ratio is aggressive to the gas turbine due to increased thermal and mechanicalloads [17].

Off-design operation strategies or a redesign need to be applied the gas turbine so that it oper-ates well below compressor surge limits. Off-design strategies can be compressor blast-off gasturbine or de-rating. Another possibility is to redesign the inlet expander nozzle to suit the gasturbine for low calorific gas. The impact of these control strategies on the performance of BIG-CC plants was analyzed earlier [18]. De-rating is the simplest approach: it allows the use ofbiomass-derived gases in gas turbines designed for natural gas. However, it is also the worstoption from the point of view of cycle performance. The compressor pressure ratio can be reducedto acceptable levels by means of reducing its maximum temperature. Less fuel is injected andthe compressor pressure ratio can be reduced to acceptable values.

Performance evaluation of co-fired combined cycles requires an off-design procedure for gasturbine simulation. The main difficulty in predicting off-design performance of gas turbines stemsfrom the fact that no actual compressor maps are available in the open literature. Consequently,it is not easy to predict accurately variations in inlet airflow as a function of pressure ratio vari-ations, nor is it easy to define surge margin. Simulation results presented in this paper are basedon a computer model previously described by Walter et al. [5,13] and by Souza et al. [18].Simplifications in the modeling led to simulation results that cannot be associated with specificcommercial gas turbines, but results show trends of certain classes of machines.

According to the simulation procedure described in [5] and [18] a generic compressor map is


assumed for the calculations. The compressor map is necessary in an off-design procedure thatpredicts variations in pressure with flow and temperature variations. The surge limit is estimatedassuming that the pressure ratio when the gas turbine operates at ISO conditions burning naturalgas is 0.75 from the pressure ratio on the surge line. A parameter named compressor map variable(CMV) was defined to monitor the margin from the pressure ratio on the surge line. A CMV isdefined as (PR - 1) / (PRs - 1), where PR is the pressure ratio at a given point and PRs is thepressure ratio that corresponds to the surge at a given running line. Thus, CMV equal to 0.75means that the compressor pressure ratio is relatively far from the surge line. It was also assumedthat the limit for continuous operation of industrial gas turbines corresponds to CMV = 0.95. Infact, the compressor in industrial gas turbines can cope with a large increase in the pressureratio [10].

On the other hand, aero-derivative engines have a more critical compressor operation and con-trol. The machines are double shafted and can withstand changes on the rotational speed of thegas generator. Due to the wider range of compressor operation stricter control is needed. Themaximum pressure ratio for the aero-derivative gas turbine was assumed to be the same as thatproposed by Consonni and Larson [19], namely corresponding to a CMV of 0.85.

An additional modeling simplification is that for all three simulated combined cycles steam israised at just one pressure level in an unfired HRSG. The reason for this simplification is thatfor BIG-CC cycles the gains with multiple pressure configurations and reheating are only marginaldue to the minimum exhaust gas temperature required for the dryer [19]. In this sense, the pro-posed configurations are not optimized for natural gas-fired combined cycles and simulation resultsof net plant efficiency are lower than performance figures listed in the literature [20]. A combinedcycle based on one LM2500 operates with two pressure levels, whereas heat recovery steamgenerators of combined cycles based on Frame 6 and Frame 7 gas turbines comprise three pressurelevels and reheat.

In these conditions, a commercial cycle based on one Frame 7 can reach a net efficiency ofabout 56% (ISO basis) [20] whereas the simulation result is 49%. This difference is mainly dueto the simplification of just one steam pressure level. However, model simplifications do notinvalidate the main conclusions of this work. Steam pressure was chosen as 8.8 MPa for thecombined cycle based on aero-derivative gas turbine and as 10 MPa for the other two cycles.The isentropic efficiency of the steam turbine is set constant and equal to 75% in all simulatedcases. Part of the steam flow is extracted at 1.8 MPa for the deaerator. The remaining steam isexpanded in the low-pressure stage and condensed at 9.6 kPa.

Gas turbine de-rating leads to a reduction in steam temperature. In the models the superheatedsteam temperature is lowered to keep the minimum HRSG approach of 30 °C. Steam temperatureis increased with co-firing up to the design point. In fact, maximum steam temperatures arereached for proportions of natural gas as low as 20% (energy basis). Depending on the fuelmixture, steam production is constrained keeping order the required minimum HRSG exhausttemperature (200 °C) that will ensure dryer operation. Varying the share of syngas in the fuelmix also interferes in the operation of the HRSG feed-water heater. When only biomass-derivedgas is used, feed-water is heated to 120 °C. This temperature is reduced when the proportion ofbiomass-derived gas in the fuel mix is decreased. The main assumptions used in modeling thecombined cycles are listed in.Table 2.


Table 2Main assumptions used for calculations

HRSG and steam cycle Approach �T = 30 °C; pinch point �T = 15 °C.Heat losses 0.7% of heat released by gas; pressure drop at the gas side 3 kPa, pressuredrop at the super heater 10%.Overall isentropic efficiency of steam turbine 0.75.Steam pressure at the condenser 9.6 kPa.Water outflow from deaerator: 488 kPa, 120 °C.Total auxiliary power=160% of the estimated power for the pumps (isentropic efficiency0.65).Cycle based on aero-derivative GTs: Steam raise at 8 MPa, 480 °C (maximum T).Medium and large cycles: Steam raised at 10 MPa, 538 °C (maximum T).

Gasifier Outlet syngas temperature 870 °C, outlet pressure 0.20 MPa, �p = 0.02 MPaDryer Biomass dried from 50 to 15% mc, exiting at 70 °CHeat exchangers �p /p 2%; heat losses equivalent to 2% of heat transferredSyngas compressor Organic and electric efficiencies 90%.Air compressor Polytropic efficiency 80%.

Organic and electric efficiency 90%.Ambient air 15 °C, p = 0.1013 MPa; humidity 60%

6. Co-firing results for gas turbine de-rating

6.1. Imposed de-rating

Co-firing results are investigated for the full range of fuel ratios (0–100% NG). As previouslymentioned, when only natural gas is considered the results of power plant performance are poorerthan those of highly optimized plants with the same capacities available on the market. The firstset of results presented in this paper is based on the use of de-rating as the gas turbine controlstrategy for biomass-derived gas burning. De-rating is gradually reduced as more natural is appliedup to the point at which temperature reduction is no longer necessary.

When the natural gas content of natural gas in the fuel mix of natural gas reaches a certainlevel gas turbine de-rating is no longer necessary. For the aero-derivative gas turbine the minimumshare of natural gas in the fuel mix that allows compressor operation at the maximum pre-estab-lished pressure ratio 19.3 and maximum cycle temperature 1258 °C is 54.2% (energy basis). Forthe gas turbine identified as medium size the minimum share of natural gas that keeps compressoroperation at 16.4 and maximum cycle temperature at 1288 °C is 36.8%. Approximately the sameproportion of natural gas is estimated for the gas turbine identified as large size, but to keeppressure ratio as 16.2 and maximum cycle temperature as 1371 °C.

6.2. Effects on cycle efficiency

Effects of co-firing on the overall cycle efficiency—as a function of the proportion of naturalgas share in the fuel mix—are presented in Fig. 6 (mass basis) and 7 (energy basis). As expected,there is a continuous increase in cycle efficiency when natural gas content in the fuel mixture is


Fig. 6. Estimated cycle efficiency as a function of the proportion of natural gas in the fuel mixture (mass basis).

Fig. 7. Estimated cycle efficiency as a function of the natural gas in the fuel mixture (energy basis).

increased. There are marked gains in efficiency when the fuel mix is within the range at which de-rating is required, i.e., up to 6–11% of natural gas on the mass basis or to 37–54% in energy basis.

When biomass-derived gas is burned, the estimated cycle efficiency lies between 80 and 87%of the efficiency achievable through the use of natural gas use only. By the end of the de-ratingregion, the cycle efficiency corresponds to 94–95% of the maximum value (burning of naturalgas only). It is noticeable that beyond the de-rating region, the difference between the mediumsize cycle and the large cycle becomes larger (up to 7%) as a result of the higher temperature ofthe largest cycle. When only biomass-derived gas is used the efficiency of the larger power unitis just 5% higher than the efficiencies of the medium size plant. On the other hand, the efficiencyof the cycle based on the aero-derivative gas turbine drops considerably in the de-rating regionbecause of more rigorous constraints imposed by the modeling.


6.3. Effects on output power

The output power is the sum of the gas turbine production and the steam power production,minus the cycle power requirements; the latter includes power consumption by the biomass gasand air compressors. For BIG-CC systems based on atmospheric gasification, the consumption ofsyngas compressor represents a considerable loss on power. In the cases in which only biomass-derived gas is used, compressor power consumption represents between 15–18% of the grosspower produced by the cycle that is based on the aero-derivative machine is the one with thehighest proportion of natural gas. For two of the three cycles studied the effects of the fuel mixon the power production for the gas turbine and the steam cycle and on the overall output arepresented in Figs. 8 and 9. In addition, the same figures show the power consumed by the biomassgas compressor. Results for the plant based on the medium size of gas turbine are similar tothose presented in Fig. 9 since medium-sized and large gas turbines have the same magnitude ofde-rating.

Gas turbine power production reaches its peak at the just at the end of de-rating region, i.e.,around 55% of natural gas in the fuel mix (energy basis) for the aero-derivative gas turbine andbetween 35–37% of natural gas for the industrial gas turbines. Gas turbine power rises as de-rating is reduced, reaching its peak when surge control is no longer needed. Gas turbines canproduce more power when biomass-derived gas is used because the mass flow is larger. However,due to the de-rating, gas turbine output power is only slightly larger when the gas is derived frombiomass. As shown by Souza et al., [18] gas turbine power production would be considerablyhigher if control strategies other than de-rating would be applied.

Power production by the bottoming steam cycle is also affected by gas turbine de-rating,maximum production is also reached at the end of de-rating region. If no de-rating is applied,the more biomass-derived gas is used, the larger is the mass flow of gas turbine exhaust gasesand hence more power is produced. However, when de-rating is imposed the temperature of the

Fig. 8. Effects of the natural gas content in the fuel mix on the power production of the cycle based on the aero-derivative gas turbine.


Fig. 9. Effects of the proportions of natural gas in the fuel mix on the power production of the cycle based on a largeindustrial gas turbine.

exhaust gases is reduced, which reduces steam production rates. Furthermore, steam temperaturealso needs to be reduced to maintain the minimum approach temperature. Simulation results indi-cate that the drop in steam temperature can be avoided if the proportion of natural gas that isabout 20–27% in energy basis.

Overall output power has a similar profile, the maximum value occurring at the end of de-rating region. The aero-derivative-based cycles produce 30% more power and the industrial gasturbines produce 15% more power compared to the natural gas burning However, as can be seenin Figs. 8 and 9 the use of higher proportions of biomass-derived gas considerably affects theoverall power of the plant. The larger the proportion of biomass-derived gas the greater is theamount of auxiliary power consumed through production of biomass gas and, more importantly,the greater is the power consumed by the biomass gas compressor.

7. Co-firing results for gas turbine retrofit

Co-firing is proposed as possible way of reducing problems of using a fuel with low heatingvalues in gas turbines originally designed for natural gas. The complete redesign of gas turbinesmight be a long-term option as the market for BIG-GT cycles develops. A complete redesignwould have to include a new combustion chamber and new expander inlet nozzles. The increasein the expander area would allow a larger flow to pass through the gas turbine without causingan increase in the compressor pressure ratio. Conceptually, the change is similar to that effectedin steam injection gas turbines—STIG that have to cope with increased mass flow. The increasein mass flow in STIG gas turbines is small when compared to the large increase in mass flowneeded when biomass is used.

A brief comparison of the performance of the co-fired schemes and BIG-CC cycles based onequivalent but retrofitted gas turbines is presented below. The increase in the expander inlet area


Fig. 10. Net plant efficiency of BIG-CC units based on retrofitted gas turbines and co-fired schemes.

would allow the gas turbine to run at the same compressor pressure ratio as when natural gas isused without gas turbine de-rating.

Fig. 10 compares the cycle efficiency of the three power plants considered. Fig. 10 is similarto Fig. 4 and introduces horizontal marks that denote the cycle efficiency of equivalent BIG-CCunits based on retrofitted gas turbines. Note that the efficiency levels that could be attained withretrofitted gas turbines can be easily exceeded by co-fired schemes with a low proportion ofnatural gas (e.g. such as 15–20% on energy basis).

However, the main gain that can be achieved with gas turbine retrofitting is output power. Itis estimated that the increase in the expander inlet area allows the power cycle to produce 12%more power than the maximum value achieved for co-fired schemes. In Fig. 11 the co-fired powerplant based on the large gas turbine is compared with a BIG-CC unit based on the retrofitted

Fig. 11. Net power production of BIG-CC units based on retrofitted gas turbines and co-fired schemes.


large gas turbine. The upper line in Fig. 11 corresponds to the net power of a BIG-CC unit basedon a redesigned gas turbine burning biomass-derived gas only.

8. Conclusions

Co-firing is a promising option for dealing with technological problems associated with the useof biomass-derived gas in gas turbines. Co-firing increases the heating value of the fuel implyingthat fewer adaptations have to be made to gas turbines when biomass-derived LCV gas is usedand no other strategy is needed to limit the pressure increase. The avoidance of a control strategyis particularly beneficial as far as de-rating is concerned. On the basis of assumptions used inthis work, it is estimated that co-firing can avoid de-rating for proportions of natural gas between35 and 50% (energy basis). The use of biomass-derived gas increases the flow that goes throughthe GT will lead to important gains in power production. On the other hand, thermal efficiencyis only about 5% lower than the thermal efficiency when natural gas is used.

With co-firing the output power peak occurs at the end of the gas turbine de-rating region inwhich the machine simultaneously achieves its maximum temperature, i.e. no de-rating is neces-sary, and the compressor pressure ratio is at its maximum. The flow of gases is larger than withthe burning of natural gas. This also contributes with the power peak because more steam isproduced in the HSRG.

Co-firing could also prevent the need for dramatic modifications of the gas turbines, the com-bustion chamber would not have to be replaced and, eventually, it might be possible to burnbiomass-derived in dry low NOx systems. However, results show that the latter would be possibleonly with very high proportions of natural gas, (e.g. higher than 90% on energy basis). Furtherstudies are needed in this subject. It is also predicted that co-firing could improve flame stability.

Co-firing was also compared with BIG-CC cycles based on redesigned gas turbines that operateexclusively on biomass-derived gas; this could become a reality in the longer term. With regardto power plant efficiency modeling results show that co-firing could be an option for shares ofnatural gas in the fuel mix as little as 15–20% of natural gas on an energy basis. However,retrofitting of gas turbines could be a better long-term option because it could raise the outputpower by about 10% compared to the peak attained by co-fired systems.

Acknowledgements

Monica Rodrigues is grateful to CNPq and CAPES for their financial support during her workat University of Campinas, Brazil and Utrecht University, The Netherlands.

References

[1] Babu SP. Thermal gasification of biomass technology development: end of task report for 1992 to 1994. Biomassand Bioenergy 1995;9(1–5):271–85.


[2] Faaij A, Van Ree R, Waldheim L, Olssom E, Oudhuis A, Van Wijk A, Daey-Ouwens C, Turkenburg W. Gasifi-cation of biomass wastes and residues for electricity production. Biomass and Bioenergy 1997;12(4):225–40.

[3] Consonni S, Larson ED. Biomass-gasifier/aero-derivative gas turbine combined cycles. Part A—technologies andperformance modeling. Journal of Engineering for Gas Turbines and Power 1996;118(3):507–15.

[4] Cohen H, Rogers CFC, Saravanamuttoo HIH. Gas turbine theory. London: Wesley Logman Limited; 1996.[5] Walter A, Llagostera J, Gallo W. Impact of gas turbine de-rating on the performance and on the economics of

BIG-GT cycles. In: Proceedings of ASME Advanced Energy Systems Division, ASME International MechanicalEngineering Congress and Exposition, Anaheim, USA 1998. pp. 67–72

[6] Walter A, Faaij A, Bauen A. New technologies for modern biomass energy carriers. Electricity production frombiomass. In: Rosillo-Calle F, Bajay SV, Rothman H, editors. Industrial uses of biomass energy—the example ofBrazil. London: Taylor and Francis; 2000. p. 217–28.

[7] Sondreal EA, Benson SA, Hurley JP, Mann MD, Pavlish JH, Swanson ML, Weber GF, Zygarlicke CJ. Reviewof advances in combustion technology and biomass cofiring. Fuel Processing Technology 2001;71:7–38.

[8] Stambler IR. epower steel mills with combined cycles to increase output and cut NOx. Gas Turbine World1999;29(3):26–30.

[9] Thoraval G. Blast furnace gas: an incentive for Italy. Power Engineering International 1998;6(4):78–83.[10] De Kant HF, Bodegom M. Studie Voorschakelling Vergassers voor aardgas Gestookte Energie-Installaties, (Study

on applying gasifiers for co-firing natural gas fired energy conversion facilities). The Netherlands: NOVEM, 2000(In Dutch).

[11] Spath P. Innovative ways of utilizing biomass in a cofiring scenario with a gas turbine integrated combined cyclesystem. Industrial Technologies Division Biomass Power Milestone completion report. Milestone Type: C Golden,CO. National Renewable Energy Laboratory, 1995.

[12] Neilson CE. LM2500 gas turbine modifications for biomass fuel operation. Biomass and Bioenergy1998;15(3):269–73.

[13] Walter A, Rodrigues M, Overend R. Feasibility of cofiring (biomass+natural gas). In: Proceedings of FourthBiomass Conference of the Americas, Oakland CA. Vol. 2. Oxford, UK: Elsevier, 1999. 1321–27.

[14] Hoppesteyn PDJ. Application of low calorific value gaseous fuels in gas turbine combustors. Dissertation presentedat the Technical University of Delft, The Netherlands, Delft: Delft University Press, 1999.

[15] Walter A, Bain R, Overend RP, Craig K. New technologies for modern biomass energy carriers. Gasification forheat and power, methanol and hydrogen. In: Rosillo-Calle F, Bajay SV, Rothman H, editors. Industrial uses ofbiomass energy—the example of Brazil. London: Taylor and Francis; 2000. p. 200–17.

[17] Johnson MS. The effects of gas turbine characteristics on integrated gasification combined—cycle power plantperformance. A Dissertation submitted to the department of mechanical Engineering and the committee on Gradu-ate students of Stanford University. Stanford, CA, September, 1990.

[18] Rodrigues M, Walter A, Faaij A. Performance evaluation of atmospheric BIG-CC systems under different gasturbine control strategies. Submitted to Applied Energy, 2001.

[19] Consonni S, Larson ED. Biomass-gasifier/aero-derivative gas turbine combined cycles. Part B—performance cal-culations and economic assessment. Journal of Engineering for Gas Turbines and Power 1996;118(3):516–25.

[20] Gas turbine world 1999–2000 Handbook. Fairfield: Pequot Publishing, 2000.

Co-firing of natural gas and Biomass gas in biomass integrated gasification/combined cycle systems

Documents

Transcript of Co-firing of natural gas and Biomass gas in biomass integrated gasification/combined cycle systems