CASE STUDY ON OPERATIONAL INSPECTION OF A FLUIDISED BED

COMBUSTION BOILER AT A TEXTILE INDUSTRY

The agenda for the visit was as below

1. Final steam temperature is not being achieved even at 85- 90 % of MCR Load.
2. Low back end temperature against design value
3. High unburnt at deferent location of the boiler i.e. Cage wall hopper, ECO, APH and ESP
4. Low Furnace gas Plenum temperature than design even at high operating bed temperature
970 - 980 Deg C

The boiler operation was seen for a duration of four days. The boiler no 1 was under shut down at
the time of the visit. The first day of the visit was used up in reviewing the past operational data and
the feedback was taken from operational team. On the second day, the proposal for revised
operational parameters was discussed & implemented. 3rd & 4th day were used for effecting gradual
change of operational parameters. The parameters were modified to a large extent. Results on these
parameters modifications with respect to agenda are briefly described below. Further to this
detailed discussion on the recommended boiler parameters is attached.

Results

Final steam temperature is not being achieved even at 85- 90 % of MCR Load.

The air flow was less to enable complete combustion of the fuel fed. After admitting the required
air for combustion, the effect could be seen. The spray took place at PSH outlet after the correction
of air flow. It is advised to maintain a minimum Oxygen level of 4.5% in flue gas after all the air
ingress locations are checked & arrested.

Low back end temperature against design value

We have to check if there is air ingress downstream in the economizer casing & APH. If outside air
ingress is present then there will be low back end temperature. The air should have got heated to
153 deg C when the gas outlet from economizer is 230 deg C. Only on shut down inspection this
can be figured out by inspecting for possible air ingress.

High unburnt at deferent locations of the boiler i,e Cage wall hopper, ECO , APH and ESP

LOI shall be reported on oven dry basis. In my presence, the bed height was lowered and air flow
was increased. The O2 level was cross checked and found to be OK with the indicated reading by
O2 analyzer. The CO ppm level is 900 with O2 at 5.5%. There can be problem of air ingress from
roof which can guide us wrongly in setting the required minimum excess air. At present the bed ash
bulk density is as high as 1450 kg/m3. The iron content in bed ash is around 14%. This can be
reduced by removing the iron in recycled bed ash. Another issue is the char % in coal. This is
known by calculating the fuel ratio known as Fixed carbon (FC)/ Volatile matter(VM). If the fuel
has petcoke / char mix then also the more unburnt will be seen. Proximate analysis is recommended
for each fuel to know the trend on FC/VM. A shut down inspection can help in fixing the problem.

On 4th day afternoon, the secondary air was set 300 mmWC at the bottom most level. This was not
reducing the unburnt level. The LOI results are to be compared for 5.5% O2 with SA & without SA.
Further SA can be set at 400 mmWC after comparing the results.
Low Furnace gas Plenum temperature than design even at high operating bed temperature 970 - 980
Deg C

The indicated temperature is found to be 100 deg C less when cross checked with a portable
instrument. This can be due to the use ungrounded thermocouple & thermowell. If there is any ash
deposit on the thermowell, then also the reading will be wrong. High temperature drop is not
possible in furnace unless there is again a source of air ingress.

Further recommendations on the parameters for better combustion.

Bed height

The bed height was seen to be as high as 750 mmWC before alteration. This was done as per advice
from OEM. This has led to frequent PA line choking. Every time the furnace is disturbed, naturally
combustion conditions will not be ideal. During the disturbances, the fuel would be consumed
additionally.

The bed height required for the immersion of the bed coils is about 350 mmWC. Further height
requirement is based on possible improvement in carbon burn up. Otherwise high bed height does
not lead to additional steam generation unless & until the waterwall takes part in heat transfer with
a good heat transfer coefficient. A refractory lined waterwall has very less overall heat transfer
coefficient. It is advised to optimize the bed height between 500 mmWC to 550 mmWC. On 16th
evening the bed height was brought down to 620 mmWC. It is recommended to hold the bed height
at 550 mmWC for few days to stabilize the combustion data.

Deeper bed above 550 mmWC is not required in any boiler. It is unnecessary to operate the boiler
at deeper bed. It gives difficulty in pumping the required rice husk in to the boiler. In fact, at 650
mmWC, the venturi does not work smooth. There will be line frequent line choking. All other
boiler makers in the field design the bed for a maximum of 500 mm bed height only. For this
reason, the bed coils are sloped at 4 deg by one experienced boiler maker, in order to obtain the
required immersion surface within the bed. Heat balance calculations are done to finalize the
amount of bed coil surface in the immersed portion of the bed. When the bed heating surface is
insufficient the bed temperature shoots up or else the excess air will have to be increased to cool the
bed.

The next day, the bed height was brought down to 550 mmWC. There was no change in generation.
In fact, the average steam generation has improved because whatever the coal fed was now burning
& releasing heat. There is saving in FD fan power when the fan head comes down. There was no
deterioration in unburnt due to reduction in bed height. This remained in the order of 14 to 18%.

Oxygen in flue gas

The oxygen in the flue gas was found to be as low as 2.5% as observed in the first day. The Oxygen
in flue gas has to be 4.5 % for complete combustion. This value corresponds to excess air of 25%.
If we increase the excess air from 25% to 40%, the reduction in efficiency is to the tune of 0.5%.
The improvement due to unburnt carbon loss will be more pronounced if we help the carbon to
burn up by providing additional air.

The oxygen which is measured has to be error free at first. If there is any air ingress any where the
indicated oxygen is wrong. The furnace will be actually starving where as the oxygen will be with
leaky doors & penthouse. It was seen that all the inspection doors at Superheater area were found to
be passing. Inspection of boiler no 1 revealed that there is air ingress in penthouse & seal box of
secondary superheater header area.

It is recommended to inspect the boiler penthouse when the boiler is shut. Any leakage shall be
arrested. Since the gas temperature is only 450 deg C at cage wall outlet, the O2 analyzer may be
shifted to cage wall outlet so that air ingress downstream will have no effect.

CO ppm

The CO ppm is to be measured for admitting the required secondary air flow. The CO ppm was
measured to be 1400 ppm on 16th November. This was too high. What is indicated is actually a
diluted value with air infiltration present in the boiler.

On 3rd day after increasing the air flow, this was measured to be 900 ppm. This can be further
brought down by operating the boiler on a steady O2 % above 4.5. The CO level was found to
reduce as the Oxygen is increased. Lowest achieved ppm in other boilers is 250 -350 ppm without
secondary air arrangement. With a steady value of O2 at 5.5% and with SA at 300 mmWC at lower
tier, the CO ppm remained at 830 at the time of my departure. As advised, secondary air can be
increased to 400 mmWC at the next step. It is advised to procure a portable CO / O2 instrument to
enable the operating team to try various combinations in order to reduce the CO ppm. The iron
reduction in bed ash can make an impact on this.

Secondary air

There was no instrumentation available for knowing the secondary air flow. Bottom tier secondary
air headers were provided with header pressure indications during my stay. In this boiler, final SH
is placed above the furnace. Secondary air can have a dilution effect if the fuel is fully burnt in the
bed itself. The regulation of secondary air is to be based on this effect.

The dampers available at secondary air headers are multi-louver type. It was suspected for leakage.
But the leakage level was checked to be 20-30 mmWC after the pressure tappings were made. This
is OK.

The selection of secondary air injection has to be at one level only at first. Seeing the effect on
reduction of CO ppm or on reduction of SH steam temperature, the air flow has to be regulated. For
high VM fuels, the secondary air admission will help in combustion and increase the free board
temperature. On Indian coal, it may have a cooling effect.

The secondary air setting can be increased in steps to know the extent of its benefit. This exercise
has to be continued level wise. The header pressure can be increased from 300 mmWC to 400
mmWC.

Furnace refractory lining & its effect on main steam temperature

Considering 110 TPH steam generation, the flue gas produced was calculated with a fuel mix of
rice husk, imported coal & Indian coal at a ratio of 10:20:70. It was assumed that the entire
waterwall is unlined for this calculation. The free board HTA is 213.86 m2. With a HTC of 50
kcal/h/m2/ deg C, the heat duty to waterwall will be 6 MMkcal/h. With a bed temperature of 980
deg C, the entire waterwall between the bed & final SH can cool down the gas to a level of 835 deg
C. However, in order to achieve the main steam temperature at a lower load the refractory lining is
required. At present the free board gas temperature read as 685 - 700 deg C. The reading was
counter checked by a meter carried by me. This is found to be +800 deg C.

Bed ash bulk density & iron content

The fresh bed material is reported to have a bulk density of 1050 kg/m3. The bulk density of bed
ash is reported to be +1400 kg/m3. This can be due to iron present in bed ash. Iron has a particle
density of 7850 kg/m3 whereas sand has a density of 2500 kg/m3. Bed ash density is around 2000
kg/m3. Heavier the bed ash particle, the fluidisation quality is affected. The heavier particles lead to
spouting bed – a bed in which air bubbles leave the bed easily pushing out the particles without
creating a fluidizing effect. A mixing effect is seen. The particles are usually seen thrown out from
the manhole. See annexure for photograph.

It is required to control iron content. The source of iron is from the imported coal. The iron is more
in certain mines. Step may be taken to trace the source of mines and to avoid the high iron source.
High iron was experienced in some lots at Madras cements and it led to bed tube failure. Later the
source was changed.

If iron can not be controlled at source, step shall be taken at plant to remove the iron before
recycling the bed ash to furnace. Used bed ash is ideal considering the particle size. Permanent
magnet arrangement shall be for removal of iron. Also, the different type of systems available is
informed. Electromagnets are not working on iron from bed ash. This is for information. Only high
power permanent magnet should be procured. A sample must be sent to vendor before placing the
order.

Bed ash removal method

Bed ash drain gates must be operated in throttled manner. Only then the coarse particles can be
removed easily. When the bed contains a wide range of particles, two fluidization regimes are seen.
See annexure 2 on fundamentals of fluidisation. The heavier / bigger particles fluidise / defluidise
close the air nozzles. When the gates are operated in throttled manner, it helps in removal of coarser
particles.

The frequency of draining is dependent on the ash content of fuel. In a fuel with high ash, the bed
ash is continually removed and hence the average particle size is dictated by the incoming particle
size range. But in a fuel with less ash content, bigger particles stay back in the bed, making the bed
to have lesser bed expansion and poor combustion. In fuels with lesser ash, the coarser particles
must be removed and replenishment with finer particles should be routine process.

The control parameter is either average particle size or the percent of oversize (+3 mm). Whichever
can be analyzed faster & reported in the beginning of a shift shall be considered. A vibratory
laboratory screen may be arranged for the use of boiler operating team.

Fuel line jamming & clearing the lines

Fuel line jamming can occur while handling different types of fuel.

 Stones in rice husk settle at bottom of the fuel pipe. This can be removed by switching of the
 feeder for some time and by opening the drain gate below the cross. While opening the gate,
fire should not be brought in to the pipe. If fire is brought inside the pipe, the choking will be
perennial problem. This is because we promote the burning of the rice husk in ash drain pipe.
 Imported coal has high VM. Simply operating the drain gate & bringing fire to drain pipe will
lead continuous fire inside cross. There will be clinker formation as well.
 Indian coal can settle in the pipe because of bigger stone / shale particles.

For all the above cases, the following practice can be of benefit.

 Never bring the fire in to the fuel drain pipe. Partially open its gate and close even before the
line gets choked. When Indian coal / Rice husk is used, periodical draining can remove the
stones / bigger particles to the drain pipe. When little amount of fuel in the drain pipe is
removed the particles can roll down to the drain pipes. During this type of operation, the fuel
feeding should be stopped. This operation like a proactive approach.
 We can switch of the rotary feeder & drag chain feeder of a compartment for 1-2 minutes. This
can help in clearing the settled particles in to the bed. In a shift, at least once this practice can be
adopted.
 If coal powder lumps are expected and if they are jamming the fuel lines, a bar screen is
recommended at inlet of pocket feeder as in photograph in annexure 1.

Using DP drop as a measurement of air flow

The airfoil meter is a fabricated item and has its limitations in giving accurate reading on flow. As
the combustion calls for excess O2 in flue gas, the same should be provided. The indicated air flow
should not be taken as important parameter. Between the indicated O2 (countercheck is advised) &
air flow meter reading, O2 percentage carries priority at least at low O2 levels (in the range of 3 –
6%).

Another way to ensure the required air flow is given for combustion is to see the DP drop. For a
mix fuel ratio of Rice husk: imported coal: Indian coal of 10:20:70 by weight, the DP drop will be
393 mmWC at full load. This is with an assumption of zero secondary air. If we find the over fire
air of 5% is beneficial, then the DP drop will be 354 mmWC. If the drilled hole is of higher
diameter than 3.2 mm, then the actual DP drop will be less than this.

Leakages in roof & seal box

Since boiler no 1 was under shut at the time of the visit, the roof was inspected. On inspection from
hot side of the roof panel, the condensation of gases is seen. This can happen when petcoke was
fired. Petcoke has high sulfur. In case of cold air ingress sulfur corrosion is seen. If this is the order
of leakage in boiler no 2, the O2 can be wrong. The leakages must be inspected for & rectified.

Over-bed feed provision of rice husk

The provision for over-bed feeding of rice husk can be discussed with OEM. This will help to attain
the SH temperature at lower load. We can review this factor after a thorough shut down inspection.

GCV and accounting of coal

The GCV reported by lab was on oven dry basis. Based on the as fired moisture at furnace / final
conveyor, the as fired GCV shall be computed. Similarly, for accounting purpose, the total moisture
as received should be immediately analyzed on receipt. Now as received GCV can be calculated.
There is always loss of moisture between receipts to firing. Hence there will be large reduction
between receipt & consumption. Attempt to tally the receipt weight and issue weight is to be made
incorporating a correction factor for drying in yard.

K.K.Parthiban
Boiler specialist / venus energy audit system
Annexure 1: Observations at plant in boiler no 2 & 1
Photo 1: The manholes fixed at SH area are not seal proof. Ash spilling is seen. This leads to air
ingress. The Oxygen as reported by O2 analyzer can not be the true value. It is logical to set air
conditions for

Photo 2: Shows the inspection door design by TBW.

Figure 1: As an alternate to cast doors, fabricated doors with refractory protection can be used for
inspection doors to prevent air ingress.

Photo 3: Flange leaks shall be attended. Though it is not related to performance, it can affect the
power consumption.
Photo 4: Secondary SH seal box is seen filled with ash in boiler no 1. The leakage location shall
be checked and attended. Similar check shall be done in boiler no 2 during shut. With air ingress
the oxygen will be wrong.

Photo 5: Photo shows the inside of penthouse in boiler no 1. Ash spillage is seen in penthouse.
Again this is a sign of air ingress. This must be attended in boiler no 1 & checked for in boiler no
1.
Photo 6: It is seen in boiler no 1, the critical spring hanger is not released yet. This can cause
strains in welds in the panel fin to tube joints.

Photo 7: The bed coils that are removed from boiler no 1 is seen polished in between the studs.
Such erosion pattern is seen when the iron content / coarse particles in the bed ash is more &
when they fluidise / defluidise at the bottom of the bed.
Photo 8: Inside the boiler no 1, the gasket in flange joint has failed. This can lead to carry over.
It is advised not to use gasket. Instead thin SS shim can be used for the flange joints.

Photo 9: water splashing is seen inside the steam drum near the steam drier assembly in boiler no
1. Similar inspection shall be carried out in boiler no 2.
Photo 10: When Indonesian coal is used in high percentage the iron concentration can be
harmful to the bed coil. It can be seen here that the slagging occurs above the fuel feed point.
This photo is from a different boiler. Iron control in bed material will call for automated system
for iron removal in bed ash before recycling back to bed.
Photo 11: When the bed ash slags, it is seen to erode the bed coils. This was observed in many
plants. The above picture is from a cement plant which used Indonesian coal without exercising
the control on iron in the bed ash / bed material.

Photo 12. The photo shows the spouting characteristics of a fluidised bed, when it contains
coarser particles. The particles are thrown out through the furnace doors. This photo is from a
different plant where the particles contained iron % & large particles as well. The efficiency is
affected by this.
Photo 12: The same spouting is seen here with oversize particles coming out from the doors.
Control of oversize particles is required for efficient combustion.

Photo13. The high density particles present in the bed, they erode the tubes in between the studs
in this fashion. The same pattern of erosion is seen in boiler no 1.
 Figure 2: Geldart classification of materials.

Geldart classic classification of particles

Not every particle can be fluidised. The behaviour of solid particles in fluidised bed depends
mostly on their size and density. A careful observation by Geldart (1973) is shown in figure 2.
` There are four different types of materials categorized.

 Group A- these are designated as ‘aeratable’ particles. The particles have small mean
particle size (d p<30µm) and or low particle density (<1.4g/cm3). Fluid cracking catalysts are
in this category. These solids fluidise easily, with smooth Fluidisation at low gas velocities
without the formation of bubbles. At higher gas velocity, a point is reached when bubbles
start to form and the minimum bubbling velocity, Umb is greater than Umf.
 Group B –these are called ‘sand like’ particles. Most particles of this group have size range
of 150µm to 500µm and particle density is from 1.4 to 4 g/cm3.
 Group C-these particles are cohesive or very fine powders. Their sizes are less than 30µm.
They are extremely ‘difficult to fluidise’ because inter particle forces are relatively large.
Examples of group C materials are talc, flour and starch.
 Group D –these materials are called, ‘Spoutable’ and the materials are either very large or
very dense. They are difficult to fluidise in deep beds. As the gas flow is increased, a jet is
formed in the bed and material may then be blown out with the jet in a spouting motion.
Roasting metal ores are examples of group D materials.

Geldart’s classification is clear and easy to use as displayed in above figure for Fluidisation at
ambient conditions and for U less than about 10*Umf. For any solid of known density and mean
particle size this graph shows the type of fluidisation to be expected.

Note 1: Iron containing bed ash will be heavier since the iron has a particle density of 7850
kg/m3 as compared to boiler ash which is around 2000-2500 kg/m3. They start spouting in bed
affecting the combustion efficiency. Particles escape out easily out of the bed.
Photo 14: It appears there is air ingress in sling tube seals at roof in boiler no 1. The roof is
filled with fly ash. If this is the condition of the boiler at roof, then O2 indicated is with error.

Photo 15. Ash craters around the sling tube in roof means there is air ingress. Slowly it leads to
tube erosion as well. The leakage shall be rectified seeing the drawing.
Photo 16: The roof panel is showing condensation. If the boiler penthouse receives high air
ingress, this sort of condensation is seen in the fins.

Photo 17: Yellow deposits usually sulfur condensation. If there is cold air ingress this
condensation occurs in the vicinity of air leak spots. The difference in color in the roof panel
indicates that there is air ingress in the roof. The condensation can happen when petcoke is
fired.
Photo 18: This is the sealbox of final SH coils at front wall. This area is to be welded as per
drawing.

Photo 19. This photo is also the same seal box, but another location. Here the welding is carried
out.
Photo 20: On inspection of the boiler no 1, it is seen that the primary SH coils have come off the
supports. This needs to be rectified to avoid erosion.

Photo 21. The photo shows the incomplete welding of the corner of the front waterwall panel to
side wall panel. Air ingress can lead to erosion of tubes / acid condensation damage in case of
high sulfur fuel.
Photo 22: It is seen that the PSH tubes pitching is not maintained. This shall be corrected to
avoid preferential erosion. The guide provided at the rear wall is not used at all. Drawing shall
be referred and corrective action shall be taken.

Photo 23. The PSH B to PSH C coils are not aligned because there is no clamping / alignment
system for top most coils. This may be corrected in consultation with OEM. This can lead to
erosion of out of lane tube over a period.
Photo 24: Bunkers do make lumps due to moisture & powder. These can be filtered above the
feeder thus avoiding the choking of the venture.
Annexure 2: Technical article on art of bed management in FBC
ART OF BED MANAGEMENT IN AFBC BOILERS

Fluidisation had been extensively researched and brought out in many papers. The fluidisation
which as originally used for chemical reactors was extended to combustion technology. It had been
nearly 30 years since the fluidisation had been commercially applied for combustion. Many
manufacturers design and supply boilers with Bubbling fluidised bed combustion technology.
Many fluidised beds are operating successfully. There are some installations wherein the
fluidisation misbehaves leading to poor combustion and bed coil failures. The subject of
fluidisation is reviewed here for the benefit of operating engineers.

The behaviour of a fluidised bed is difficult to predict since the parameters such as the particle size,
shape and density are not simply some numbers. These parameters in real world fall in to wider
range. There are many Fluidisation regimes which have been well explained in Fluidisation
engineering by Kunii & Levenspiel.

Review of Fluidisation basics

What is Fluidised Bed?

When air or gas is passed through an inert bed of solid particles such as sand supported on a
perforated plate, the air, initially, will seek a path of least resistance and pass upward through the
sand. With further increase in the velocity, the air starts bubbling through the bed and the particles
attain a state of high turbulence. Under such conditions, the bed assumes the appearance of a fluid
and exhibits the properties associated with a fluid and hence the name 'Fluidised Bed'. If velocity is
too low, Fluidisation will not occur, and if the gas velocity becomes too high, the particles will be
entrained in the gas stream and lost. Hence, to sustain stable operation of the bed, it must be
ensured that gas velocity is maintained between minimum Fluidisation velocity and particle
entrainment velocity. Fluidisation is widely used for many commercial operations, such as
transportation, heat treatment, absorption, mixing, combustion, chemical reactions.

Regimes of Fluidised bed combustion (source: Fluidisation engineering-Kunni & Levenspiel)

When the solid particles are fluidised, the fluidised bed behaves differently as velocity, densities of
gas & solid particles are varied. It has become evident that are number of regimes as shown in
figure 1;

Regime A- Fixed bed- the particles are at rest.

Regime B- Minimally fluidised bed: The upward moving gas is able to overcome the gravitational
force on the particles. The voidage increases slightly.
Regime C- As the velocity is increased further, the bubbles are generated. The bubbles coalesce and
grow as they rise to the top of bed.
Regime D-If the ratio of the height /diameter of the bed is high, the size of the bubbles becomes
equal to the diameter of the bed. This is called slugging.
Regime E-When the particles are fluidised at a high enough rate, the upper surface of the bed
disappears. Instead of bubbles one observes turbulent motion of solid clusters and voids of various
sizes and shapes.
 Figure 1: Different regimes of fluidisation
Regime F- With further increase in gas velocity, eventually the fluidised bed becomes an entrained
bed in which the pneumatic transport of solids take place.
As you might have guessed we are supposed to operate in regime C. But invariably we may land in
regime E, where the particles have become coarser. Even coarser particles are seen getting out of
the bed.
Geldart classic classification of particles
Not every particle can be fluidised. The behaviour of solid particles in fluidised bed depends mostly
on their size and density. A careful observation by Geldart (1973) is shown in figure 2. There are
four different types of materials categorized.

Figure 2: Geldart classification of particles (Geldart-1973)

• Group A- these are designated as ‘aeratable’ particles. The particles have small mean particle
size (dp<30µm) and or low particle density (<1.4g/cm3). Fluid cracking catalysts are in this
category. These solids fluidise easily, with smooth Fluidisation at low gas velocities without the
formation of bubbles. At higher gas velocity, a point is reached when bubbles start to form and
the minimum bubbling velocity, Umb is greater than Umf.
• Group B –these are called ‘sand like’ particles. Most particles of this group have size range of
150µm to 500µm and particle density is from 1.4 to 4 g/cm3.
• Group C-these particles are cohesive or very fine powders. Their sizes are less than 30µm. They
are extremely ‘difficult to fluidise’ because inter particle forces are relatively large. Examples of
group C materials are talc, flour and starch.
• Group D –these materials are called, ‘Spoutable’ and the materials are either very large or very
dense. They are difficult to fluidise in deep beds. As the gas flow is increased, a jet is formed in
the bed and material may then be blown out with the jet in a spouting motion. Roasting metal
ores are examples of group D materials.

Geldart’s classification is clear and easy to use as displayed in figure 2 for Fluidisation at ambient
conditions and for U less than about 10*Umf. For any solid of known density and mean particle size
this graph shows the type of fluidisation to be expected.

Fluidisation velocity
Many fluidisation velocity correlations are available in text books. A typical correlation found in
fluidisation engineering by Kunii & Levenspiel is presented below.

g (ρs − ρ g ) ε m3 1.75 D pVm ρ g 2 150(1 − ε m ) D pVm ρ g D p3 ρ g ( ρ s − ρ g ) g

Vm ≈ ⋅ φs 2 D p2 3
( ) + 3 2
( ) = 2
150 µ 1− εm ε mφs µ ε mφs µ µ
D pVm ρ g D pVm ρ g
when, < 20 when, > 20
µ µ
The above equation is usable for The above equation is usable for coarser particle system
smaller particle system

Legend in above equations are as below:

Vm = Minimum Fluidisation velocity g = acceleration

εM = Void fraction ρp = density of particles
ØS = Sphericity Dp = Diameter of the particle
µ = viscosity of gas ρg = density of gas

We need to notice the following in the above equations-

**Min Fluidisation velocity is proportional to Dp^2 or Dp^3 – depending on the particle size.
**Min Fluidisation velocity is proportional to particle density.

The above equations are based in narrow particle distribution. In real life fluidised bed system, we
have wider particle size distribution. For example in coal preparation plant from one plant to
another we may not have same particle size distribution. From mines to mines the coal ash will be
different in terms of composition and thus particle density is going to be different. In some case, we
may have complete segregation of particles in to a region of predominantly small particles and a
region of predominantly larger particles. The segregation is characterized by the abrupt change in
porosity. The picture is same in case the bed has two different particle densities. The picture is
shown in figure 3. When the particle sizes fall into a relatively narrow group we would have an
intermediate case, as shown in figure 4. The larger size particles (or denser particles) are found
grouping at bottom. There is zone where we have a mix of particles. The above theoretical
exposure would now help in analyzing our problems in fluidised bed combustion boilers.

Figure 4: Fluidisation with wider (not too wide)

Figure 3: Fluidisation with wider particle
particle distribution / wider particle density
distribution / wider particle density difference
difference

Real life Fluidised bed combustion system

In a fluidised bed combustion system, the fuel / bed material is a source of the particles. In the case
of coal, the burning particles & and the burnt particles are the main contributors which decide the
fluidisation behaviour in the bed. The bed material which is used for start up of a bed may be
particles which have been sieved from the ash obtained from Fluid bed combustor or it is generally
sieved river sand / sieved crushed refractory bricks...

About sand

When sand is used as bed material, particles less than 1.5 mm should be used. These particles are
generally round in shape. When the sand particle happens to be on the higher size, we can notice
that the sand is sharp in nature. Such sand particles are found to be erosive when the bed is not
properly fluidised. It is not necessary to remove the finer portion of sand. The sand size range of 0-
1.5 mm gets the best combustion of fuels in many installations. Sand is practically a crystallized
material and does not have pores.

About crushed refractory

The refractory bricks are made from recycled bricks & fresh refractory clay. The old refractory
bricks dismantled from ore melting furnaces and many other refractory furnaces are crushed to
minus 3mm as part of raw material preparation in refractory manufacturing industry. The refractory
grog prepared in this way is screened in 20 & 8 mesh and the material that is lying in between is
used as bed material for FBC. This bed material is found to be less erosive. As the density of this
material is lesser as compared to sand, the size range is 2.35 to 0.85 (corresponding to mesh
numbers 20 & 8). Higher size has been permitted here as the particle density is lower due to
presence of pores.

About fuel Particle size- coal fired boilers

Even if the sand / refractory bed material conforms to required size range, over a period of FBC
operation, the entire particles are going to be replaced with the ash particles generated from fuel.
Thus the fuel sizing is very critical for trouble free FBC operation. Ignorance to maintain screen
system at coal handling plant can lead to accumulation of oversize particles. Once the over size
particles are accumulated, the bed goes for spouting. Even the heavier particles get thrown out of
the bed. The loss on ignition would go up if the furnace residence time is less. Otherwise we may
not notice it. Again if the coal is reactive with high volatile matter, the unburnt in ash may not go
high. In the case of under bed firing, the erosion of bed coils is accelerated due to violent
turbulence at fuel feed points. The generation would come down as the particle size increases. This
is due to the fact that smaller particles have more surface area to conduct the heat to bed heat
transfer surfaces.

About fuel Particle size- agro fuel fired boilers

In the case of agro fuels such as rice husk, de-oiled bran (DOB), ground nut shell the ash from fuel
does not contribute to bed particle size. Whatever the bed material used gradually disintegrates and
the fines go out of bed. Regular addition of bed material is required in order to maintain the bed
height. DOB fired boilers experience peculiar problem of heavier particle generation due to melting
of ash. DOB ash melts and agglomerates to over the sand / bed material. More sand will be required
to offset the bed particles becoming heavier. In general agro fuels are having lesser density as
compared to bed material and they try to leave the bed earlier. Using finer material helps to achieve
a good bed expansion and binds the fuel particles better.
Stone ingress in husk fired boilers is well known among old installations. More drains had been
added in some cases to bring out the stones. The stone removal by mechanical screen system is not
effective. Only way to get the best from these combustors is to resort to frequent draining and
recharging screened bed material. Fine sand is the right option for husk fired boilers as the
disintegration of the refractory type bed material could increase the operational cost.

About fuel Particle Dolachar fired boilers

Dolachar & coal are fired together in boilers used in sponge iron industries. The Dolochar contains
iron particles which are heavier. As one can expect the bed can have segregation of particles to
bottom of bed. In many cases, the operators ignore this. Once a bed lands in to segregation of
particles, there is no way out in the case of flat distributor plate designs. Only open bottom design
can pull out the oversize particles which settle at bottom. Alternately sloped distributor plate / DP
with many drains can help to control the particle separation. Particles separation when takes place
preferential erosion of bed coils is seen.

In these combustors, the generation of bed ash is also more. We may not find a requirement to add
bed material. Except for a fresh start up we may not use iron free bed material. Bed coils in these
boilers call for early replacement and the availability of the boiler is greatly disturbed due to
unscheduled shut downs. In under bed feed system, the erosion of bed coils is localized whereas in
overbed the erosion is not so, provided we keep removing heavier lot. If heavier lot is allowed to
accumulate, the bed coil erosion is of different kind. The coil is seen eroded in between studs to as
in photograph 4. There is a gross erosion of bed coil. All these troubles are related to iron
accumulation. They solution could be a continuous bed ash iron removal system. The bed ash must
be cooled and discharged to an automated magnetic separator system & over size screen system.
The screened bed material should be recharged to bed through a bed material silo and feeder
system. In a case where I had visited, the plant in charge had resorted to magnetic plate for removal
of iron from bed ash. See photograph 5. Yet the rate of accumulation was so high the iron % in bed
ash did not come down. The customer was advised to resort to continuous bed ash removal and
recharge system. More bed drains have been advised so that the removal is effective.
Case study 1- related to Fluidisation

This is a 90 TPH coal fired AFBC boiler with over bed feed arrangement. The customer was
restarting the boiler after a shut down from cold condition after 5 days shut period. The bed
material was not changed. Whatever remained before the shut had been used as such. Naturally we
can expect that particle range will be right from fly ash dust to coarser ash. As this boiler was
provided with over bed start up burners, the height of the bed material had to be kept to 600-700
mm in order to cover up the bed coil. The burner was put on for 3 hrs to raise the boiler pressure
slowly and to promote circulation in bed coils & other evaporative circuits. In over bed burner start
up system, the bed never gets heated up beyond 150 deg C. The airbox pressure was raised to 800
mmWC and the coal feed was done. The powder content in bed material got separated to the top
and gave a picture as if the bed was fluidizing. The bed plate design was checked for MCR pressure
drop. It worked out to be 300 mmWC. That means nearly 900 mmWC air nozzle pressure drop will
be required to set a cold Fluidisation. When the FD fan can not give 1500 mmWC, the question of
thorough fluidisation can not be achieved. There were temperature differences between bed
thermocouples even after 12hrs after start up. In this situation whatever fluidisation which was
observed was clearly a regime of fluidisation with two sets of particle groups. Now to correct the
situation I simply drained the bed material from several drain points simultaneously to bring out the
coarser fraction of material at the bottom. After 4 hrs, we had seen that all the bed temperatures
read the same value. Further, the customer was advised to use finer bed material. I recommended
minus 1.5 mm sand. I advised not to discard the finer fraction of bed material that is less than 0.5
mm.

Case study 2- related to Fluidisation

This is a coal fired AFBC boiler with under bed feed arrangement. Within 2 months of changing
over to new lot of imported coal, the plant suffered bed coil erosion. The unit experienced severe
iron accumulations in bed ash. It was as high as 33%. The source was found to be the high iron
content in coal ash itself. A model calculation done proved that we need to increase the fresh bed
material feed rate to 25% of fuel feed in order to limit the iron in bed ash to 5.6%. Incidentally
more the iron content in coal, the ash fusion temperatures could be lower. The bed may begin to
generate clinkers as well (see photo).

Table 1- Ash analysis summary – Fe2O3 has increased in recent coal supply.

Ash Bed ash at Bed Coal used before Failure Coal –used
constituents present material Coal 1 Coal 2 of late
SiO2 27.11 53 48.27 56.79 35.09
Al2O3 15 34.5 26.31 27.36 19.46
Fe2O3 33.4 0.8 6.12 6.19 29.88
CaO 0.234 3.5 9.65 3.19 5.28
MgO 0.56 2.1 1.4 0.53 1.91
Na2O 2.898 - 1.2 0.81 1.34
K2 O 0.31 - 0.87 1.41 2.2
SO3 - - 0.244 0.2 0.49

Case study 3- related to Fluidisation

This was a rice husk fired boiler. The customer experienced severe polishing of bed tubes. The
steam generation had come down drastically. Rice husk fired boilers need regular make up of bed
material. When I visited I found the bed ash drained from the bed was heavier. When checked up
with a magnet we noticed that the bed ash contained 40% iron. The source of iron was found to be
the bed material itself. The crushed refractory as explained earlier is made from crushing used
bricks removed iron melting furnaces. Naturally the iron content in the crushed material has to be
more. The vendor did not separate the iron which was the normal process used for making
refractory bricks to meet the IS 8 specification bricks. The customer was advised to change over to
fine river sand.

Case study 4- related to Fluidisation

This case is a 100 TPH Boiler which was designed for high DP drop operation by design. The cold
Fluidisation was checked by me in person. With a bed height of 300 mm and with air box pressure
of 1000 mmWC, the bed did not fluidise. I requested the customer to change over to finer bed
material for start up. I explained that the bed would not fully fluidise. I have enclosed photos 5 & 6
which proved that the bed did not fluidise even after 12 hrs after the coal feeding was commenced.
I recommended that fresh fine bed material had to be added to large extent simultaneously drain the
bed that did not fluidise initially. Cold fluidisation is being ignored by many in the recent boilers
which were provided with overbed start up burners. When the bed is designed with hot air
generator this would not be a problem. Had the boiler been provided with charcoal start up burner
lances, the boiler would have never started. It would have experienced clinkering every time. In
fact I had attended a case where the client got the entire set of air nozzles replaced with higher
diameter holes to have the cold fluidisation.

Final word

Boiler operators need to realize the importance of the right size of bed material with the right
particle density is important for a thorough Fluidisation. Use of mechanized bed ash / bed material
screen system, regular sieve & ash analysis of fuel, regular measurement of sieve analysis bed ash,
frequent check in iron content of bed ash & bed material would help to improve get the best from
the AFBC boilers. Manufacturers need to realize that the additional drains are a must for smooth
operation of FBC boilers. High DP drop in bed would give problems in setting the cold
Fluidisation. Cold Fluidisation inspection should be part of start up process and can not be
compromised.
 Photo 2- heavier particles collected at
Photo 1-heavier particles are thrown in bed inspection door.

Photo 4- bed coil erosion in overbed fed

boiler. The heavier particles settling at
Photo 3-Open bottom furnace originally bottom of the bed has caused erosion.
designed by B&W

Photo 5-Improper fluidisation Photo 6- Improper fluidisation