SEARCH

SEARCH BY CITATION

Keywords:

  • re-circulating fluidised bed reactor;
  • Chemical-looping combustion;
  • cold model experimental study

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

A re-circulating fluidised bed (RCFB) reactor is proposed here for gaseous fuel combustion using chemical-looping combustion (CLC). A single RCFB reactor was operated in alternate oxidation and reduction cycles for simulating the interconnected reactors arrangement for CLC. For this, a Perspex made transparent, concentric, semicircular cold model of RCFB reactor was used with three grades of Indian Standard sand. Operational parameters like bed inventory, pressure drop, particle size, suspension density, solid circulation rate, bed voidage and fluidisation gas velocity were studied to how they affect the performance of the RCFB reactor. The solid circulation rate has been found to increase with the increase in the bed inventory. Proper solid circulation rate is very much required for proper heat and mass transfer in an interconnected fluidised bed reactors system. The bed voidage is observed to be high during the aeration cycle and low during the reduction cycle. A high bed voidage during the oxidation cycle is appropriate for the fast oxidation reaction of the metal oxides. However, a low bed voidage is required for proper heat transfer for an endothermic reduction reaction in the fuel reactor. Suspension density in the downcomer of the RCFB reactor has been found to be increasing with bed inventory for both the oxidation and reduction cycles, which is good for the bed-to-wall heat transfer. A good observed solid circulation rate, operating voidage and suspension density add to the capability of RCFB to run CLC.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

Atmospheric CO2 concentration has increased from a pre-industrial level of 230 to 430 ppm.[1] Fossil fuels based power plants contribute about one-third of global CO2 emissions.[2] Renewable energy resources such as solar energy, wind energy and biomass may provide one possible way to minimise CO2 emissions.[3] However, most of these alternate energy technologies are in their developmental stage and cannot replace the existing fossil fuel based power generation. So, fossil fuel based power generation with CO2 capture seems to be a future strategy.[4, 5]

Currently, there are a number of available techniques for CO2 capture such as pre-combustion, post-combustion and oxy-combustion techniques.[2] These capture techniques are highly energy intensive, resulting in a 15–20% decrease of overall combustion efficiency of power plants and an increase in the price of electricity produced.[4, 6] It is necessary to find out economical and efficient CO2 capture methods.

Chemical-looping combustion (CLC) is an emerging technology that appears to have the potential for an efficient and economical CO2 capture technology. Initially, this process was proposed to increase the efficiency of thermal power plants; however, later on it was found to have inherent advantage of CO2 capture avoiding costly equipments and energy consumption.[7, 8]

Chemical-Looping Combustion (CLC)

CLC, consisting of an air reactor and a fuel reactor, is a two-step gas combustion process for gaseous fuels and gases produced from gasification that produce almost pure CO2 stream ready for capture, compression and sequestration.[9]

CLC technology involves the use of metal oxides (Fe, Ni, Cu and Mn based), which transfer oxygen from air to fuel avoiding the direct contact between air and fuel. Metal oxides need to circulate between the air reactor and fuel reactor.[10] Interconnected fluidised beds may provide an optimal option for bulk transfer of solids from one chemical environment to another, which is a basic requirement for chemical-looping. The fuel reactor has reduction cycle where oxidised metal oxides, MeO, are reduced by the gaseous fuels according to Reaction (1)[3, 4, 8, 9, 11, 12]:

  • display math(1)

The exit gas stream from the fuel reactor contains CO2 and H2 O vapours, and nearly pure CO2 can be collected after H2 O vapours are condensed. The reduced metal oxides, Me, are transferred to the air reactor where the metal is oxidised (oxidation cycle) according to Reaction (2)[3, 4, 8, 11]:

  • display math(2)

The flue gas stream from the air reactor contains only N2 and some un-adsorbed O2.

CLC Reactor Requirement

For heterogeneous reactions in fluidised, beds proper circulation of bed material has always been considered useful. Bed circulation results in intimate contact between gas and solid particles and an improvement in reactor throughput.[13] Gas and solid phase contact is directly related to reactor design, its configuration and other hydrodynamic considerations like residence time, particle entrainment and pressure drop. Several laboratory designs have been proposed and tested for CLC using cold models and hot prototypes. These designs have indicated that CLC can be run in a variety of configurations with a two-reactors arrangement[2, 5, 7, 11, 14] and a single reactor arrangement.[3, 8, 12, 15, 16] In a single reactor arrangement, the same reactor acts as the air reactor and the fuel reactor alternatively; however, in a two reactors arrangement, one reactor acts as an air reactor and the other as a fuel reactor. Air reactors and fuel reactors have been studied with different operating conditions by various researchers, and a brief overview is presented in Table 1.

Table 1. Operating conditions for the air reactor and fuel reactor in the interconnected fluidised bed arrangement by different researchers
   Bed particlesAir reactorFuel reactor
S. no.Refs.Fuel/gasTypeSize (μ>m)Density (kg/m3)TypeFluidisation velocity (m/s)Temperature (K)Bed voidageParticle residence time (min)TypeFluidisation velocity (m/s)Temperature (K)Bed voidageParticle residence time (min)
1Lyngfelt et al.[11]CH4NiO/YSZ2005000High velocity bed5.712430.6Bubbling bed0.5312230.6
2Johansson et al.[45]AirSilica1502600Fast fluidised bed0.75–1.15Bubbling bed0.09–0.31
3Brandvoll and Bolland [46]CH4NiO/YSZFast fluidised bed14730.95Bubbling bed8330.45
4Pröll et al.[34]AirBronze powder558730High velocity bed4.25298Low velocity bed1.21298
5de-Diego et al.[47]CH4Cu14Al’γ Al2 O3100–300 and 200–5001500 and 1560Bubbling Bed0.51073Bubbling Bed0.11073

The most talked about interconnected fluidised beds configuration for CLC is presented in Figure 1. Here, the reactor is a high velocity bed with a fast fluidisation regime, and the fuel reactor is a low velocity bubbling bed.

A cyclone separator is used to recover the bed material, leaving the air reactor then is diverted to fuel reactor. The bed material flows from fuel reactor to air reactor under the influence of gravity, so the fuel reactor has to be placed slightly elevated as compared to the air reactor. Loop seals are used in the CFB configuration: one between the air and fuel reactor, and the other between the cyclone separator and fuel reactor. Loop seals avoid leakage of N2 into the fuel reactor and CO2 into the air reactor. High velocity in the air reactor provides the required driving force for solid circulation between the two reactors.[11]

Issues Related to Existing CFB Configuration for CLC

In any fluidised bed system like CLC, for example, the bed material is in intense turbulence and thus inevitably subjected to mechanical stress due to inter-particle collisions and bed-to-wall impacts, which is known as attrition. The main consequence of attrition in CLC is the generation of fines, which are finally passing through the dust recovery systems resulting in a loss of valuable metal oxides. The generation and loss of fines from the system will lead to decrease in the combustion efficiency.[17] In a CFB arrangement, three regions can be identified as main attrition sources, namely the cyclone separator section, the grid jets and the bubbling bed itself.[18] Werther and Reppenhagen[17] found the cyclone separator attrition to be higher than the jet and bubble induced attrition at higher velocities; however, the gas distributor is the main attrition source in the lower velocities.

Transferring bed material from one reactor to another is of great concern in any fluidised bed combustion system, and nearly half of the chemical plants involving fluidised beds suffer from this problem.[19] Non-mechanical valves like loop-seals are commonly employed to control solid flow rates in CFB configuration. These have a special application in high pressure and elevated temperature combustion systems, like CLC.[20] However, loop-seals in the CFB require a proper pressure balance around them. Otherwise, their operation becomes irrational and seal failure may occur.[21] In this scenario, the aeration air from the air reactor and un-burnt gaseous fuel from the fuel reactor may escape through the cyclone separator, affecting the overall efficiency of the CLC system. Above all, at a given gas velocity and solid circulation rate, pressure drop across the loop-seal increases linearly with increasing solid inventory in the bed. Also, excessive aeration may increase the voidage in the loop seal and hence low heat transfer.[20]

In an interconnected reactors configuration, the heat loss may occur through cyclone separator, loop-seal and stand pipe, other than the reactor's walls.[22] Post-combustion of combustibles (carbon monoxide and residual carbon) are often found in the cyclone separators of the CFB boilers. Post-combustion could result in a temperature increase of flue gas in the cyclone separator by 30–50C. This may cause an overheating problem with the re-heaters, and if un-considered, may cause extra heat loss and thus lower CLC boiler efficiency.[23]

The fast fluidisation regime in the air reactor is a complex phenomenon.[24] In the air reactor (riser), particle cluster formation takes place.[25] The core region of the CFB riser has a very few upward moving clusters; however, the annulus region has large downward flowing clusters. This flow pattern in the riser results in the back mixing of bed particles, which ultimately affects the overall heat transfer rates and the chemical interactions between metal oxides and the gases.[26, 27]

Various reactor modifications and alternative designs have been proposed to overcome these problems. In this paper, a re-circulating fluidised bed (RCFB) reactor is proposed for CLC with a feeling that it can overcome the problem of gas leak, low residence time of the bed material and proper solid circulation between the air and fuel reactors.

image

Figure 1. Two interconnected fluidised beds (CFB) configuration for CLC.[11]

Download figure to PowerPoint

Re-Circulating Fluidised Bed (RCFB) Configuration for CLC

RCFB is a spouted fluidised bed with a draft tube.[28] It involves two concentric tubes; the inner tube is known as draft tube, and the outer tube is known as downcomer. Draft tube is used to induce a high solid circulation rate in the reactor.[29] Aeration gas is fed to the base of an open draft tube section such that the bed material is picked up pneumatically in the draft tube. This creates an overall bed density difference between the draft tube and downcomer section, which drives the solid circulation pattern upward through the draft tube and downward in the downcomer. Bed material and gases flowing out of the draft tube moves into an expanded cross-section of the reactor (freeboard), which is used to retain the outgoing bed particles. Solids from the top of the draft tube flow into the downcomer and enter the base of the draft tube.[30]

RCFB reactor configuration was initially developed for maintaining uniform bed temperature in oil and coal gasifiers[30]; however, later on Yang and Keairns[31] proposed RCFB boiler as an alternative to the pressurised fluidised bed combustion boiler, where combustion would take place in the draft tube. RCFB reactor has never been used for CLC. This paper explores the usability of RCFB for CLC. RCFB can be used in the following configurations for chemical looping combustion:

  1. An interconnected reactor arrangement, where one RCFB reactor is the air reactor and the other a fuel reactor.
  2. A single RCFB reactor where the central draft tube acts as the air reactor and the downcomer as the fuel reactor. In this arrangement, proper seal may be required to avoid inter-mixing of the exhaust gases between the air reactor and the fuel reactor.

Simulating the first configurations, a cold model RCFB was operated in alternate cycles as the air reactor and fuel reactor. This paper describes the basic principles and the reactor architecture for the first reactor configuration only, together with results obtained from cold model testing. Use of a freeboard in place of a cyclone separator in the RCFB configuration may be a positive point for this arrangement, as the cyclone separator does not account for any heat loss. A cyclone separator free arrangement, in addition, will ensure lesser generation of fine particles due to low attrition. Absence of loop-seals makes this reactor configuration simple. Absence of loop-seals and standpipe, all together, will reduce the heat loss from the system. Also, draft tube in RCFB configuration minimises the back mixing phenomena of the bed particles and hence improves the bed to wall heat transfer. Reactor architecture and the cold model results for second RCFB configuration for CLC will be discussed in a forthcoming paper.

COLD MODEL RCFB CONFIGURATION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

Fluidised beds are designed on the basis of previous experience and hydrodynamic experiments in a half to full sized cold model experimental facility, since cold models of reduced capacity and size can be operated very easily.[32-34] Operating parameters are easy to measure in reactors operating at ambient conditions than in reactors operating at higher temperature or pressure. Felice et al.[35] has experimentally validated a large number of cold model fluidised beds. However, in case of non-fluid dynamic effects, like particle sintering at higher temperature, cold model predictions may lose their reliability.[13]

A Perspex-made transparent, concentric, semi-circular RCFB reactor was fabricated (Figures 2 and 3). Berruti et al.[36] has quoted that hydrodynamic data obtained from a semi-circular column are same to those obtained from a full column. Above all, transparent cold models facilitate visual observation of the bed operations.[29, 37] The cold model has four sections, namely jet tube section, spacer section, draft tube section and freeboard section. The draft tube section has semicircular concentric pipe arrangement and is of 1 m height. The inner diameters of the outer pipe and inner pipe are 0.15 and 0.05 m, respectively. The detailed dimensions are shown in Figure 2. Spacer section mentioned as ‘x’ was changed during experiments (0.03, 0.08 and 0.15 m). A minimum of 4 kg and a maximum of 10 kg of bed inventory were used during the experiments.

image

Figure 2. Recirculating fluidised bed cold model experimental facility.

Download figure to PowerPoint

image

Figure 3. Picture of the cold model experimental facility.

Download figure to PowerPoint

Operating parameters like bed height, bed inventory, pressure drop, bed particle size, solid circulation rate, bed density, bed voidage and fluidisation air velocity were studied, since they affect the performance of fluidised beds. Solid circulation rate is a very sensitive parameter in operating chemical looping, since solids act as oxygen and heat carriers between oxidation and reduction sections.

In the present study, a single RCFB reactor was operated in the alternate oxidation (air reactor/aeration cycle) and reduction (fuel reactor/fuel burning cycle) cycles by providing the appropriate regimes of the fluidisation. In this reactor configuration, there was a fast fluidisation regime in the riser of the RCFB reactor during the oxidation cycles, such that the spout of the bed material emanating from the riser reached 0.5 m above the top of the riser tube and did not escape out of the reactor. However, in the reduction cycle, there was a slugging bed regime in the downcomer of the RCFB reactor, such that there was proper circulation of bed material from the top of the draft tube to the top of the downcomer section. High velocity aeration cycle is for low residence time of bed material, and low velocity fuel burning cycle is for long residence time of bed material. The regimes were identified after visual observation. The generated experimental data, corresponding to these regimes in the RCFB, is presented here.

In an actual RCFB CLC system with metal oxides, the first cycle will be an oxidation cycle, where air will be used for fluidisation of the metal oxides. After the completion of the oxidation cycle, the metal oxides will have fuel gas for fluidisation. An inert gas can be purged from the RCFB reactor between the oxidation and reduction cycles to avoid the intermixing of the atmospheric air and fuel gas. However, in this study, in place of metal oxides, Indian standard sand of three different sizes was used as bed material. The properties of the bed material are given in Table 2. Ambient air was used for fluidisation of the bed material in oxidation as well as reduction cycles.

Table 2. Characteristics of the bed material
S. no.Sand gradeGeldart's classificationAverage particle size (mm)Density (kg/m3)Dense bed voidage
1.Grade I (2–1 mm)D1.326220.35
2.Grade II (1–0.5 mm)B0.526060.613
3.Grade III (0.5–0.09 mm)B0.3526000.707

MEASUREMENTS AND COMPUTATIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

Pressure taps were provided for measuring differential pressure drops between points 1 to 4. Pressure drop was measured in the riser ΔP1–2, the downcomer ΔP4–3 and between the riser bottom and the downcomer bottom ΔP1–4 for the different operating conditions using Mercury (Hg) filled U-tube manometers. Air from the air compressor was supplied to the cold model using 0.025 and 0.03 m semicircular jet tubes; flow of air was monitored by a rotameter. The aluminium-made jet tube supplied air for fluidisation to the bed material resting on a stainless steel perforated plate. Wire mesh was provided along with perforated plate to avoid weeping of fine bed material during operation. The main reactor was separated from the jet tube section with a spacer. The particle velocity in the downcomer Upd was estimated using coloured tracer particles of the same size as that of the fluidising media. A sufficient amount of tracer particles were mixed with the bed material to facilitate easy and early estimation of particle velocity. The velocity was calculated by measuring the time required to travel a known distance by the coloured particles. Five coloured particles were traced down in the downcomer section, such that their arithmetic mean represents the particle velocity in the downcomer. The operating voidage in the riser er was calculated by the equation suggested by Alappat and Rane[38]:

  • display math(3)

where Hdt is the height of the spout under stable operation, Hs is the height of the bed in the riser when the air flow is suddenly cut off and the riser bed is static, and eis the dense bed voidage of the bed material as presented in Table 2. The operating voidage in the downcomer ed was calculated using the equation[38]:

  • display math(4)

where i is the total bed material in the reactor for a particular run, Hdc is the operating downcomer bed height, Ad is the cross-sectional area of downcomer, Ar is the cross-sectional area of the riser, and ρp is the particle density. These values were noted for every run corresponding to the stable operation in the RCFB reactor. The particle velocity in the riser Upr was calculated using the following relation[39]:

  • display math(5)

Solid circulation rate Ws was calculated by Equation (6), in which the particle velocity in the downcomer Upd was calculated using tracer particles (measuring the time required by the particles to travel a known distance):

  • display math(6)

The suspension density ρsus of the bed at the downcomer wall was calculated as quoted by Wang et al.[40]:

  • display math(7)

For the design of experiments, variables like air flow rate, spacing between draft tube and the jet tube section and bed inventory were varied. The experiments were divided into 18 series (Table 3). There is a limit on the minimum and maximum inventory used inside the reactor once the reactor dimensions have been fixed. Below 6 kg of bed inventory, there was no proper operation of the fluidised bed. Also, above 10 kg of bed inventory, the reactor cannot be operated due to choking. The fluidisation air velocity for the aeration cycle (fast fluidisation regime) was in the range of 7.13–9.17 m/s. On the other hand, for the fuel burning cycle (slugging bed regime), it was in the range of 1.01–2.03 m/s. Similar considerations was applied for the selection of the spacer section between the perforated plate and the bottom of the riser tube. A very small spacer section will lead to problems related to the proper circulation of the bed material in the reactor. However, a very large spacer section will lead to high gas bypassing from the jet tube to the downcomer section.

Table 3. The scheme of the cold model experiments
TestJet tubeSpacerSandInventory
seriesdiameter (m)section (m)grade(kg)
10.0250.03I4, 6, 8, 9
20.0250.08I6, 8, 9
30.0250.15I8, 9, 10
40.0250.03II4, 6, 8, 9
50.0250.08II6, 8, 9
60.0250.15II8, 9, 10
70.0250.03III4, 6, 8, 9
80.0250.08III6, 8, 9
90.0250.15III8, 9, 10
100.030.03I4, 6, 8, 9
110.030.08I6, 8, 9
120.030.15I8, 9, 10
130.030.03II4, 6, 8, 9
140.030.08II6, 8, 9
150.030.15II8, 9, 10
160.030.03III4, 6, 8, 9
170.030.08III6, 8, 9
180.030.15III8, 9, 10

EXPERIMENTAL RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

In this study, a single RCFB reactor was operated in the alternate oxidation and reduction cycles. There was the fast fluidisation regime in the riser of the RCFB reactor during the oxidation cycles and the slugging bed regime in the downcomer of the RCFB reactor during the reduction cycle. The generated experimental data, corresponding to these regimes in the RCFB, are presented here in Tables 4 and 5.

Table 4. Experimental data for different operating conditions corresponding to 0.025 m jet tube diameter
x (m)i (kg)Sand gradeFluidisation air velocity during aeration cycle (m/s)Fluidisation air velocity during the fuel burning cycle (m/s)Suspension density in the downcomer during fuel burning cycle (kg/m3)Suspension density in the downcomer during aeration cycle (kg/m3)Suspension density in the riser during fuel burning cycle (kg/m3)Suspension density in the riser during aeration cycle (kg/m3)Solid circulation rate during fuel burning cycle (kg/s)Solid circulation rate during aeration cycle (kg/s)ΔP1–4 for fuel burning cycle (cm of H2 O)ΔP1–4 for air reactor cycle (cm of H2 O)ΔP4–3 for fuel burning cycle (cm of H2 O)ΔP4–3 for air reactor cycle (cm of H2 O)ΔP1–2 for fuel burning cycle (cm of H2 O)ΔP1–2 for air reactor cycle (cm of H2 O)
0.034I9.174.071756.741835.40175.57136.560.070.3840.543.21.41.235.137.8
0.034II5.091.012136.921120.58128.44128.440.070.165456.744.256.754
0.034III5.091.011222.00936.00106.2091.030.130.2078.378.310.41064.881
0.036I9.173.051809.181861.62214.59175.570.170.3940.5541.84.240.543.2
0.036II5.091.012189.041537.54176.61144.540.110.1775.670.29.79.656.754
0.036III5.091.011820.001430.00166.89121.370.330.3783.7811110.678.381
0.038I9.173.051835.401861.62214.59175.570.170.4340.575.62.46.240.556.7
0.038II5.091.012189.041980.56192.67144.500.110.2097.291.811.613.267.567.5
0.038III5.091.011976.001898.00166.89151.710.330.4589.181111181.581
0.039I9.173.051861.621887.84234.10195.080.180.4545.986.43.87.840.567.5
0.039II5.091.012189.042110.86208.72176.610.280.28110.799.912.413.674.275.6
0.039III5.091.012132.002132.00212.40166.890.430.4589.189.1111183.781
0.086I9.174.071599.421782.96370.66214.590.050.3362.959.23.53.237.854
0.086II6.111.012058.741641.78208.72160.560.080.1578.386.414.414.478.3108
0.086III5.091.011716.001352.00136.54121.370.120.1581102.66878.394.5
0.088I9.173.051756.741782.96370.66234.100.170.3867.575.66643.256.7
0.088II6.111.012110.861954.50224.78160.560.090.1783.710814.816.678.3116.1
0.088III5.091.012158.001820.00136.54136.540.130.1783.7110.711.2138199.9
0.089I9.173.051809.181835.40370.66234.100.190.4067.581.0785467.5
0.089II6.111.012110.861980.56224.78160.560.10.1897.2110.7162099.9118.8
0.089III5.091.012236.002028.00182.06166.890.130.1891.8129.613.414.2102.6108
0.158I9.173.051782.961782.96448.70292.630.100.689.189.1316.451.354
0.158II8.151.012032.681667.84208.72144.500.080.24108118.824.423110.7108
0.158III7.131.012184.001690.00151.71121.370.030.12135121.5302597.2108
0.159I9.173.051782.961809.18448.70312.140.180.6794.591.85.217.65467.5
0.159II8.151.012058.741876.32256.89176.610.100.24113.4124.227.823121.5110.7
0.159III7.131.012236.001924.00151.71121.370.050.14148.5129.63033124.2135
0.1510I9.173.051809.181809.18448.70312.140.230.68113.4108619.664.867.5
0.1510II8.151.012084.802006.62289.00176.610.260.34121.5132.33029.8126.9121.5
0.1510III7.131.012262.002132.00151.71121.370.060.14162156.63037128.2162
Table 5. Experimental data for different operating conditions corresponding to 0.03 m jet tube diameter
x (m)i (kg)Sand gradeFluidisation air velocity during aeration cycle V (m/s)Fluidisation air velocity during the fuel burning cycle V (m/s)Suspension density in the downcomer during fuel burning cycle (kg/m3)Suspension density in the downcomer during aeration cycle (kg/m3)Suspension density in the riser during fuel burning cycle (kg/m3)Suspension density in the riser during aeration cycle (kg/m3)Solid circulation rate during fuel burning cycle (kg/s)Solid circulation rate during aeration cycle (kg/s)ΔP1–4 for fuel burning cycle (cm of H2 O)ΔP1–4 for air reactor cycle (cm of H2 O)ΔP4–3 for fuel burning cycle (cm of H2 O)ΔP4–3 for air reactor cycle (cm of H2 O)ΔP1–2 for fuel burning cycle (cm of H2 O)ΔP1–2 for air reactor cycle (cm of H2 O)
0.034I8.154.071730.521520.76409.68234.100.070.4451.345.933.243.245.9
0.034II7.132.031667.841329.06176.61160.560.040.1948.662.12.6951.348.6
0.034III7.132.031976936151.71121.370.070.2548.65453.243.235.1
0.036I8.154.071887.841887.84409.68234.100.290.525462.13.21048.645.9
0.036II7.132.031772.081693.90192.67160.560.0850.2178.394.511.4205467.5
0.036III7.132.0320281404166.89121.370.130.2672.994.51219.264.845.9
0.038I8.152.031914.061914.06409.68234.100.300.5956.778.3819.648.656.7
0.038II7.132.032058.741772.08192.67176.610.150.5694.594.5232359.467.5
0.038III7.132.0320801898166.89121.370.160.448194.5192367.559.4
0.039I8.152.031940.281914.06409.68253.610.320.6767.594.51428.85475.6
0.039II7.132.032189.042032.68192.67192.670.321.1397.297.224.82659.472.9
0.039III7.132.0321582132166.89121.370.180.448197.22023.870.272.9
0.086I8.153.051730.521678.08370.66253.610.040.3254818840.540.5
0.086II6.112.032006.621824.2192.67144.500.140.2945.9543.6443.248.2
0.086III6.112.0316121352151.71106.200.170.3054544243.245.9
0.088I8.153.051782.961809.18390.17253.610.170.601081089205454
0.088II6.112.032058.741902.38192.672890.160.32132.3148.57125454
0.088III6.112.0317941794151.71136.540.170.3397.294.5121259.456.7
0.089I8.153.051835.401835.40390.17273.120.210.62108113.410205454
0.089II6.112.032058.741928.44208.72433.510.180.32162167201964.859.4
0.089III6.112.0320542028166.89166.890.200.34108108202062.162.1
0.158I9.174.071678.081704.30409.68312.140.140.5368.886.431756.754
0.158II8.151.012058.741798.14192.67160.560.040.3191.891.8161570.267.5
0.158III6.111.0121061950166.89121.370.030.26126.994.517.41775.681
0.159I9.174.071704.31730.52429.19331.640.280.7072.986.48.41856.754
0.159II8.151.012084.81928.44208.72160.560.050.33124.2121.516168170.2
0.159III6.111.0122622002182.06136.540.030.2813513517.42475.681
0.1510I9.174.071861.621756.74487.72331.640.380.7194.5121.5132872.994.5
0.1510II8.151.012084.801980.56256.89208.720.050.35129.6140.4162686.483.7
0.1510III6.111.0123402132182.06136.540.070.29151.2162242491.8108

Fluidisation Air Velocity

The fluidisation air velocity V in the riser (corresponding to which there was fast fluidisation regime and a slugging bed regime in the RCFB reactor) were noted down and are presented in Tables 4 and 5. The effect of spacer section, jet tube diameter and the bed inventory on the fluidisation air velocity necessary for maintaining the required regimes in the RCFB reactor are discussed here. During the air reactor cycle, the fluidisation air velocity was in the range of 5.09–9.17 m/s. However, during the fuel reactor cycle it was in the range of 1.01–4.07 m/s. For a given spacer section and grade of sand, the fluidisation air velocity in the riser of the air reactor is independent of change of bed inventory. In case of the fuel reactor for a given spacer section and grade of sand, the fluidisation air velocity in the riser for the low bed inventory is higher compared to high bed inventory (test series 1 [6 kg], test series 2 [6 kg] and test series 10 [4 and 6 kg]), which could be due to insufficient bed inventory and inability of the reactor to achieve the desired regimes of fluidisation in stable condition.

Keeping the spacer section fixed, the fluidisation air velocity required to maintain the desired regime of fluidisation in the RCFB decreases with the decrease in particle size. During aeration cycle corresponding to the 0.15 m spacer section and 0.03 m jet tube diameter, the fluidisation air velocity was 9.17 m/s for sand grade I, 8.15 m/s for sand grade II and 6.11 m/s for sand grade III. For both of the air reactor and fuel reactor, the fluidisation air velocity in the riser required to maintain the desired regime of fluidisation increased with an increase in the spacer section. For the aeration cycle corresponding to the sand grade I with 0.03 m jet tube diameter, the fluidisation air velocity in the riser was 8.15 m/s for 0.03 m spacer section, 8.15 m/s for 0.08 m spacer section and 9.17 m/s for 0.15 m spacer section. This is due to the gas bypassing phenomenon that increased with increase in the space between the perforated plate and the bottom of the draft tube.

Solid Circulation Rate

After the identification of the desired regime of fluidisation and the corresponding draft tube superficial air velocity in the reactor, the values of the hydrodynamic parameters were noted. Solid circulation rate is the most critical variable in predicting the performance of a RCFB reactor.[38] The values of the solid circulations rate observed during the air reactor and fuel reactor cycle are presented in Tables 4 and 5.

The variation of solid circulation rate with the change of bed inventory and bed particle size for 0.03 m spacer section, 0.025 and 0.03 m jet tube diameter are presented in the Figure 4. A similar trend was observed for 0.08 and 0.15 m spacer section for 0.025 and 0.03 m diameter jet tube section. For 0.025 m jet tube diameter, a maximum solid circulation rate of 0.68 kg/s was observed for the air reactor and 0.43 kg/s was observed for the fuel reactor (series 9). However, with 0.03 m diameter jet tube section, a maximum of 1.13 kg/s was observed for the air reactor and 0.38 kg/s was observed for the fuel reactor (series 18).

image

Figure 4. Variation of the solid circulation rate with bed inventory during (a) the aeration cycle (b) fuel burning cycle for 0.03 m spacer section with 0.025 m diameter jet tube.

Download figure to PowerPoint

As expected, for a given spacer section the solid circulation rate increased with an increase in the bed inventory for each grade of sand. Also, for a given grade of sand, the solid circulation rate increased with increase in the space between the perforated plate and the bottom of the draft tube in the range tested. This could be due to the enhanced entrainment zone. On the other hand, the solid circulation rate increases with increase in the particle size for a fixed spacer section and jet tube diameter. This could be due to fact that the coarse particles are more spoutable (sand grade I) compared to the fine sand particles (sand grade II and III).[41]

Operating Voidage

The operating voidage plays an important role in the design of the fluidised beds since it affects the erosion, mass transfer and bed-to-wall heat transfer.[42, 43] Voidage study also helps in the scale-up and design of the fluidised beds.[44] The effect of the spacer section, jet tube diameter and the grades of sand on the operating voidage were studied. The results of the operating voidage for the riser and the downcomer of the aeration cycle and fuel burning cycle for sand grade I and jet tube diameter 0.025 m are presented in Figures 5 and 6. A similar trend was observed for the riser and downcomer in the case of sand grade II and sand grade III with jet tube diameter 0.03 m.

image

Figure 5. Variation of the operating voidage with bed inventory in the (a) riser (b) downcomer during the aeration cycle for sand grade I and 0.025 m diameter jet tube.

Download figure to PowerPoint

image

Figure 6. Variation of the operating voidage with bed inventory in the (a) riser (b) downcomer during the fuel burning cycle for sand grade I and 0.025 m diameter jet tube.

Download figure to PowerPoint

For a given grade of sand and jet tube diameter, the operating voidage decreased with an increase of the bed inventory in the range of bed inventory tested in this study. This low operating voidage at high bed inventory may be good for proper heat transfer in the RCFB combustion system, since it will increase the bed-to-wall heat transfer coefficient. Moreover, the operating voidage is higher in the riser of the air reactor (0.84–0.94 for 0.023 m jet tube and 0.73–0.92 for 0.030 m jet tube diameter) and fuel reactor (0.77–0.93 for 0.023 m jet tube and 0.75–0.90 for 0.030 m jet tube diameter) compared to the operating voidage in the downcomer (Table 6). This is because the jet tube supplies aeration air directly into the riser and there is no aeration mechanism for the downcomer.

Table 6. Operating voidage range for the aeration and fuel burning cycles
   SandOperating voidage range forOperating voidage range for
S. no.CyclesSectionSand grade0.025 m diameter jet tube diameter0.03 m diameter jet tube diameter
1AerationDowncomerI0.28–0.320.27–0.48
2AerationDowncomerII0.19–0.570.22–0.49
3AerationDowncomerIII0.18–0.640.18–0.64
4AerationRiserI0.84–0.930.83–0.88
5AerationRiserII0.89–0.920.73–0.90
6AerationRiserIII0.89–0.940.89–0.92
7FuelDowncomerI0.29–0.390.26–0.37
8FuelDowncomerII0.16–0.220.16–0.36
9FuelDowncomerIII0.13–0.530.17–0.38
10FuelRiserI0.77–0.910.75–0.81
11FuelRiserII0.82–0.920.84–0.89
12FuelRiserIII0.86–0.930.88–0.90

In the reactor arrangement with 0.025 m diameter jet tube section (as compared to 0.03 m diameter jet tube section), the operating voidage was high in the riser of the RCFB system. This may be attributed to less gas bypassing in the 0.025 m diameter jet tube.

For a particular grade of sand, the operating voidage in the riser during the aeration and fuel burning cycle decreased with an increase in the spacer section in the range tested. This could be due to the gas bypassing to the downcomer section in case of an increased spacer section. Conversely, for the downcomer during the aeration and fuel burning cycle, the operating voidage increased with an increase in the spacer section. General operating voidage ranges in the riser and downcomer for the aeration and fuel burning cycles are mentioned in Table 6.

Pressure Drop

The experimental pressure drop across the riser ΔP1–2, the downcomer ΔP4–3 and between the riser bottom and the downcomer bottom ΔP1–4 were measured and are presented in Tables 4, 5 and Figure 7. The effect of different operating conditions on pressure drop was found to be similar for the riser and downcomer. Hence, the effects of different parameters are discussed only for the pressure drop across the riser.

image

Figure 7. Variation of the pressure drop with bed inventory in the riser during the (a) aeration cycle (b) fuel burning cycle for sand grade I with 0.025 m diameter jet tube.

Download figure to PowerPoint

For a given grade of sand and jet tube diameter, the pressure drop is increased with increase in the bed inventory in the reactor for each grade of sand. This is due to the fact that the higher inventory has more resistance to the flow and hence a high-pressure drop. Another observation was that the pressure drop varies considerably with inventory for fine sand particles and was less prominent with coarse sand particles. For example, in the riser of the fuel reactor, for 0.025 m jet tube and 0.15 m spacer section and when the bed inventory is increased from the 8 to 9 kg, the pressure drop increases from 51 to 54 cm of water in the case of sand grade I (coarse sand), while for sand grade II (fine sand), it increases from 110 to 121 cm of water. The effect of inventory is more significant with the fine particles compared to the coarse particles.

For a particular grade of the sand, the pressure drop increased with an increase in the spacer section in the range investigated here. This may be due to the increased bed material between the draft tube bottom and the perforated plate. This bed inventory causes the gas by passing and reducing the flow rate of the aeration air through the riser section.

For a given spacer section and jet tube diameter, the pressure drop is more for the fine sand (grade II and III) than the coarse sand (sand grade I). This may be due to the fact that the fine sand particles have a large surface area, offer larger friction between the wall and have higher tendencies for cluster formation.[39] For example, for the riser during the aeration cycle with 0.025 m jet tube diameter and the 0.08 m spacer section, the pressure drop in the riser for sand grade I particles for 8 kg bed inventory was 57 cm, whereas the same for the sand grade III particles was 100 cm of water.

Suspension Density

The suspension density for the downcomer during the aeration and fuel burning cycles was measured, and the results are presented in Tables 4, 5 and Figures 8, 9. The suspension density plays an important role in the bed-to-wall heat transfer coefficient.

image

Figure 8. Variation of the suspension density with bed inventory in the riser during the (a) aeration cycle (b) fuel burning cycle for sand grade I with 0.025 m diameter jet tube.

Download figure to PowerPoint

image

Figure 9. Variation of the suspension density with bed inventory in the downcomer during the (a) aeration cycle (b) fuel burning cycle for sand grade I with 0.025 m diameter jet tube.

Download figure to PowerPoint

For a given grade of sand and jet tube diameter, the suspension density increased with an increase of the bed inventory. This is due to reduction of the operating voidage in the downcomer. For example, for the 0.025 m jet tube diameter and 0.15 m spacer section and the 8 kg bed inventory, the suspension density for sand grade II was 1667 kg/m3, whereas for 9 kg bed inventory, it was 1876 kg/m3.

The suspension density decreased with an increase of the spacer section for a given grade of sand. This could be due to gas passing, which causes the flow of aeration air to the downcomer and thus reduces the suspension density. For example, for the downcomer of the fuel reactor with 0.025 m jet tube, sand grade II and 8 kg bed inventory, the suspension density for the 0.03 m spacer section is 2189 kg/m3, whereas for 0.08 m spacer section, the suspension density is 2110 kg/m3 and 2032 kg/m3 for 0.15 m spacer section.

The suspension density decreased with the decrease of the particle size of the bed inventory for a fixed spacer section and jet tube diameter. For example, for the downcomer of air reactor with 0.025 m jet tube diameter, 0.03 m spacer section and 9 kg of bed inventory, the suspension density for the sand grade I is 1887 kg/m3, whereas for sand grade II and III, it is 2110 and 2132 kg/m3, respectively.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References

An RCFB reactor has never been used for CLC. A cold model of an RCFB reactor was fabricated and tested for its usability for CLC. A single RCFB was operated alternatively as an air reactor with fast fluidisation regime in the riser and as a fuel reactor with slugging bed regime in the downcomer. The effects of various operating and design parameters were studied on the fluidisation air velocity required to maintain the required regime of fluidisation in the reactor, solid circulation rate, operating voidage, pressure drop and suspension density for Indian standard sand of grade I–III. The fluidisation gas velocity in the riser was in accordance with the required fluidisation regime in the reactor.

It was found that the fluidisation air velocity in the riser was independent of the bed inventory; however, it was decreasing with a decrease in the particle size. The increase in the spacer section length increased the fluidisation air velocity in the riser required for maintaining the desired regime of fluidisation for both aeration and fuel burning cycles.

The solid circulation rate increased with an increase in the bed inventory and length of the spacer section; however, it decreased with an increase of particle size. Also, operating voidage decreased with an increase in the bed inventory. It increased with an increase in the jet tube diameter and with a decrease in the particle size.

The pressure drop in the riser increased with an increase of the bed inventory and spacer section. For fine bed particles (sand grade II and III), the pressure drop was larger than the coarse one (sand grade I). Suspension density increased with an increase of bed inventor. However, it decreased with an increase of spacer section and decrease of particle size.

With oxidation of metal oxides, an exothermic reaction and reduction and an endothermic reaction, proper circulation of bed material and heat plays an important role in maintaining high combustion efficiency of the fuel in the CLC reactor system. A good solid circulation rate has been observed between air and fuel reactors. The low observed operating voidage in the downcomer of RCFB will ensure good bed-to-wall heat transfer, which is very much required in a combustion system. All these merits add to the usability of RCFB for CLC. Another RCFB configuration is possible for CLC, where the central draft tube acts as the air reactor and the downcomer as the fuel reactor. The reactor architecture and cold model results for this configuration will be discussed in a forthcoming paper.

NOMENCLATURE

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References
Ad

cross-sectional area of downcomer (m2)

Ar

cross-sectional area of riser (m2)

e

dense bed voidage

ed

operating voidage in the downcomer

er

operating voidage in the riser

Hdc

operating downcomer bed height with respect to the perforated plate (m)

Hdt

operating height of the spout with respect to the perforated plate (m)

Hs

height of the bed in the riser with respect to the perforated plate when the air flow is suddenly cut off (m)

i

total bed material in the reactor (kg)

ΔP1–2

pressure drop in the riser (kg/m s2)

ΔP1–4

pressure drop in between the riser bottom and downcomer bottom (kg/m s2)

ΔP4–3

pressure drop in the downcomer (kg/m s2)

Upd

particle velocity in the downcomer (m/s)

Upr

particle velocity in the riser (m/s)

Ws

solid circulation rate (kg/s)

x

length of spacer section (m)

Greek Symbols

ρp

solid particle density (kg/m3)

ρsus

suspension density (kg/m3)

Abbreviations

CFB

circulating fluidised bed

CLC

chemical-looping combustion

Me

reduced metal oxide

MeO

oxidized metal oxide

RCFB

re-circulating fluidised bed

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COLD MODEL RCFB CONFIGURATION
  5. MEASUREMENTS AND COMPUTATIONS
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSIONS
  8. NOMENCLATURE
  9. References